chapter
stringlengths 1.97k
1.53M
| path
stringlengths 47
241
|
|---|---|
International production has shaped the career trajectories of film professionals in specific ways. Organizational concepts such as boundaryless careers 36 and semipermanent work groups 37 go some way to explaining how this phenomenon has taken shape; however, these are limited as explanatory frameworks because they do not take into account the transnational processes that accelerate some workers’ careers while restricting others to low-level positions, particularly those specializing in major Anglo-American productions. The latter find themselves in the paradoxical position of being well-paid mobile workers, thanks in part to a lack of union regulations, but with little chance of professional upward mobility. They remain trapped in a segregated work world, deprived of either the financial incentive to work on local productions or any realistic chance of the type of career development enjoyed by the foreigners running the international productions on which they work.
American-born production managers are often fast-tracked. They typically skip arguably the two most challenging career steps: being given access to the industry and being socialized in aspects of it.38 Instead, they acquire prized locational knowledge and develop marketable specializations at a rate impossible in Western media hubs like London and Los Angeles. As Minkowski put it, “I could have gone back to LA and become one of thousands fighting to work on films, or I could stay here and strike out on my own.”39 By contrast, for local production management, the collapse of the old hierarchical state-owned studios brought uncertainty and unemployment, but a rapid generational change granted some in their ranks swift access to the industry. The fortunate ones developed hybrid professional identities, claiming to “behave like Americans” without leaving their homeland.
To gain insight into the differences and mediating mechanisms that underpin communities of cultural workers, we can benefit from the self-reflexive comments of Czech personnel. Even those struggling to progress in the industry highlight the experience of learning rather than the feeling of being exploited. This sentiment is bound up with their construction of hybrid cultural identities. Thus the Czech soundman Petr Forejt describes himself as becoming an American filmmaker in Prague, distanced from the trivialities of a local industry in which wages and standards are low and improvisation and multitasking high.40 Similarly, Milan Chadima, a camera operator who has worked on such projects as The Brothers Grimm (2005), spoke of American producers helping him escape the frustrations of shooting low-budget Czech films and commercials.41
These cases notwithstanding, it is clear that the careers of even the most successful Czech service production workers are characterized by striking limitations. Such individuals are not promoted to higher creative positions like department heads. They work in other international media hubs only when their employers move a project overseas and rarely take part in prestigious domestic projects. Coming closest to the privileged position of the department head were several Czech art directors, yet only one, Ondřej Nekvasil, has built what could reasonably be considered a career of international standing. Nekvasil switches between working on Czech art-house fare, teaching production design, and working as a production designer on international productions like The Illusionist (2006) and Snowpiercer (2013). Two factors underwrite Nekvasil’s distinctive transnational career trajectory. A reputation-making Emmy for Anne Frank: The Whole Story (2001) brought him to the attention of American producers such as David R. Kappes, who hired him for the Sci-Fi Channel miniseries Children of Dune (2003). He is also fortunate to specialize in the aspect of local production services most valued by American producers—set design and construction, which, in spite of its high standards of craftsmanship, can be obtained 50 percent cheaper in Prague compared to Los Angeles. I asked Nekvasil what he felt sets him apart from those art directors who also work on medium-to-big-budget productions but have failed to match his level of professional success. Nekvasil said nothing of differences in skill, but instead suggested that they may prefer the relative calm of the art department over the greater responsibility of face-to-face interaction with foreign producers.42
7.06: Conclusion A Two-Tier Department
To gain a better understanding of the contemporary production world of Prague, we require a more balanced approach than those focusing primarily on the supposed exploitation of the global labor force, as neo-Marxism does, or on city development strategies, as creative industries and cluster theories do. Cultural intermediaries, knowledge transfers, and learning effects play major roles in a postsocialist, non-English-speaking country like the Czech Republic. As a result of historically specific experiences—communists discrediting labor unions, the interventionist yet selective cultural politics of the state—local film workers tend to contradict conclusions derived from studies of cultural imperialism or NICL. They criticize local policymakers rather than Hollywood producers and focus on learning and mobility barriers rather than exploitative working conditions. This is true even of individuals whose livelihood is threatened by Hollywood moving runaway productions to neighboring countries like Hungary. A new model of globalization is clearly needed if we are to gain deeper insight into the interplay between global forces “from above” like GPN’s “flagships” and those from below, such as local workers. As economic geography has shown, we also need to understand the relationships between local and translocal transactions,43 whose interaction allows for extralocal knowledge flows. In the case of the Prague screen industries, such an approach might involve examining mediating mechanisms and agencies like the service production sector in terms of their interaction with local and international partners and competitors. The production culture of Prague is effectively a two-tier system split between production services and domestic productions, which are characterized by different salaries, career patterns, and work practices. Recognizing it as such opens up new avenues of investigation. We might, for example, consider the extent to which this instance of multitrack globalization precipitates “departmentalized” thinking, especially in service productions. We might also wish to consider the implications of the “glass ceilings”44 that have prevented many local workers from moving into original projects and securing high-level creative jobs. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/07%3A_Transnational_Crews_and_Postsocialist_Precarity__Globalizing_Screen_Media_Labor_in_Prague/7.05%3A_Career_Patterns_and_Precarity_in_.txt |
This chapter is based on a long-term research project focusing on the history of state-socialist and postsocialist production systems and production cultures in Prague and East-Central Europe, supported by the Czech Science Foundation (grant nr. P409/10/1361) and by the Fulbright Scholar Program. It is also based on data generated by the EU-funded FIND Project, which uses student internships in production companies to combine job shadowing with ethnographic research of production cultures (see www.projectfind.cz). The author would like to thank Richard Nowell for his editing assistance.
1 The interviews were conducted in two phases: in 2009 and 2010 in Prague and Los Angeles, and in 2013 and 2014 in Brno, Prague, and Budapest. The questions solicited responses about the involvement of specific personnel in international productions and how working on these projects affects career trajectories. The interview subjects were: Ludmila Claussová (film commissioner, Prague); Radomír Dočekal (Barrandov studio’s former president, Prague); Petr Forejt (sound recordist, Prague), Daniel Frisch (production manager, head of a production-service firm, Prague/Los Angeles), Thomas Hammel (producer, executive producer, Los Angeles); Michael Hausman (executive producer, first assistant director, New York); Tom Karnowski (production manager, producer, Los Angeles), Aleš Komárek (production manager, a head of a production-service firm, Prague); Tomáš Krejčí (head of a production-service firm, Prague/Los Angeles); Cathy Meils (former Variety correspondent in Prague); David Minkowski (production manager, head of a production-service firm, Prague/Los Angeles); Ondřej Nekvasil (production designer, art director, Prague); Steven North (producer, executive producer, Los Angeles); Rusty Lemorande (producer, Los Angeles); Cathy Schulman (producer, Los Angeles); Steven Lane (producer, Los Angeles); William Stuart (Barrandov Studios representative in Los Angeles); Jaromír Švarc (art director, Prague); Michelle Weller (former production manager in Prague, currently out of film business in Texas); Tomáš Hrubý (producer, Prague); Gábor Krigler (head of development, HBO, Budapest); Viktória Petrányi (producer, head of a production-service firm, Budapest); Pavel Strnad (independent producer, Prague); Petr Bílek (head of a production-service firm, Prague); Viktor Tauš (director, Prague).
2 From mid 2012 to mid 2014, the EU-funded FIND Project (www.projectfind.cz) organized over one hundred student internships. As assistants for international film and television productions, interns conducted participant observations and kept field diaries.
3 See Toby Miller, Nitin Govil, John McMurria, Richard Maxwell, and Ting Wang, Global Hollywood 2 (London: British Film Institute, 2008).
4 See Ben Goldsmith and Tom O’Regan, The Film Studio: Film Production in the Global Economy (Lanham, MD: Rowman & Littlefield, 2005).
5 Neil M. Coe, Peter Dicken, and Martin Hess, “Global Production Networks: Realizing the Potential,” Journal of Economic Geography 8 (2008): 271–295. For examples of applying GPN theory to global media industries, see Neil M. Coe and Jennifer Johns, “Beyond Production Clusters: Towards a Critical Political Economy of Networks in the Film and Television Industries,” in The Cultural Industries and the Production of Culture, ed. Dominic Power and Allen J. Scott (London: Routledge, 2004), 188–204; Hyejin Yoon and Edward J. Malecki, “Cartoon Planet: Worlds of Production and Global Production Networks in the Animation Industry,” Industrial and Corporate Change 19.1 (2010): 239–271; Terry Flew, Understanding Global Media (New York: Palgrave Macmillan, 2007).
6 Dieter Ernst and Linsu Kim, “Global Production Networks, Knowledge Diffusion, and Local Capability Formation,” Research Policy 31.8–9 (2002): 1417–1429.
7 See Candace Jones, “Careers in Project Networks: The Case of the Film Industry,” in Boundaryless Career: A New Employment Principle for a New Organizational Era, ed. Michael Bernard Arthur and Denise M. Rousseau (New York: Oxford University Press, 1996), 58–75.
8 Anders Malmberg and Peter Maskell, “Localized Learning Revisited,” Growth and Change 37.1 (March 2006): 1–18.
9 This claim is based on information gleaned from files at Barrandov Studios archive, the National Film Archive in Prague, a register of secret police agents, and interviews with current production managers.
10 Jonathan Kandell, “Americans in Prague: A Second Wave of Expatriates Is Now Playing a Vital Role in the Renaissance of the Czech Capital,” Smithsonian Magazine, August 2007, www.smithsonianmag.com/people-places/americans-in-prague-160121860/?no-ist=&story=fullstory&page=1.
11 Ibid.
12 Daniel Rosenthal, “Czech Out Hollywood’s New Euro Options,” Weekly Variety, March 9, 1998.
13 See Olsberg SPI, “Economic Impact Study of the Film Industry in the Czech Republic” (London, 2006).
14 “The Hungarian Tax Credit System and the 20% Rebate Scheme,” Studies in Eastern European Cinema 1.1 (2010): 127–130.
15 See government-sponsored reports on audiovisual industry: Olsberg SPI, “Economic Impact Study of the Film Industry in the Czech Republic”; “Strategie konkurenceschopnosti českého filmového průmyslu 2011–2016” (MK ČR, 2010).
16 Will Tizard, “Czech Republic Sees Jump in Foreign Production Spending as Incentives Grow,” Variety, July 9, 2014, http://variety.com/2014/film/news/cz...ked-1201259106; see also Goldsmith and O’Regan, The Film Studio.
17 Cathy Meils, “Czech Film Tax Incentives Shrink,” Filmneweurope.com, June 24, 2013, www.filmneweurope.com/news/czech/105714-czech-film-tax-incentives-shrink/menu-id-150.
18 Leo Barraclough, “Film and TV Production Heats Up in Budapest: Flowing Coin Attracts Hollywood Films,” Variety, May 16, 2013, http://variety.com/2013/film/interna...nbc-1200466459.
19 Alexandra Zeevalkink, “Hungary Ups Film Tax Rebate to Attract More Foreign Productions,” KFTV.com, July 17, 2014, www.kftv.com/news/2014/07/17/hungary-ups-film-tax-rebate-to-attract-more-foreign-productions.
20 See “Barrandov Studios,” www.barrandov.cz/clanek/reference-zahranicni-film.
21 Based on interviews with Cathy Schulman, October 23, 2009; Tom Karnowski, December 4, 2009; Radomír Dočekal, June 5, 2009.
22 See Petr Szczepanik, “The State-Socialist Mode of Production and the Political History of Production Culture,” in Behind the Screen: Inside European Production Cultures, ed. P. Szczepanik and Patrick Vonderau (New York: Palgrave Macmillan, 2013), 113–134.
23 Interview with Pavel Strnad, December 7, 2012.
24 Interview with David Minkowski, May 19, 2009.
25 Ibid.
26 Interview with Tom Karnowski, December 4, 2009.
27 Spillover effects are highlighted by reports on the potential effects of the Czech incentive program commissioned by the government. See, for example, an industry report, EEIP, “Hodnocení dopadů regulace (velká RIA) k části návrhu zákona o kinematografii vztahující se k úpravě pobídek filmovému průmyslu (fiskálním stimulům)” (Prague, 2009).
28 Project FIND student intern field diaries.
29 Interview with David Minkowski, May 19, 2009.
30 Petr Szczepanik, “Globalization through the Eyes of Runners: Student Interns as Ethnographers on Runaway Productions in Prague,” Media Industries Journal 1.1 (2013), www.mediaindustriesjournal.org/index.php/mij/article/view/23/67.
31 Q&As with producers Karla Stojáková (Axman), Ondřej Beránek (Punk Film), and Pavel Berčík (Evolution), 2012 and 2013.
32 See Gary R. Edgerton and Jeffrey P. Jones, The Essential HBO Reader (Lexington: University Press of Kentucky, 2008).
33 Antony Root, quoted in Gary Smitherman, “The HBO Treatment,” C21Media, June 19, 2012, www.c21media.net/the-hbo-treatment/.
34 Informal interview with Tomáš Hrubý, March 10, 2014.
35 Interview with Gábor Krigler, March 28, 2014.
36 Jones, “Careers in Project Networks.”
37 Helen Blair, “You’re Only as Good as Your Last Job: The Labour Process and Labour Market in the British Film Industry,” Work, Employment and Society 15.1 (2001): 149–169.
38 See Jones, “Careers in Project Networks.”
39 Minkowski, quoted in Kandell, “Americans in Prague.”
40 Interview with Petr Forejt, June 17, 2009.
41 Quoted in Tomáš Baldýnský, “Quentin za to může!” Reflex, February 17, 2006.
42 Interview with Ondřej Nekvasil, June 11, 2009.
43 Peter Maskell, Harald Bathelt, and Anders Malmberg, “Building Global Knowledge Pipelines: The Role of Temporary Clusters,” European Planning Studies 14.8 (2006): 997–1013.
44 See Martha M. Lauzen’s annual reports: “The Celluloid Ceiling,” http://womenintvfilm.sdsu.edu/files/...ing_Report.pdf. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/07%3A_Transnational_Crews_and_Postsocialist_Precarity__Globalizing_Screen_Media_Labor_in_Prague/7.07%3A_Notes.txt |
On September 11, 2001, Afghanistan’s media sphere was one of the sparsest in the world. Few newspapers had survived the previous half-decade of Taliban rule, during which the nation devolved into a “country without news or pictures,” according to Reporters Without Borders.¹ A single radio station, Radio Sharia, was in operation—the lone remnant of a bygone Soviet era marked by relatively sophisticated, if centrally controlled, broadcasting practices. According to the architects of the American-led invasion that would ultimately overthrow the Taliban regime, this lack of media was not only a symptom of totalitarianism but also a cause.
• 8.1: Introduction
Overview of the chapter's aims: to understand the specific gender dynamics that have emerged in Western-funded Afghan media production, comparing the nonprofit and commercial sectors, and to identify the instability affecting workers in both sectors, especially women workers.
• 8.2: Rightful Suspicions
The Western emphasis on media assistance as a form of women's liberation in Afghanistan post-9/11; the "top-down" nature of that gender rights discourse that assumes Afghan women lack the agency to empower themselves and asserts a neoliberal, individualist ideal of women's empowerment.
• 8.3: Case by Case – Women in For-Profit TV
The youth-dominated Afghan for-profit mediasphere, and some of the factors that bring women to producerial positions in this sphere but keep them from retaining stable positions there.
• 8.4: The Noncommercial Sector
Examining the role of and the challenges faced by women in Afghanistan's noncommercial, Western-funded media sector, focusing on the case studies of Zakia Zaki and Farida Nekzad.
• 8.5: Conclusion
Considering the ways in which the American-funded media system in Afghanistan has created new opportunities for women, while at the same time creating additional precarity for women working in media.
• 8.6: Notes
08: The Cost of Business Gender Dynamics of Media Labor in Afghanistan
On September 11, 2001, Afghanistan’s media sphere was one of the sparsest in the world. Few newspapers had survived the previous half-decade of Taliban rule, during which the nation devolved into a “country without news or pictures,” according to Reporters Without Borders.1 A single radio station, Radio Sharia, was in operation—the lone remnant of a bygone Soviet era marked by relatively sophisticated, if centrally controlled, broadcasting practices. According to the architects of the American-led invasion that would ultimately overthrow the Taliban regime, this lack of media was not only a symptom of totalitarianism but also a cause. Free media, argued President George W. Bush during a speech to the National Endowment for Democracy in 2003, was a central pillar of “successful societies” that rejected terrorism, engaged productively with the international system, and respected the rights of all people, especially women.2 To “fix” Afghanistan would be, in part, to fix Afghan media.
Accordingly, America’s war on terror brought an unprecedented level of media intervention into Afghanistan. As early as November 2001, just weeks after the start of the invasion, American media consultants were in Kabul, laying the groundwork for a hybridized public-private broadcasting system that would flourish, at least numerically, in the years to come. Today, Afghanistan’s media market is as crowded as it was once barren. Countless television stations compete for viewer attention, ranging from the megalith Tolo TV, backed by the United States and co-owned by Newscorp, to the dozens of tiny outlets owned by local politicians, businessmen, and warlords. Bolstered by both direct American investment and a local economy supported by international aid, Afghan’s media system today appears not only vibrant but in many ways progressive. Minority cultural groups and women are represented on screen in a fashion that only years ago would have seemed impossible. It is also now commonplace to find women in key production positions, a phenomenon unseen in Afghanistan in the years between the end of Soviet control and the arrival of NATO forces in 2001.
This chapter details the ways Western media intervention in Afghanistan has aimed to foster cultural and economic environments that encourage female media participation, as well as the significant costs and limitations that have come with it. Drawing upon a five-week research trip to Kabul as well as extensive documentary research and personal communications over several years, I aim to move beyond the sweeping ideological accusations of critics to understand the specific gender dynamics that have emerged in the world of Western-funded Afghan media production. In doing so, I argue that Western efforts to bolster Afghan broadcasting have resulted in a limited but identifiable success with regard to greater female participation in the mediasphere. Noting the differences in gendered media labor between the nonprofit and commercial spheres, I foreground the intersection of economics and culture that complicates these two approaches to media assistance. I argue, however, that both approaches have fostered a sense of precariousness in the lives and careers of all media workers, with particular instability affecting women. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.01%3A_Introduction.txt |
If any element of recent Western foreign policy ought to be eyed with suspicion, it is the unique confluence of gender rights discourse and media assistance that has occurred in Afghanistan since 2001. Although understood largely as separate issues, both relate directly to the post–September 11 moment, in which the United States desperately attempted to articulate a worldview in which military and cultural intervention in Afghanistan would offer a sense of increased American security. As early as September 20, President George W. Bush offered the American people a picture of an Afghanistan plagued by the dual cultural blind spots of patriarchy and media deficiency. Detailing the need to destroy the Taliban and its support for Al Qaeda, Bush noted that in Afghanistan “women are not allowed to attend school. You can be jailed for owning a television.”3 Unable to offer much in the way of concrete objectives for America’s military intervention, Bush bolstered his moral standing by positioning the Taliban’s Afghanistan as a perfectly inverted version of America’s idealized neoliberal self-image.
Immediately, American politicians and media outlets began to construct a discourse in which foreign intervention was required both to avoid future terrorist attacks and to save Afghan women from Afghan men. In a striking example, CNN aired the BBC documentary Beneath the Veil multiple times in the months leading up to war, offering what Lynn Spigel describes as a chance for U.S. viewers “to make easy equivocations between the kind of the oppression the women of Afghanistan faced and the loss of innocent life on American soil on September 11.”4 Stabile and Kumar go farther in their analysis of American media coverage of Afghan women in the period following September 11, arguing that “women’s liberation” amounted to “little more than a cynical ploy” used to “sell the war to the US public.”5 American media, in their analysis, offered a decidedly selective vision of Afghan history in which the Taliban played the role of the dark villain and the United States was portrayed as the white knight rushing to save Oriental damsels in distress. Effaced from this account is the uneven, hard-fought struggle of Afghan women’s groups, such as RAWA, as well as the significant women’s rights violations committed by the United States and local allies like the Northern Alliance.6 Most provocatively, Stabile and Kumar go on to accuse the United States of using women as a tool through which to justify “imperialist domination,” rendering the West just as guilty of erasing Afghan women’s agency as the Taliban government against which it fought.
In a book-length study of the extensive aid aimed at improving the lives of women in postinvasion Afghanistan, Lina Abirafeh argues that willful blind spots produced by Western media had a direct impact on the sorts of programs that received funding and support in the country.7 In particular, stereotypical images of oppressed chaddari-clad Afghan woman seem to have dominated the mindset of NGO and Western government decision makers, much as they had captivated American readers of best-selling nonfiction, such as Zoya’s Story 8 and My Forbidden Face.9 Official American voices emphasized the importance of undoing the drastic restrictions on women’s liberty enforced by the Taliban beginning in 1990, with little attention paid to the diverse history of women’s experiences in Afghanistan. Abirafeh identifies an overtly “top-down” orientation to women’s rights programming, much of it embedded with a sense that Afghan women are “unable to empower themselves.”10 Echoing Islah Jad’s work on the “NGO-isation” of global women’s movements,11 Abirafeh notes that Western feminism, with its emphasis on individual, often economic rights, blinded the Western aid apparatus to the traditional strengths of the Afghan women’s movement.12 In this sense, the NGO world can be understood as being in line with the emphasis on free agency and mobility that marks the landscape of globalized labor markets and contributes to the sense of precariousness that plagues media workers in every subfield. Perhaps most damningly, however, Abirafeh declares that in the rush to provide them with a dramatic and politically popular salvation, the West “forgot to consult Afghan women at all.”13
Scholarly accounts of Western media assistance—defined here as the provision of Western funding and training to local media workers—to Afghanistan, though sparse, are hardly kinder than the critiques of gender aid. This pattern of suspicion follows a broader critical concern with media assistance, a field that remains steeped in the work of post-WWII modernization theorists, particularly Daniel Lerner.14 James Miller usefully summarizes this critique, noting that media assistance is “fundamentally about universalizing the local and assuming an unjustifiably near causal relationship between media . . . and self governance.”15 Turning a blind eye to the idiosyncrasies of Western media and Western democracy, media assistance advocates tend to assume that the two are inherently good and fundamentally intertwined entities.
A more subtle assumption built into media assistance work is an emphasis on the individual journalist or producer as the fundamental unit of a successful mediasphere. Although money is certainly devoted to institutional capacity building in international media projects, the trend toward contract, mobile labor found throughout the American media industry inevitably affects the training that aid recipients encounter. Furthermore, as Rao and Wasserman note, this preoccupation with individualism serves as a linchpin between the economic logic of the media business and hegemonic Western understandings of media ethics, which downplay communal interests in evaluating media quality.16 Miller suggests that such assumptions may be exacerbated by the well-meaning individuals on the frontlines of media assistance projects in places such as Afghanistan. The Western journalists and NGO workers who enact the on-the-ground aspects of media assistance often embrace the sort of precarious labor conditions with which this volume is concerned.17 Moving from place to place to provide training, these individuals bring with them the sense that media work is an independent endeavor often in direct tension with geographical and financial stability. Though this ought not impugn the intentions of Western NGO workers and media trainers, it is impossible to ignore the tension that exists between the radical individualism that might encourage someone to move from a European capital to war-torn Afghanistan and the more communal goals of local institutions.
This emphasis on market-oriented, entrepreneurial media systems is fully apparent in the reality of postinvasion Afghan media. Alongside the military onslaught of late 2001 that brought down the Taliban regime in Kabul came a concerted and highly coordinated effort to supply Afghans with a new, ostensibly independent media system. In addition to commandeering the state radio system, American forces, through USAID’s Office of Transition Initiatives (OTI), underwrote the production of a remarkably broad and diverse Afghan mediasphere. Within five years, a once broadcast-free rural Afghanistan was dotted with local radio stations surviving on a combination of foreign largesse and local advertising revenue. In addition to playing programming aimed at articulating the intentions of NATO forces and “rural transition teams,” these stations offered a mix of locally produced shows and foreign-funded public service material.
Kabul quickly emerged as a true media capital, as the vacuum produced by the Taliban’s near-total elimination of broadcasting gave way to a chaotic landscape in which outlets run by NGOs, warlords, and entrepreneurs competed for economic footholds and political influence. Security circumstances aligned with the “logic of accumulation” identified by Curtin in the development process of media capitals, bringing thousands of young aspiring professionals home to Kabul after years of exile in Iran, Pakistan, and the West.18 The first great success in this new environment was Arman FM, a purely commercial radio station that nonetheless received a large initial investment from USAID’s OTI. The relationship between USAID and Arman’s owners, the Australian-Afghan Mohseni brothers, would continue and grow, with the United States eventually providing over \$2 million in grants to Mohseni’s Tolo TV, a commercial station that now dominates the crowded field of Afghan television through a mix of programming tilted heavily toward Western-style game shows and dramas. Perhaps predictably, in popular accounts of Afghan’s new mediasphere, Tolo president Saad Mohseni is positioned as the protagonist of a story that emphasizes the individual entrepreneur over the realities of communal and government cooperation that make his station possible.
The unabashedly capitalist orientation of this project has drawn the ire of numerous critics, most notably Mark J. Barker, who argues that the newly oligarchic orientation of the Afghan mediasphere confirms American desires to foster a friendly “polyarchy” in the country, as opposed to a true democracy geared toward expressing the will of the people. To Barker, such a tactic emerges from the same strategy that led to the overtly deceitful content produced by the American-run Iraq Media Network in the wake of the fall of Baghdad. In each case, he argues, the United States took the steps it deemed necessary to ensure friendly leadership in occupied spaces, always at the expense of democracy and social justice.19 The Mohseni family, in this telling, represents an oligarchic regime that the United States supports due to its willingness to engage fully in the system of global capitalism. The local NGO elites favored by America play a similar rule, inculcating Western conventional thinking and contributing to the production of a mediasphere that embraces the values underpinning the neoliberal order. I will not evaluate this broader claim here. It does, however, offer a useful starting point from which to inquire into the relationship between media assistance and egalitarianism at the level of gender in Afghanistan. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.02%3A_Rightful_Suspicions.txt |
The post-Taliban period by no means represents the first time women held prominent positions in the Afghan mediasphere. The period of Soviet influence and control from the 1960s to 1989 brought a number of women into the field of journalism, as a select class of urban elites prospered while others across the country faced tremendous violence and persecution.20 Today, older Afghan media institutions, such as the state-run Kabul Times newspaper and Radio Afghanistan, employ a small but significant number of veteran female journalists, most of whom left as refugees during the Taliban years and returned after NATO and the Northern Alliance took control of Kabul. Like all Afghan media workers of this era, however, these Soviet-trained individuals are routinely dismissed as “unprofessional” by younger figures in the field. In addition to lacking training in contemporary media technology, they also, according to multiple sources interviewed for this study, are believed to lack the audience-focused approach to production required to succeed in contemporary Afghanistan. Whereas in other environments wartime experience might be a source of cultural capital, in contemporary Afghanistan the wholesale remaking of the local media system on primarily capitalist principles has largely marginalized the older professionals whose experiences are tainted by the communist era.21
The vast majority of media producers in Afghanistan, male or female, have thus emerged over the past decade, as the country moved from a single radio broadcaster to a loosely organized system in which hundreds of outlets compete for creative talent, spectrum space, and audience attention. As a result, the overwhelming majority of television producers in Afghanistan are comfortably under the age of thirty, with radio workers skewing only somewhat older. Top-rated television programs such as Tolo’s Afghan Star and On the Road, for example, are both lead-produced by men under twenty-five.
The rapid ascent of the Afghan mediasphere offers a unique set of obstacles to female participation. As was emphasized in American discourse surrounding Afghanistan in the preinvasion period, formal education for women was virtually annihilated in the country during the Taliban’s reign. Thus the desire to quickly craft a robust media system in the postwar period left little time to train and recruit young women who could balance the gender aspect of Afghan media labor. Instead, labor needs were filled largely by a combination of returning refugees from Iran and Pakistan and local men with basic educational backgrounds. As Barker points out, subsidies for new stations, both local and national, were granted overwhelmingly to politically connected, well-resourced individuals identified by Americans as entrepreneurial enough to thrive in a commercial environment.22 Such individuals, by local definition, had to be males able to curry favor either with urban political elites or rural community leaders with religious legitimacy. A combination of local resistance to female participation in the public sphere and foreign demands to quickly establish a commercially viable system left little opportunity for women in media during the earliest stages of Afghanistan’s reconstruction and established a system in which men currently possess a near monopoly on “experience” and “professionalism.”
However, the profit motive of Tolo TV, combined with the organization’s interest in establishing itself as capable of relating to Western supporters, has advanced the place of female producers in remarkable ways. In part, this results from the financial strength of the station, which draws upon the resources of its partner organization Newscorp to provide expensive services such as child care and door-to-door shuttle services, which are particularly important to women working in the dangerous environment of contemporary Kabul. These benefits are, for many potential female employees, absolute necessities that are often unavailable at the smaller-scale media operations that exist throughout the city.
Tolo has also made a concerted effort to hire women as producers, particularly in the areas of family and lifestyle programming, which are associated with primarily female audiences. Tania Farzana, for example, was recruited back to Kabul, after years in the United States, to produce a local adaptation of Sesame Street. Numerous other women, many of whom grew up in Afghanistan during the Taliban period, have risen to similar roles as producers within the organization. However, in speaking to a dozen female producers in Kabul in the spring of 2014, I was unable to locate one who considered an Afghan, not Western, woman to be her ultimate boss.23
In my attempt to identify the most experienced female producers in Afghan commercial television, I was consistently steered toward women between the ages of twenty and twenty-three. Rokhsar Azamee ranks as one of the most experienced female producers in Afghanistan, despite having left the industry at twenty-two. Feverishly working from the age of seventeen after being introduced to Tolo management by a neighbor, Azamee produced several programs, primarily in the health and morning talk show genres. Having freelanced at a number of local stations in Kabul, Azamee enthusiastically attests to the freedom allowed women at Tolo TV as well as Ariana TV, another for-profit station. She suggests that these outlets, especially Tolo, encourage female freedom of expression by never introducing the concerns of “the government” or religious leaders into programs on sensitive topics such as health and education. This is not to say, however, that working at Tolo comes without risk. As Wazmah Osman notes in her history of postinvasion Afghan culture wars over television, women who work at Tolo, particularly on air, face precariousness in the most literal sense. A famous, tragic example is that of Shaima Rezayee, an on-air personality murdered after months of criticism from conservative cultural elements.24
Perhaps with such factors on her mind, Azamee, at an age at which her Western counterparts would have been fighting over volunteer internships at local stations, reached what she felt to be a natural conclusion to her television career. She moved into the more lucrative and stable telecommunications industry.25 This remarkable trend toward youth currently cuts across gender lines at Tolo, although trends suggest that young men are more likely to remain with the organization for the long term. Although Kabul University offers a degree in journalism, Tolo TV recruits its creative staff by casting an enormously wide net, bringing in large numbers of young people with negligible skill sets and quickly assigning them surprising levels of responsibility. Most recruits wash out quickly, while the survivors take on relatively high-ranking producing roles within months.
This system succeeds in bringing in a fair number of women alongside a much higher proportion of men. However, Tolo’s trial-by-fire approach is far better suited to the lifestyles of young Afghan men. The hours are long, sometimes bordering on abusive.26 In a cultural space in which women working at night and women engaging in the public sphere are both points of great controversy, this system of long hours and high stakes at young ages is particularly precarious for women. Ultimately, it is untenable for most Afghan women to continue working such hours for the pedestrian pay that even the well-funded Tolo is able to offer.27 There are many men willing to endure these conditions during their twenties, gaining professional experience and prestige while putting off family life and the economic exigencies that come with it. However, this is less of an option for women, many of whom wish to marry during this time period.
As a result, the Afghan for-profit mediasphere is remarkably successful in bringing women to positions of responsibility in production but is far less successful in keeping them there. In my interviews with producers at Tolo TV, the station was often described as a benevolent institution insofar as it granted expressive freedom to young women and opened doors, including opportunities at Western media organizations like the BBC. It is not, however, a stabilizing force for women wishing to gain an economic foothold in Afghanistan’s uncertain economy. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.03%3A_Case_by_Case__Women_in_For-Profit_TV.txt |
As the United States supported Tolo TV in the hopes that market forces would, among other things, encourage greater female media participation, numerous smaller, nonprofit projects were put forth with this explicit goal. To a significant extent, the United States outsourced this aspect of the Afghan media sphere, building gaps into a system of small local radio stations to be filled by allied nations with projects geared specifically toward female empowerment. Although American policy papers written throughout the 2000s emphasize the importance of nonprofit, women-produced media, they also cede many logistic elements to the French and Canadian governments.28 This strategy makes a certain sense. America had overtly promised increased female access to the public sphere in the buildup to the war. However, once the invasion began, the realpolitik of post-Taliban Afghanistan provided ample incentive to create distance between the U.S. government and controversial media initiatives.
The most striking example of this danger is the story of Afghan Radio Peace, a single-room station broadcasting from Jabal Saraj in Parwan Province. Uniquely, the station’s origins predate September 11. In a meeting in France in March 2001, a group of French women’s rights leaders challenged Northern Alliance leader Ahmad Shah Masoud in exactly the way Abirafeh argues postinvasion NATO forces could not challenge Afghan leaders.29 As the world focused on the evaporation of women’s rights under the Taliban, French leaders pressed Masoud on his position on women in the Afghan public sphere. Putting the famed military leader on the spot, the group offered to fund a radio station in Northern Alliance territory, provided a woman be made manager. Masoud accepted and Zakia Zaki, a controversial local advocate for women’s rights, took charge of Afghan Radio Peace. The French agreed to pay for a year of Zaki’s salary, oil for an electric generator, and fifteen days of basic radio training with French broadcasting professionals.
By the time the station opened in October 2001, much had changed. Masoud was assassinated on September 10, one day before America’s invasion of Afghanistan was made inevitable. However, operating in a semiautonomous space at the mouth of the Panjshir Valley, Zaki’s small station was truly revolutionary. For six years, the station functioned on a combination of foreign money and local revenue strategies built primarily on classified-style hyperlocal advertising. As NATO and USAID took over rural radio broadcasting in Afghanistan, the French withdrew support, forcing the station to play American-funded public service programming to bring in the money necessary to remain on air. From 2005 to 2007, the station flourished under this model, airing foreign-produced material as well as a selection of local programs rarely focused specifically on women’s issues but steeped in the values that brought Zaki to anti-Taliban activism.
In 2007, however, tremendous changes took place. NGO support in Afghanistan began to falter from fatigue, and the Taliban began to reassert itself nationally, largely through increasingly daring suicide attacks. With these disruptions came a campaign specifically targeting female journalists and other public figures. On June 6, 2007, Zaki was murdered. Her death had a massive effect on both Afghan Radio Peace and female media participation throughout the country. In the years following Zaki’s assassination, little support has flowed to the station, with no foreign institution wishing to rebuild it to its previous place. Zaki’s husband, Abdul Ahad Ranjbar, has taken over what has become a quiet outlet broadcasting twelve hours a day, half of which is American-funded national programming for which the station receives a few hundred dollars to keep the gas generator running.30
If there is a figure who has picked up Zaki’s mantle, it is Farida Nekzad, whose career has run the gamut of Western-supported noncommercial media in Afghanistan. Nekzad began her media career as a refugee, working in Pakistan when the Taliban controlled Kabul. Inspired by a neighbor who worked for state media during the years of Soviet control, Nekzad received an education in journalism at Kabul University, before the Taliban banned female enrollment. In Pakistan, Nekzad found work with the BBC. When she returned to Kabul in 2002, Nekzad was a rare woman with the credentials necessary for managerial-level work with Western-funded media institutions. This experience allowed her to find employment with the Institute for Media, Policy and Society (IMPACS), a branch of the Canadian government encouraged by USAID to create three rural women-run stations in the relatively peaceful northern region of Afghanistan.
According to Sarah Kamal, a scholar who worked at the IMPACS station in Mazar I Sharif, the project fell into many of the unreflective patterns so often seen in Western approaches to women’s development. In addition to early difficulties in recruiting women to work at the station due to local cultural resistance, the outlet was plagued by a disconnect between the needs of listeners and the expectations of international organizations. Foreign funders and local religious leaders required all programming be preapproved and “pressed the radio station towards adopting a scripted and more formal radio voice over spontaneous conversational dialogue in its programming.”31 Ultimately, the station failed to reach its intended female audience, as Western ideas of individualism and journalistic professionalism created a growing gap between the voices on the air and those listening.32
However, this professionalized, Westernized understanding of how to run a radio station has had one significant side effect: it has produced numerous female journalists prepared to succeed in Westernized organizations, often at better pay. The station’s current director, Mobina Khairandish, describes the outlet’s function as much in terms of training women producers as reaching the sort of audience envisioned by IMPACS when it made the original investment in 2005. Now financially stable, though reliant on occasional contracted projects for NGOs, the station has become a training ground for local women wishing to gain a foothold in the world of media. Despite ongoing difficulties with local groups who question the appropriateness of women on the radio, recruitment issues have more or less disappeared, with small but consistent numbers of young women arriving at the station to work each year.33
Although critics like Kamal question the station’s ability to truly engage with large numbers of local listeners, it is undeniable that journalists trained at the station have moved on to jobs both within Afghanistan and abroad. In a creative maneuver, the outlet has taken young women whose families discouraged them from taking on public roles and positioned them as journalistic trainers. For example, local offices of Nai Supporting Open Media in Afghanistan, an Internews-funded institution for media education, now feature alumna from the Mazar I Sharif station. Other former employees work across the globe, primarily in media assistance organizations whose goals model the idealized, arguably disconnected, approach to journalism put forth by IMPACS. Although IMPACS perhaps failed in its attempt to train producers capable of connecting to a local audience, it succeeded in training Afghan women for a world of precarious labor in the overlapping fields of media production and media assistance. A project originally intended to suture community bonds may ultimately serve as a training ground for individuals entering an era of transitory and inconsistent labor conditions.
Farida Nekzad became the IMPACS project’s greatest local success, leveraging her time with the Canadian organization to procure a position as the codirector of Pahjwok News. Set up as an independent NGO, Pahjwok—funded by Internews (and thus USAID) and based in Kabul—serves as the main domestic wire service in Afghanistan. While at Pahjwok, Nekzad oversaw tremendous changes in Kabul’s journalistic landscape, particularly with regard to women. As universities began producing the first generation of post-Taliban journalism graduates, a number of women sought work at organizations like Pahjwok.
Nekzad, in a unique position of power, made a number of policy changes that have had a significant impact. Most obviously, she instituted a hiring quota for women, arguing that the increase in female journalism graduates required a change on the part of the organization. Remarkably, for a brief period in 2006, Pahjwok employed more female than male journalists. More subtly, Nekzad fundamentally changed local newsroom culture, temporarily suspending the traditional practice of separating men and women at company lunches. In Nekzad’s view, lunchtime gender segregation, which remains prevalent throughout Afghan business, government, and NGO culture, represented a significant stumbling block to gender equality. “They make big decisions over lunch,” she notes. “People think that men are funny and women are quiet because they are in the other room.”34
The ugly events of 2007, however, forever changed Nekzad’s career and undid a number of the changes she had instituted. After Zaki’s death, Nekzad began to receive increasingly violent threats, followed by multiple attempts on her life. Women no longer applied for positions at Pahjwok at the same rate, for fear of reprisals. When Nekzad became pregnant, she quit the organization for America to safely raise her child. Upon returning in 2010, she found a very different landscape. Pahjwok was once again dominated by men, with lunches resegregated and only men in managerial positions.
In response, she took over a fledging competitor, Wakht News. Housed in a three-room apartment off of a main street in Kabul, Wakht currently operates a frequently updated web site that is routinely cited by mainstream Afghan media. Nekzad has attempted to restaff the organization with women but has found the task nearly impossible. As Western NGOs have fled Kabul over the past five years, little grant money is available for anything but well-established, typically male-dominated institutions. As a result, Nekzad was forced to pare back her budget. Although women can sometimes be hired for lower salaries, their presence adds to a company’s expenses. While traveling in the city is dangerous to anyone, men are able to ride public transportation and hire taxis when they are unable to walk. For women, this is too dangerous, adding significant private transportation costs to employer budgets. Nekzad’s new organization thus faces a familiar conundrum in public service Afghan media. Those organizations that are big enough to maintain a large, diverse staff are deeply entrenched in traditional modes of office culture that work to the disadvantage of female workers. Those that might be willing to challenge these norms, however, are unlikely to receive enough funding to successfully hire women, given the significant additional expense. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.04%3A_The_Noncommercial_Sector.txt |
Reflecting on the popularity of the Afghan women’s cause in the period following September 11, Lila Abu-Lughod notes the deeply troubling parallels between the American rhetoric of the time and the colonial era discourse of “saving Muslim women.”35 The American government, media, and seemingly much of the population understood the plight of women under Taliban rule in terms that were both culturally reductive and historically myopic. Generalizations about Muslim and Afghan women abounded, while reflections on the role of the West in shaping Afghan history were all but absent.36 The American-led war effort was praised for freeing Afghan women from the Taliban but rarely critiqued for assuming that all people ought to pursue Western, neoliberal visions of individual agency and free expression.
It is both easy and entirely appropriate to set American media assistance efforts in Afghanistan within this context. Although efforts were made to incorporate local voices in the new Afghan media system, this system was nonetheless built upon presumptions of the superiority and universality of Westernized media systems. Whenever possible, wealthy entrepreneurs were afforded benefits. When for-profit media was not suitable, America and its allies employed an NGO model that David Harvey found to be deeply intertwined with the neoliberal state system, often emphasizing individual rights over community needs.37 Both of these approaches appealed to the notion of “saving” women critiqued by Abu-Lughod, in the former case through the magic of the profit motive, in the latter via the largesse of Western cosmopolitanism.
In this chapter I have strived to move beyond the simple neoliberal critique, attempting to consider more closely the specific, concrete impact of American policies on the work of female Afghan media workers. It would be foolish and dishonest to deny that the American-imposed system of media that currently dominates Afghanistan has brought hundreds of women into the public sphere in ways previously impossible. In the nonprofit realm, rare, privileged, and remarkably determined individuals like Farida Nekzad have succeeded in using small openings imposed by the West to create new opportunities for female voices. Furthermore, in considering the words and experiences of women working in the field, it is apparent that, given the circumstances, the profit-oriented media systems decried by Barker do, in fact, offer a greater range of expression to women. Although the Afghan government attempts to exert control over all media, the economic might and global cachet of Tolo TV have allowed the station to push boundaries, thus providing greater autonomy for producers like Rokhsar Azamee.
And yet it is necessary to note that, despite the rhetoric of security and nation building surrounding American media efforts in Afghanistan, increased female expression has by no means removed the precariousness of Afghan labor. In some cases, it has actively encouraged new elements of uncertainty, particularly for women. Most obviously, violence, death threats, and terror still plague the lives of female producers, although this situation predates the immediate post-2001 American involvement in the region.38 More subtly, both NGO media initiatives and for-profit businesses place women in disproportionately precarious circumstances. Although NGOs train and hire women at admirably high rates, their funding is fickle, with donors often falling away over time. As seen in the case of Nekzad’s Wakht News, women are often the first to lose their jobs.
In the realm of commercial outlets, such as Tolo TV, female producers are valued as a short-term means to attract women viewers and positive global press. Perhaps, over time, economics will encourage the outlet to offer long-term stability to the most successful female producers. However, given Afghan economy’s remarkable instability and foreign dependence, this seems unlikely. Women will likely continue to leave for more lucrative, stable, and culturally acceptable positions, leaving most of the prestigious yet highly taxing production jobs to the men. Yes, Afghan women now have access to jobs that did not exist fifteen years ago and would never have been open to females even if they had. However, these new opportunities have combined the precariousness of war and reconstruction with the sorts of precariousness described throughout this volume in even the calmest mediaspheres. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.05%3A_Conclusion.txt |
1 Vincent Brossel, Jean-François Julliard, and Reza Moini, Afghanistan: What Gains for Press Freedom from Hamid Karzai’s Seven Years as President? (Paris: Reporters Without Borders, 2009), 3.
2 George W. Bush, “Remarks by President George W. Bush at the 20th Anniversary of the National Endowment for Democracy,” National Endowment for Democracy, www.ned.org/remarks-by-president-george-w-bush-at-the-20th-anniversary/.
3 George W. Bush, “Transcript of President Bush’s Address,” CNN, September 21, 2001, http://edition.cnn.com/2001/US/09/20...sh.transcript/.
4 Lynn Spigel, “Entertainment Wars: Television Culture after 9/11,” American Quarterly 56.2 (2004): 249.
5 Carol A. Stabile and Deepa Kumar, “Unveiling Imperialism: Media, Gender and the War on Afghanistan,” Media, Culture & Society 27.5 (2005): 766.
6 Ibid., 773.
7 Lina Abirafeh, Gender and International Aid in Afghanistan: The Politics and Effects of Intervention (Jefferson, NC: McFarland, 2009).
8 Zoya, John Follain, Rita Cristofari, and Rita Wolf, Zoya’s Story: An Afghan Woman’s Struggle for Freedom (New York: HarperCollins, 2002).
9 Latifa, My Forbidden Face (London: Virago, 2002).
10 Abirafeh, Gender and International Aid in Afghanistan, 28.
11 Islah Jad, “The NGO‐isation of Arab Women’s Movements,” IDS Bulletin 35.4 (2004): 34–42.
12 Abirafeh, Gender and International Aid in Afghanistan, 30.
13 Ibid., 16.
14 Daniel Lerner, The Passing of Traditional Society: Modernizing the Middle East (New York: Free Press, 1958).
15 James Miller. “NGOs and ‘Modernization’ and ‘Democratization’ of Media: Situating Media Assistance.” Global Media and Communication 5.1 (2009): 16.
16 Herman Wasserman and Shakuntala Rao, “The Glocalization of Journalism Ethics.” Journalism 9.2 (2008): 163–181.
17 Miller, “NGOs and Modernization,” 16.
18 Michael Curtin, Playing to the World’s Biggest Audience: The Globalization of Chinese Film and TV (Berkeley: University of California Press, 2007).
19 Michael J. Barker. “Democracy or Polyarchy? US-Funded Media Developments in Afghanistan and Iraq post 9/11,” Media, Culture, and Society 30.1 (2008): 119.
20 Abirafeh, Gender and International Aid in Afghanistan, 19.
21 Habib Amiri, personal communication, June 2013.
22 Barker, “Democracy or Polyarchy?” 109–130.
23 This is not to say that such a person might not exist. However, the trend of young women serving as producers under slightly less young males was remarkable in its consistency during my Skype and Facebook interviews.
24 Wazhmah Osman, “Thinking outside the Box: Television and the Afghan Culture Wars” (PhD diss., New York University, 2012), 132.
25 Rokhsar Azamee, personal communication, June 2013.
26 I realize this is a serious accusation, but I firmly believe it to be the case. In my discussions, multiple young male producers, whom I will not name in this context, attested to working shifts lasting over twenty-four hours. To them, these tasks seemed more or less akin to that of a college student pulling an all-nighter—stories recounted with nostalgia but nonetheless indicative of labor exploitation.
27 A lead producer of multiple high-rating Tolo shows might make as little as \$1,200/month in a city that suffers from considerable housing and food inflation.
28 Colin Soloway and Abubaker Saddigue, USAID’s Assistance to the Media Sector in Afghanistan (Washington: USAID, 2005).
29 Abirafeh, Gender and International Aid in Afghanistan.
30 Abdul Ahad Ranjbar, personal communication, June 2013.
31 Ibid., 408.
32 Ibid., 409.
33 Mobina Khairandish, personal communication, June 2013.
34 Farida Nekzad, personal communication, June 2013.
35 Lila Abu-Lughod, “Do Muslim Women Really Need Saving? Anthropological Reflections on Cultural Relativism and Its Others,” American Anthropologist 104.3 (2002): 785.
36 Ibid., 784.
37 David Harvey, A Brief History of Neoliberalism (London: Oxford University Press, 2005), 78.
38 Which is not to say, however, that America did not play a role in bringing Afghanistan to the state it is currently in. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/08%3A_The_Cost_of_Business__Gender_Dynamics_of_Media_Labor_in_Afghanistan/8.06%3A_Notes.txt |
“Immy, can you tone it down a bit?” I was watching a film shoot at the Marriott Renaissance in Mumbai in July 2014, and the director, Vikram Bhatt, was instructing the lead actor, Emraan Hashmi, about his body language during a tracking shot where he had to stride resolutely across the hotel’s ballroom. When Bhatt’s assistant director began to block the shot for Hashmi, Bhatt bellowed in Hindi from his position in the back of the room, “Arre beta, thode dheere se jaao! [Hey son, go a little slower!].”
• 9.1: Introduction
Overview of the chapter's goals: to examine the structural changes in the Hindi film industry since the mid-2000s that have led to film dialogue increasing in regional specificity of Hindi while privileging English over Hindi on the production side.
• 9.2: English as a Lingua Franca in the Hindi Film Industry
The historically multilingual nature of the Hindi film industry, and how this led to the adoption of English as an industry lingua franca starting in the early sound era. Highlighting the difference in the modern elevation of English over Hindi as the language of creativity and decision-making, and its historical status as a strictly necessary means of communication.
• 9.3: The Precarious Status of Hindi in the Hindi Film Industry
Some of the current factors leading to the decline of Hindi in favor of English within the Hindi film industry, including the rise of multi-generation film families whose children are educated in English-only contexts, the increased adoption of script-writing software that only supports Roman script, and actors who are unwilling to improve their Hindi skills and must be accommodated by writers under the star-centric film system.
• 9.4: Hindi “Indie”
Exploring contributing factors to the trend of post-2006 films using highly localized registers of Hindi, in contrast to the generic Hindustani of earlier films. Factors include outsider filmmakers gaining traction in the formerly metropolitan-dominated film industry and seeking to increase their perceived authenticity in depicting Indian life, and the rise of corporate production and distribution decreasing the need for any single film to be a "universal hit."
• 9.5: Conclusion
Implications of the ascendance of English over Hindi in the Hindi film industry, including the relegation of workers only fluent in Hindi to below-the-line status and the extent of English's naturalization as the unofficial language of film production such that films can gain unwarranted praise for authenticity due only to their use of localized Hindi registers.
• 9.6: Notes
09: No One Thinks in Hindi Here Language Hierarchies in Bollywood
“Immy, can you tone it down a bit?” I was watching a film shoot at the Marriott Renaissance in Mumbai in July 2014, and the director, Vikram Bhatt, was instructing the lead actor, Emraan Hashmi, about his body language during a tracking shot where he had to stride resolutely across the hotel’s ballroom. When Bhatt’s assistant director began to block the shot for Hashmi, Bhatt bellowed in Hindi from his position in the back of the room, “Arre beta, thode dheere se jaao! [Hey son, go a little slower!].” He then spoke on the phone in Gujarati with a marketing representative from the music company that had released the soundtrack of his soon-to-be-released film. I noticed that the sheets of Hindi dialogue used by another assistant director to monitor actors’ accuracy were written in Roman rather than Devanagari script.
A Hindi film set is a highly multilingual environment, and it is common to hear several languages spoken, but the linguistic bifurcation illustrated in this example, where Bhatt spoke in Hindi with his assistants and in English with his lead actor and screenwriter, is a manifestation of the increasing presence of English in the everyday life of the Hindi film industry. In 1996, when I started fieldwork on the production culture of the film industry, I was surprised by how prevalent English was as a lingua franca, especially among the actors, directors, writers, art directors, designers, and others responsible for the creative labor that goes into a film. For below-the-line workers, Hindi was merely one language in a complex linguistic universe that included Marathi, Bengali, Tamil, Telugu, Gujarati, and Punjabi. This is a testament to the tremendous linguistic diversity of India—18 official languages but 122 languages with at least 10,000 native speakers—and the cosmopolitan nature of the Hindi film industry, where people hail from every linguistic region of India as well as other parts of South Asia (or beyond) and are not necessarily native Hindi speakers. According to the 2001 census, while Hindi is spoken by 53.6 percent of the population, there are fifty different types of Hindi.1
India is perhaps unique among film-producing nations for having at least eight major film industries, all distinguished by language, and for producing films in about twenty languages every year. The polyglot nature of the contemporary Hindi film industry fits into the broader history of filmmaking in Mumbai. Mumbai, as a colonial center of commerce, has always been marked by tremendous linguistic, ethnic, and religious diversity, and this diversity has been apparent in the world of filmmaking from its origins: early Indian cinema featured Parsi and Gujarati capital, Marathi directors, and Anglo-Indian performers.
With the advent of sound in 1931, Mumbai filmmakers had to choose which language to make films in; Hindi offered the largest market, but which type of Hindi? Filmmakers finally settled on a version, referred to as Hindustani by the British, that had operated as a sort of lingua franca throughout northern India.2 Thus Mumbai became the only city in India where the language of the film industry’s output was not congruent with the dominant languages of the region, Gujarati and Marathi. This was in direct contrast to other major centers of film production in India, such as Kolkata, Hyderabad, Chennai, and Trivandrum. Thus the Hindi film industry, unlike other Indian language film industries, has not had recourse to a regional state apparatus to promote its interests. Other states in India promote filmmaking in their official languages by offering incentives and subsidies, whereas Hindi films are not identified with any one particular state.
Whether it is the earmarking of subsidies for filmmaking in specific languages, the promotion of a particular dialect as a normative standard in advertising, the daily translations undertaken by news agencies, or Hollywood studios’ local language production strategies, language—as a category of socio-political identity, a form of labor, a set of commodified skills, and an object of market exchange— plays a critical role in the political economy of media industries.3 Referring to the increasing opportunities and attractions afforded by the Hindi film industry and the growing international profile of Bollywood, this chapter discusses how changes in language or code choice within Hindi cinema and the increasing significance of English in the production culture of the film industry concretely animate the transformations that have taken place in the political economy and social world of the Hindi film industry since the advent of neoliberal reforms in India mandated by the International Monetary Fund in 1991.
The changes I have characterized elsewhere as gentrification have resulted in a situation in which two apparently contradictory phenomena are taking place within the contemporary industry:4 the spoken language in many contemporary Hindi films is much more diverse and regionally specific than in films from earlier decades, at the same time that fluency in Hindi appears to be waning among certain elite categories of creative workers (writers, directors, actors, producers), resulting in a situation where English has attained a certain primacy and status and putting those whose primary language is Hindi in a far more socially and economically precarious position within the industry. This chapter discusses the reasons for and consequences of this paradox and illustrates how language and linguistic competence become sites for the elaboration of distinction, the performance of cultural capital, and the enactment of new hierarchies within the Hindi film industry. I argue that the turn toward localized registers of Hindi in film dialogue is integrally connected to the increased prevalence of English within the film industry, as both phenomena emerge from structural transformations that have beset the industry since the mid-2000s.
These transformations have reduced the economic precarity that typified Hindi filmmaking for much of the industry’s history. Flexibility, fragmentation, decentralization, and their associated occupational/employment insecurities, which are cited as characteristics of a global late-capitalist order, have actually been defining features of the Hindi film industry since the end of World War II. Dramatic changes in the structure of the Hindi film industry were initiated after the Indian state recognized filmmaking as a legitimate industrial activity in 2000. Official designation as an industry paved the way for a greater variety of financing for filmmaking, including loans from banks and other financial institutions, and initiated a number of structural changes commonly characterized as “corporatization,” where high-profile Indian conglomerates established new production-distribution companies or existing production, distribution, or exhibition concerns became public limited companies listed and traded on the Indian stock market. These new regimes of finance and organization in the film industry transformed it from a very undercapitalized enterprise (with accompanying high rates of attrition and stalled films) to one where raising capital was no longer an obstacle. However, these very conditions have produced a scenario where Hindi has become marginalized within the Hindi film industry.
This chapter is divided into three main sections, based on fieldwork conducted with screenwriters, writer-directors, directors, and journalists in Mumbai in August 2013, January 2014, and August 2014. First, I provide historical background on the multilingual nature of the Hindi film industry, including the long-standing presence of English. Then, I discuss how contemporary members of the film industry assess the relationship between English and Hindi within the industry and outline the impact, especially on screenwriting labor, of the growing reliance on English within the creative process. Finally, I examine how certain filmmakers deploy their linguistic skill in Hindi as a form of cultural capital and a mode of elaborating distinction within the film industry. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/09%3A_No_One_Thinks_in_Hindi_Here__Language_Hierarchies_in_Bollywood/9.01%3A_Introduction.txt |
If I’m making a Hindi film, I’m a Tamilian, but my DOP is a Malayali, and my editor is from Gujarat, so our common language is English.
—Sriram Raghavan, writer-director
When reflecting upon the linguistic history of the Hindi film industry, two features stand out: the relative insignificance of fluency in Hindi/Urdu as a prerequisite for acting, directing, or even writing; and the consistent presence of English as a language of trade discourse, commentary, and professional nomenclature. As mentioned previously, the industry emerged in multilingual Mumbai rather than the regions of northern or central India referred to as the Hindi “heartland” and drew personnel from all over the subcontinent and beyond. While standard histories of Indian cinema point out the diverse ethnic and linguistic backgrounds of actors and directors during the silent era, making special mention of how actors and actresses who could not speak Hindustani were displaced by stage actors and courtesans with the arrival of sound, the scenario is actually more complex.5
Even with the advent of sound in 1931, directors and actors came from diverse linguistic and national backgrounds. For example, Bombay Talkies, which left an important legacy in the postindependence Hindi film industry in terms of stars and directors, had many Germans in its employ. One of its directors, Franz Osten, who did not know any Hindi, directed some iconic Hindi films from this era. Throughout its history and continuing till the present, there have always been a few directors working in the industry who knew very little or no Hindi at all. This holds true for actresses as well. One of the top stars of the 1930s was the Australian-born Mary Evans, renamed “Fearless Nadia,” who gained fame in action/stunt films despite her heavily accented Hindi. Presently, women from non-Hindi-speaking parts of India as well as from as far afield as Australia, Brazil, Canada, Great Britain, Sweden, and the United States continue to try their luck in Bollywood.6
Even in the scripting process, English has played an important role starting from the early sound era. During the 1930s, in studios like Bombay Talkies, scripts and dialogues were initially conceived of in English by the writer (who was referred to as a “scenario” writer), after which the dialogue writer translated them into Hindustani.7 Since scripts were often written by individuals who were not proficient in Hindustani, the autonomous dialogue writer emerged as a staple of Hindi cinema. The credits for a script were broken down into three components: story, screenplay, and dialogue, with each element attributed to a different individual—a practice continuing into the present. As a result of the varying ethno-linguistic backgrounds of Bombay film personnel, writers who were fluent or had a facility in Urdu were in great demand as dialogue and lyric writers, since Persianized Urdu was a valorized register for song lyrics and dialogues.8 Many well-known Hindi/Urdu poets, playwrights, and novelists supported their literary endeavors by working in the Hindi film industry, and scholars have pointed out that after the partition of British India into India and Pakistan, whereby Urdu became the official language of Pakistan, the only site in India where Urdu was kept alive, and even flourished, was the Hindi film industry.9
In this multilingual context—Urdu writers, German directors, Bengali actors, Marathi singers, Parsi producers, and so on—it is not surprising that English emerged as a lingua franca for cultural producers based in a British colonial port city. While Hindi was (and remains) important as the language within the diegesis and that of consumption, English served (and continues) as the primary language of professional nomenclature and discourse about Hindi cinema and filmmaking. The English terms director, producer, writer, actor, and film are part of daily parlance within and outside the industry, rather than the Hindi equivalents nirdeshak, nirmata, lekhak, abhineta, and chalchitra. In contrast to films made in other Indian languages, the opening and closing credits for mainstream Hindi films have been in English since the 1930s. The Devanagari (Hindi) and Nasta’liq (Urdu) scripts make an appearance in a film’s title, but only after the prominent appearance of the title first in Roman script. One reason is perhaps because Hindi films are the only ones to have been consistently distributed nationally since the 1930s and internationally since the 1940s.
The most prominent forms of journalistic, critical, and trade commentary about the Hindi film industry have been in English since the 1930s. Film reviews, interviews with stars, industry news, celebrity gossip, and trade reports are carried out in English-language periodicals, whether Filmland, Filmindia, or Blitz in the 1930s–1960s; Filmfare, Trade Guide, or Screen since the 1950s; or Film Information or Stardust since the 1970s. While there are several Hindi fanzines and newspapers that cover Hindi cinema, they are quite marginal in terms of their impact or readership within the industry. The main trade journals, Trade Guide, Film Information, and Box Office India, which carry box-office figures and report about the business of the film industry, are in English.
If English has always played a prominent role in the Hindi film industry, how is the contemporary moment distinctive? The most drastic difference is English’s changed status and value relative to Hindi. While English has served as a necessary lingua franca throughout the industry’s history, it is increasingly operating as a language of production, creativity, and decision making since the mid-2000s. This change has to do with key demographic shifts in the film industry: namely, the intensification of kinship networks whereby a significant number of leading actors, directors, and producers represent the second, third, or even fourth generation within the industry; and a larger number of creative personnel drawn from urban social elites whose formal schooling has been wholly in English. Since the turn of the millennium, as Hindi filmmaking became more lucrative and rationalized, taking on an aura of professionalism and respectability that it had not traditionally enjoyed, social elites and film industry progeny gravitated toward the film industry as a viable career path.10 In the next section I discuss how these shifts are implicated in industry members’ assessments of the state of Hindi within the film industry. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/09%3A_No_One_Thinks_in_Hindi_Here__Language_Hierarchies_in_Bollywood/9.02%3A_English_as_a_Lingua_Franca_in_the_Hindi_Film_Industry.txt |
During my fieldwork in Mumbai in 2013 and 2014, several observers and members of the Hindi film industry lamented that the knowledge of Hindi had become so abysmal that the language appeared to be in a precarious position within the industry. Anupama Chopra, a noted film critic and television host, stated bluntly during our conversation, “Hindi is a secondary language now.” She relayed the travails of producing a Hindi version of her popular English-language weekly television show, Front Row with Anupama Chopra—a talk show that mixes film reviews, interviews with actors and directors about their upcoming releases, and group discussions about important issues or key trends within the industry. The challenge of producing the Hindi version, according to her, was that “no one thinks in Hindi here,” especially the younger generation of actors, who, though quite voluble and articulate in English, were unable to express themselves in Hindi. Chopra recounted how since it was so difficult for many actors to speak entirely in Hindi, they frequently devolved into English. She noted how in an episode on the relationship between Bollywood and fashion, eight minutes had to be cut from a thirty-minute segment because of the inability of the guests to converse about the topic in Hindi. Chopra quipped, “We would all breathe a huge sigh of relief after we finished the Hindi version, and then sit back and think [referring to the English version], ‘Now we can relax and have fun!’”
Social class, generation, and geography are the central reasons offered by industry observers for waning fluency in Hindi. The two main social groups identified as having a poor knowledge of Hindi are actors and directors who grew up within the film industry, nicknamed “star kids,” and upper-middle-class residents of Mumbai, dubbed “South Bombay types.”11 What these groups have in common is limited formal education in Hindi as a result of going to elite English-medium schools in India or boarding schools abroad, as well as the absence of a Hindi-speaking milieu by virtue of growing up in an elite social world in Mumbai where the primary language is English. Ajay Brahmatmaj, the film editor for the Hindi-language Dainik Jagran, the most widely circulated newspaper in India, discussed how in the current generation of actors, those who are from Mumbai and especially from film families speak Hindi only when they are compelled to with their domestic labor and household staff, and hence their knowledge of Hindi is limited to a very simple register. He said (in Hindi), “Many of them say they practice their Hindi, but with whom? With their cook, driver, and vegetable vendor. Now, the conversations with such individuals will be limited in terms of the vocabulary, not more than one hundred to two hundred words. At the most it will be ‘gaadi lao’ [bring the car], now ‘gaadi lao’ is hardly Hindi!”12 Screenwriter Kalpana Chadda, a native of Delhi, who started working in the film industry in the early 2000s and who had learned and spoke English only in school, described how colleagues and friends regard her as an anomaly for being comfortable in Hindi, asking her frequently, “Why do you speak in Hindi so much?”13 She reflected, “Delhi is very Hindi, friends speak to each other in Hindi, but in Bombay it seems not to be appropriate to speak in Hindi and to date that’s the joke about me.”14
Chadda spoke at length about the challenges faced by screenwriters like herself who “think in Hindi” in an industry run by people who primarily “think in English.” One particularly ironic manifestation is when she is hired to write a screenplay but not dialogue. Since a screenplay has to have dialogue, the screenwriter will put in “dummy” or placeholder dialogue, after which the dialogue writer takes over and crafts the speech in the film. Although she is instructed to write the screenplay in English, Chadda ends up writing her dummy dialogue in Hindi because of her facility with the language, but then has to translate them into English for the director, producer, and actors, even though the film will ultimately be in Hindi. Chadda said she felt like telling filmmakers, “Why don’t you just keep this dialogue and throw it away and let the writer write something else because it is double work for me to make the dialogue into English.”15 She also mentioned that she was much less precise in English, but according to her, most directors and producers from Mumbai are unable to comprehend an entire screenplay in Hindi. She asserted, “They won’t be able to listen to a script written completely in Hindi. They won’t get it. When it is in English, they’ll get the craft and say, ‘Oh this scene is tight’ because English lends itself to crispness. Hindi is very difficult for you to go crisp on it. And we can’t use difficult words because everybody is not familiar. If I use good Hindi words, I’ll write a crisp Hindi script, but I can’t do that—I have to use colloquial and general words.”16
Notice that Chadda mentions “listening” to rather than “reading” a script. The dominant convention in the film industry is to orally recount a script, and it is commonplace to hear actors assert in interviews that they decided to work in a particular film after “hearing the script.” Key members of the production team gather to hear the writer or director relay the film’s screenplay. These sessions, referred to as “narrations” in the industry, are undertaken for the purpose of pitching or having a project green-lighted as well as recruiting the cast and crew. Since a script is often judged on how well it is narrated, Chadda explained that the practice of narrating a script disadvantaged writers who had limited proficiency in English.
While Chadda related the difficulties writers face with producers and directors, others spoke of the challenges of working with actors who had limited Hindi skills. Kamlesh Pandey, president of the Film Writers Association, who has written the dialogue or screenplay for a number of prominent films starting in the late 1980s, was vociferous in his criticism of the state of writing and Hindi in particular. Pandey blamed urban, English-educated writers and industry insiders for the poor state of Hindi, and criticized the prevalent practice of having to write Hindi dialogues in Roman rather than Devanagari script because of the inability of many younger actors to read Hindi. Pandey complained, “Hindi has come to such a state that it has to be read in Roman, and hence I’m afraid the lipi, the script will soon become extinct. In cinema, Devanagari lipi [script] has more or less disappeared.”17 For those writers who specialize in writing dialogue in Hindi, either from the outset or adapting someone else’s English dialogue, an actor’s facility with the language has significant consequences for the writer’s creative labor. For an individual who is fluent in Hindi, which is a phonetically based language and alphabet, having to write dialogues in Roman script involves more effort, especially since the screenplay of a Hindi film on average comprises about seventy-five to eighty scenes and tends to be dialogue-centric.18
Another impetus to transliterate Hindi into Roman script is connected to broader efforts to refashion the film industry into a professional, corporatized site with greater emphasis on planning, preproduction, and rationalization of the production process.19 An important artifact of such planning is the “bound script,” which has achieved a near-totemic status within the film industry. The desire for a complete typed script with dialogue available in advance, supported by a younger generation of computer-literate screenwriters and assistant directors who have had some formal film training, has led to an increase in the use of screenwriting software such as Final Draft, which is an English-only application. Hence, even if actors can read Hindi, screenwriters who utilize such software have to write their dialogues in Roman script, and then may have to transcribe the dialogue separately into Devanagari for veteran actors who find it alienating to have to read Romanized Hindi. Chadda, who uses Final Draft, remarked, “It’s so strange that we have a multibillion-dollar Hindi film industry, but we are slaves to English. We even write the Hindi word in English.”20
Writers also related that they had to think harder about vocabulary and syntax when actors were not fluent in Hindi. Pandey complained that he was unable to be subtle in his dialogue writing since actors did not understand nuance or idioms specific to Hindi. Writer-director Sriram Raghavan mentioned that he had to keep in mind an actor’s facility with Hindi when composing dialogues because good lines could ring false depending on the actor’s ability to deliver them. Sameer Sharma, a writer-director who has written dialogues for films helmed by directors who knew little to no Hindi, related his frustration: “I think the sad part is that most actors today have a diction problem, so they don’t really try, and there are directors who don’t correct them because they themselves have a problem. That’s very visible, and it’s very irritating for somebody who knows the language, but they get away with it so they don’t work hard.”21
Writers thus feel they have to work harder to make it easier for actors to read and speak Hindi, rather than actors expending the effort to improve their language skills. This appears as another manifestation of the star-centric nature of the Hindi film industry. Ever since the decline of the studio system in the aftermath of World War II, the Hindi film industry is star-oriented, star-driven—and, many would complain, star-controlled. In the next section I discuss how language becomes critical to some filmmakers’ attempts to redefine or challenge mainstream paradigms of filmmaking. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/09%3A_No_One_Thinks_in_Hindi_Here__Language_Hierarchies_in_Bollywood/9.03%3A_The_Precarious_Status_of_Hindi_in_the_Hindi_Film_Industry.txt |
From 2006 on, a number of films produced by A-list production companies have utilized local registers of Hindi that set them apart from earlier films.22 With a few exceptions, these films forgo the use of major stars and are set in small towns or subaltern spaces of large cities. Screenwriter Anjum Rajabali commented that the generic Hindustani of earlier eras of filmmaking was disappearing and that in contrast to the past, when characters spoke in the same dialect and register regardless of region or social class, in contemporary films, “characters’ language is rooted in the cultural milieu in which they exist—not just region-specific, but also area-specific, city-specific, and locality-specific.”23 Rajabali explained this shift in terms of a greater concern with authenticity and realism. Devika Bhagat, a writer-director trained at New York University’s Tisch School of the Arts, asserted that it was not possible to have a uniformity of speech across an entire film when its characters represented a wide array of socioeconomic backgrounds. Discussing her directorial debut, Bhagat described the characters and their various milieus: “I have a boy who’s from the Dharavi slums, he speaks in Mumbaiya tapori language. And then there is a middle-class IT professional who speaks in Hinglish, so for each character, the language is specific to their background, their region, their attributes, so that’s why there cannot be a pure form of Hindi anymore.”24
This quasi-ethnographic attention to linguistic detail reflects the tremendous concern of a newer generation of filmmakers with gesturing toward a form of realism in mainstream cinema. One way to index the “real” is through spoken words and dialogue. Even if the rest of the production design is in the realm of fantasy and spectacle, dialogue can mark the rootedness of a film. In many recent films, the setting is actually not integral to the narrative. For example, certain films that showcase a Delhi Punjabi vernacular could have been set anywhere in India, as Delhi was not crucial to their plots; they could have easily been set in Mumbai. These films are anchored to a particular place not by the story but by the language and the register of the dialogue.
Language has become an important way to distinguish among films; it is being foregrounded not just in songs but also in dialogue and speech. Thus language, in terms of dialect, accent, slang, and proverbs, has become an important part of the mise-en-scène, akin to songs, action, locales, and sets. Dialogue has always been important in Hindi cinema, but the turn to the colloquial helps to “dress the dialogue” in a different way. Writers are less reliant on clever turns of phrase or memorable dialogue because mere showcasing of the vernacular is enough. Writers can demonstrate skill not by cleverly crafting witty or memorable dialogue but by merely sounding nonstandard and “rustic.”
Industry professionals offered two main explanations for this turn to the vernacular. One was framed in terms of a backlash of sorts by filmmakers, who were mostly from the Hindi-speaking north and outsiders to both Mumbai and the industry, against the dominant paradigm of filmmaking in the late 1990s and early 2000s. Referring to filmmakers like Anurag Kashyap, who is heavily identified with gritty, violent, dark dramas frequently set in nonmetropolitan sites, writer-director Sriram Raghavan stated, “There was that big phase of the Yashraj and Karan Johar films, which were shot largely abroad. Anurag and that group made it even more specific about certain areas [in India] because it was a reaction to some of these big films—that there were too many of them and they were seeming fake.”25
Director Tigmanshu Dhulia asserted that the artifice of films from this period was a result of Mumbai-bred filmmakers catering only to diasporic audiences: “Because films were being catered to the sensibility of the NRIs [nonresident Indians], the language became fairly easy, poetry was lost, subjects and story lines became very frivolous. Suddenly we stopped making rooted films, and because we were getting revenues from abroad, we stopped making films for Indians.” Dhulia went on to describe how filmmakers who grew up in the Hindi film industry—the second- and third-generation professionals who began their careers in the 1990s—“had not seen India; they’d only seen Bombay.”26 Due to their limited experience of India, according to Dhulia, such filmmakers only made “films about films; they were creating characters out of filmi characters, because they had no experience of India, of life; they thought Bombay is India.”27
Dhulia claimed that people coming from outside Mumbai played a significant role in transforming Hindi cinema and the film industry. In Dhulia’s words, outsiders enabled cinema to “find geography.” Referring to himself and a number of other filmmakers, Dhulia remarked, “We came with our experiences, and so we started making films about the characters we knew, about the region we knew, so that is why the change of character and of language; we became very area specific. Now films have a geography, whereas earlier films didn’t have a geography at all.”28 Sameer Sharma pointed out how even if the “bosses”—those who control finance and distribution—are from Mumbai, many of the directors, such as Anurag Kashyap, are outsiders who are becoming producers themselves and thus are able to green-light or foster films that are set in nonurban or nonmetropolitan settings. With reference to his own directorial debut, Luv Shuv Tey Chicken Khurana, a quirky comedy produced by Kashyap and set in Punjab, which employed Hindi heavily laced with Punjabi expressions, idioms, and humor, Sharma explained that although Kashyap is not from Punjab, he was able to understand a script that was rooted in a small town milieu by virtue of not being from Mumbai. Sharma stated, “It’s important that people who are actually from outside can influence the making of certain films, which may not be understood by a producer who is only from Bombay.”29
Sharma’s remarks illustrate how even filmmakers who self-identify as “outsiders” or as “indie” work very much within mainstream structures of finance, distribution, and exhibition. In fact, their use of Hindi can be seen as another way to assert and perform their “independence” from “Bollywood,” so that their linguistic ability becomes an important form of cultural capital that allows them to distinguish themselves within the industry.30 The overwhelmingly positive critical reception of such directors—including Dibakar Banerjee, Vishal Bharadwaj, Abhishek Chaubey, Tigmanshu Dhulia, and Anurag Kashyap—by the English-language media in India, mostly for their “authentic” portrayals of the “Hindi heartland,” illustrates the success some filmmakers have had with an outsider or renegade image.
However, rather than indexing the “real” or “geography,” the Hindi in such films circulates as an exotic parlance or a simulacrum of the Indian hinterlands within English-speaking cultural spheres. The widely divergent responses between English-language and Hindi-language media regarding Anurag Kashyap’s Gangs of Wasseypur provide a case in point. Screened at Cannes in the Directors Fortnight in 2012, this tale about a long-running blood feud between the families of an outlaw and a corrupt politician in Bihar was widely celebrated in the international and Indian press for “redefining” Indian cinema and identified as a potential crossover success. Interestingly, some media analysts noticed that the Hindi press was quite underwhelmed with the film and pointed out that reviewers for Hindi newspapers dismissed the film’s claims to authenticity and argued that it simply pandered to metropolitan stereotypes under the guise of realism.31 With respect to the dialogue, one analyst pointed out, “The English-language media were fawning about precisely the sort of things the Hindi reviewers noticed as false, including the language with its extravagant crudity.”32 I contend that it is only as a result of English becoming the unmarked, naturalized language of production and discourse within the film industry that filmmakers are able to deploy Hindi as a self-consciously marked commodity.
The second reason for the turn to the colloquial has to do with changes in the political economy of the film industry. The fact that some filmmakers are able to utilize language in a way that would have been regarded in an earlier era of filmmaking as limiting or alienating one’s audience has to do with changing structures of finance, production, distribution, and exhibition that have reshaped the Hindi film industry’s audience imaginaries—issues I have explored in detail elsewhere.33 For example, Tigmanshu Dhulia, referring to his 2012 film Paan Singh Tomar, a biopic about a celebrated and medaled Indian steeplechase runner who is forced by circumstances to become a bandit, stated, “The language was Bundelhi; it was not even Hindi and I was scared that . . . I thought the audience would not even understand the language, but they did! So now cinema has changed, and I think it has changed for the better.”34
One of the biggest changes in the political economy of the Hindi film industry since the advent of multiplexes and corporate production and distribution companies is the diminished significance of the “universal hit”—films that do well all over India and across all demographics. This is due to the structural transformations in filmmaking caused by the entry of corporate production companies and multiplexes, which have altered ideas of commercial success in the industry. Multiplexes, with their high ticket rates, revenue-sharing arrangements, and financial transparency, have managed to transform even low to moderate audience attendance or ticket sales into a sign of success. The entry of the Indian organized industrial sector into film production and the ability of established producers to raise money from the Indian stock market have diminished the role of traditional territorial distributors, who were always perceived by filmmakers as averse to cinematic experimentation. Many corporate producers have ventured into both all-India and overseas distribution and possess a much higher threshold for financial risk. These corporate distributors can either rely on profits from some territories to offset losses from others or profit from their investment by reselling distribution rights to individual territorial distributors. A universal hit is simply not as necessary within this new financing and distribution scenario. Thus there is less anxiety on the part of the financing side of the industry if a film appears limited in its appeal.
Sameer Sharma asserted that filmmakers now have a greater opportunity to express their individual style: “Previously you had to cut across to a whole section of the audience, and your way of making was dictated by the fact that a film should work both in New York and in Patna. But today, it’s become more flexible. I think people have gotten more confident that you don’t have reach out to everybody.”35 Sharma’s statements illustrate how the reduced value of the universal hit within the industry has expanded the criteria of success to the benefit of filmmakers. While the previous structure of the industry rewarded—in terms of both economic and symbolic capital—only filmmakers who strived for universal hits, the contemporary structure enables those filmmakers who are unable to achieve or are unconcerned with broad appeal to also raise money and earn prestige and status within the industry. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/09%3A_No_One_Thinks_in_Hindi_Here__Language_Hierarchies_in_Bollywood/9.04%3A_Hindi_Indie.txt |
Filmmakers’ prestige and status is critically connected to the ability to circulate within elite social spheres, such as international film festivals, and to garner praise from the English-language press in India and abroad, as seen in the divergent responses to Gangs of Wasseypur. Not all “outsiders” or primary Hindi speakers from northern India are able to leverage their linguistic skills in the same manner as the filmmakers mentioned in this essay. Linguistic skill or fluency in Hindi serves as a form of capital only for those who are also fluent in English—that is, filmmakers like Anurag Kashyap, who are internationally celebrated in prestigious film festivals, such as Cannes and Toronto, and garner a great deal of media and critical attention within India. Those film professionals who know only Hindi, with limited proficiency in English, are condemned to remain assistants (to a variety of department heads), dialogue writers for hire, or language tutors, and are frequently marginalized in the social networks that provide a chance at upward mobility in the industry. Therefore, it appears that while not knowing Hindi is not much of a setback or obstacle to participating in the Hindi film industry, not knowing English can be a problem.36
The Hindi film industry has always been and had to be self-conscious and reflexive about language because of its commercial box-office orientation. In the early years of the film industry, language choice was thought about in terms of intelligibility and access to the largest market. Here I have argued that filmmakers consciously consider code choice as a way of marking a film as distinct within a crowded marketplace and of garnering symbolic capital within the film industry. Both are choices born of commercial considerations, but they speak to different moments and transformations in the political economy of the Hindi film industry. Thus, language/code choice helps make visible, or perhaps more appropriately, audible, the changing political economy of the film industry, as well as the changing social relations within it.
9.06: Notes
1 Ananya Vajpeyi, “Hindi, Hinglish: Head to Head,” World Policy Journal 29 (Summer 2012).
2 Hindi, Urdu, and Hindustani are not self-evident and neutral terms, but rather index a long, complex history starting from colonialism when British administrators along with language activists drew boundaries between Hindi and Urdu. At the level of colloquial speech, Hindi and Urdu are mutually intelligible and interchangeable.
3 Tejaswini Ganti, “Mumbai vs. Bollywood: The Hindi Film Industry and the Politics of Cultural Heritage in Contemporary India,” in Global Bollywood, edited by Anandam P. Kavoori and Aswin Punathambekar (New York: New York University Press, 2008); Arlene Davila, Latinos Inc. (Berkeley: University of California Press, 2001); Lucile Davier, “The Paradoxical Invisibility of Translation in the Highly Multilingual Context of News Agencies,” Global Media and Communication 10.1 (2014); Courtney Brannon Donoghue, “Sony and Local-Language Productions: Conglomerate Hollywood’s Strategy of Flexible Localization for the Global Film Market,” Cinema Journal 53.4 (2014); Judith Irvine, “When Talk Isn’t Cheap: Language and Political Economy,” American Ethnologist 16.2 (1988).
4 Tejaswini Ganti, Producing Bollywood: Inside the Contemporary Hindi Film Industry (Durham, NC: Duke University Press, 2012).
5 Eric Barnouw and S. Krishnaswamy, Indian Film (New York: Oxford University Press, 1980); B.D. Garga, So Many Cinemas: The Motion Picture in India (Mumbai: Eminence Designs, 1996).
6 Such actresses are usually dubbed over by voice artists fluent in Hindi.
7 Debashree Mukherjee, “Bombay Modern: A History of Film Production in Late Colonial Bombay 1930–1948,” (PhD diss., New York University, 2015).
8 Ibid.
9 Sumita Chakravarty, National Identity in Indian Popular Cinema 1947–1987 (Austin: University of Texas Press, 1993); Mukul Kesavan, “Urdu, Awadh and the Tawaif: The Islamicate Roots of Hindi Cinema,” in Forging Identities: Gender, Communities and the State, ed. Zoya Hasan (New Delhi: Kali for Women, 1994).
10 For a more in-depth discussion of the social and structural transformations that have beset the industry since the late 1990s, see Ganti, Producing Bollywood.
11 “South Bombay” refers to the oldest and most expensive parts of the city, often synonymous with “old money” and an Anglicized elite, and also happens to be the farthest geographically from the northern and western suburbs that comprise the heart of the film industry.
12 Interview with Ajay Brahmatmaj, Mumbai, August 7, 2014.
13 I have assigned a pseudonym as she is still trying to establish herself within the industry and did not want her frank remarks attributed to her.
14 Interview with Kalpana Chadda, Mumbai, August 2, 2014.
15 Ibid.
16 Ibid.
17 Interview with Kamlesh Pandey, Mumbai, August 28, 2013.
18 The Hindi alphabet contains fourteen vowels, forty-one consonants, and fourteen conjunct consonants—a total of sixty-nine characters, conveying a much wider phonetic range than the twenty-six-character Roman alphabet.
19 See Ganti, Producing Bollywood; and Tejaswini Ganti, “Sentiments of Disdain and Practices of Distinction: Boundary-Work, Subjectivity, and Value in the Hindi Film Industry,” Anthropological Quarterly 85.1 (2012).
20 Interview with Kalpana Chadda, Mumbai, August 2, 2014.
21 Interview with Sameer Sharma, Mumbai, September 2, 2013.
22 Omkara (2006), Ishqiya (2010), Band Baaja Baraat (2010), Paan Singh Tomar (2010), Vicky Donor (2012), or Gangs of Wasseypur (2012).
23 Anjum Rajabali, personal communication, March 27, 2014.
24 Interview with Devika Bhagat, Mumbai, August 29, 2013.
25 Interview with Sriram Raghavan, Mumbai, August 28, 2013.
26 Interview with Tigmanshu Dhulia, Mumbai, January 25, 2014.
27 Ibid.
28 Ibid.
29 Interview with Sameer Sharma, Mumbai, September 2, 2013.
30 Pierre Bourdieu, Distinction (Cambridge, MA: Harvard University Press, 1984).
31 Anand Vardhan, “A Landmark Departure,” The Hoot, July 5, 2012, www.thehoot.org; Shougat Dasgupta, “Art or Artifice,” Tehelka, January 21, 2012, www.tehelka.com.
32 Dasgupta, “Art or Artifice.”
33 Ganti, Producing Bollywood.
34 Interview with Tigmanshu Dhulia, Mumbai, January 25, 2014.
35 Interview with Sameer Sharma, Mumbai, September 2, 2013.
36 Knowledge of English is more important for above-the-line workers than those below the line. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/09%3A_No_One_Thinks_in_Hindi_Here__Language_Hierarchies_in_Bollywood/9.05%3A_Conclusion.txt |
The transformation of the Latin American television industry clearly exposes the profound impact of neoliberal policies throughout the region, including the multiplication of distribution windows, trends toward media concentration, and changes in the modalities by which global media corporations are rooting in local and national television industries. Miller and Leger argue that runaway productions are the means by which Hollywood outsources production to developing countries to realize cost advantages via flexible labor, low wages, low prices, tax incentives, cheap accommodations, and access to material, cultural, and symbolic infrastructure, all the while maintaining tight central administrative and financial control.
• 10.1: Introduction
Overview of the goals of this chapter: to explore the ways in which multinational corporations have approached Latin American indie television companies to create local programming, and the effect of this on labor conditions in these indies.
• 10.2: A Brief Overview of the Latin American Television Industry
A brief history of the Latin American television industry, including nationally owned networks establishing their dominance early on and global conglomerates later attempting to expand into this market approaching local indies for the talent and labor needed to compete with the national networks.
• 10.3: Independents in the Contemporary Landscape of Television Industries
Challenges faced by indie production houses in Latin American television, with a focus on the high level of control exerted by national networks on what projects are able to reach mass audiences.
• 10.4: Genre and Formats Division of Labor
Labor considerations, including cost and the duration for which a production house will be employed, for different genres (in both fiction and nonfiction) and formats of television.
• 10.5: The Asymmetrical Relationship Between Indie Houses and Television Networks
The dependency of indie studios on major national or multinational networks for funding, and the differing levels of creative control exerted by these two network types on the studios. Focuses on the case of the indie studio Argos in Mexico.
• 10.6: Labor Struggles, Accommodations, and Strategies of Survival
Examining the staffing patterns, labor conditions, and emotional/artistic connections common in these indie studios.
• 10.7: Conclusion
The precarity of working at indie production houses, as shown by the emergence of “indies” owned by or closely tied to transnational conglomerates and the diverse strategies needed by other indies to financially survive. Narratives common within these indies that frame this precarity as emblematic of freedom, creativity, and innovation.
• 10.8: Notes
10: Complex Labor Relations in Latin American Television Industries
The transformation of the Latin American television industry clearly exposes the profound impact of neoliberal policies throughout the region, including the multiplication of distribution windows, trends toward media concentration, and changes in the modalities by which global media corporations are rooting in local and national television industries. Miller and Leger argue that runaway productions are the means by which Hollywood outsources production to developing countries to realize cost advantages via flexible labor, low wages, low prices, tax incentives, cheap accommodations, and access to material, cultural, and symbolic infrastructure, all the while maintaining tight central administrative and financial control.1 This New International Division of Cultural Labor (NICL) allows global corporations to expand their transnational presence; however, capital accumulation and profit revenues stay close to the conglomerates’ homes. Given this new media industrial order, Hesmondhalgh argues that media conglomerates acting as large bureaucracies increasingly rely on professionals from small production houses to provide creativity and innovation.2
Given this scenario, labor conventions in Latin American television are changing dramatically. The incursion of global media conglomerates in local markets across the region has caused unanticipated alliances with local independent production houses, or “indies,” which have traditionally been subject to the disproportionate power of the major national television networks in their respective countries. This combination of circumstances has led to disparate outcomes. On the one hand, the presence of global conglomerates has problematically spurred the region’s further incorporation into global capitalism, allowing penetration of Western media in countries where they were formerly confronted by institutional, linguistic, and cultural barriers tied to dynamics of local television consumption. On the other hand, these circumstances present an opportunity for indies to produce different kinds of television projects for national and regional markets, bypassing the long-standing monopoly of national television networks.
The fact that there are just a few indies in each domestic market with the capacity to produce for television networks with national distribution is symptomatic of their precarious status, and exposes their vulnerable position within an industry dominated by national or multinational media conglomerates.3 Following Miller’s reasoning, “Cultural labor incarnates this latter-day loss of life-long employment and relative income security among the Global North’s industrial proletarian and professional-managerial classes”;4 accordingly, the very existence of indies relies on access to media professionals obeying the dynamics of flexible labor, nonunion status, and lack of long-term health or retirement benefits, which are crucial to the sustainability of their business models. These labor conditions are at the core of what Curtin and Sanson describe as “precarious livelihoods,” which are “indicative of a new world order of social and economic instability.”5
I argue that there is a fundamental difference among the production dynamics of film and television in transnational settings that are largely shaped by the relation with local/national audiences. So it comes as no surprise that some authors in this volume, such as Szczepanik in the case of transnational ventures with television conglomerates in the Czech Republic 6 or Keane in the case of television in China,7 recognize the impact of the NICL and its insidious effects on creative labor and local economies and also complicate the landscape, avoiding a causal and unidirectional effect of these global processes. For instance, in Latin America, where Hollywood cinema overwhelmingly dominates the box office in every domestic market, a different scenario appears in television, where the programming of national/regional networks overwhelmingly dominates prime time and achieves high ratings because of the cultural proximity of their products.8 As a result, while multinationals still rely on the advantageous agreements they receive in national markets and the flexible labor their productions require, these global television corporations also need indies and local professionals, who become valuable assets, albeit temporarily, helping to connect their content with local audiences.9
So indies of different sizes with diverse financial and technical infrastructures now participate in projects ranging from low-budget documentaries and journalism to midbudget games and reality shows to the most expensive fictional series, telenovelas, and movies. Indies’ status across the region has grown because they can provide professional staff and talent to produce innovative narratives along with linguistic and cultural input that transnational corporations need to penetrate national and regional markets. In spite of these new opportunities for indie producers, it’s not clear that these new opportunities have as yet had a positive effect on wages and working conditions.10 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.01%3A_Introduction.txt |
Throughout Latin America since the 1950s, television has been the primary audiovisual medium. During this decade, the growth of a national television institution was central to the process of modernization and nation building across the region.11 Simultaneously, U.S.12 technologies and the commercial model of broadcasting 13 were important engines for the growth of the medium across the region; Sinclair and Straubhaar have described this process as an interdependent relationship in which U.S. economic, commercial, and technological interests intertwined with the interests of national economic and political elites.14 Within this pattern, national markets developed in relatively distinctive ways.15 The development of Latin American television was shaped by specific conditions that include market size (population and purchasing power), market structure (monopoly, oligopoly, and competition), broadcasting regulations (private-commercial, state-owned, or hybrid), social and political instability (military coups, social movements, revolutions, and repression), and the visibility and impact of previously established cultural flows across the region, particularly radio and film.
As for content, U.S. programming became a television staple across Latin America during the early years, but national productions took the lead in countries where producers were able to achieve comparable production values.16 Primetime television in major markets, such as Mexico and Brazil, has been largely monopolized by national networks that produced most of their programming in-house. These not only dominated their national markets, they also became leading exporters, and their studios became the primary employers of the television labor force in their domestic markets. In smaller markets with lower production capacities, prime time has been populated with content coming from these major regional producers rather than U.S. studios, due largely to audience preferences for culturally proximate programming.17
In major markets in Latin America, the leading networks were free to produce and broadcast their own content with little regulatory oversight. They built impressive studio facilities and dominated distribution, thereby limiting the possibilities for independent producers. However, the development of the industry across the region did not follow the same patterns. In Mexico, Brazil, and Venezuela, hegemonic networks dominated domestic markets from the very beginning; in contrast, in Argentina, Chile, and Peru, TV production was disrupted or fragmented as a result of dramatic political changes brought about by military coups in the 1970s and authoritarian forms of government throughout the 1980s and the 1990s.18 Colombia, which was similarly affected by political uncertainties, developed a hybrid television model wherein networks were state-owned but the content was largely produced by private production companies (programadoras).19 In the long run, certain levels of media atomization, with an array of independent production entities and market competition, boosted creativity and stylistic diversity in these countries.20
During the 1990s, the implementation of media policies of privatization, deregulation, and liberalization coupled with new technological scenarios triggered the emergence of new windows of delivery through broadcasting, cable, satellite, and the Internet. Deregulation also encouraged a trend toward vertical and horizontal integration that enhanced the muscle of the already powerful national networks, allowing them to build alliances and expand into new sectors.21 However, new national television networks have been launched in parallel, establishing new competitors that are scrambling to secure talent and content to ensure their economic survival. This scenario has been complicated by the slow but steady increase in the presence of transnational networks through cable and satellite television. An increasing localization effort made by global conglomerates (Comcast/NBC-U, 21st Century Fox, Disney/ABC, Sony, Viacom, and TimeWarner/HBO) has resulted in them tailoring programming that could be successful locally in a Latin American market, as well as luring audiences in the U.S. Hispanic market and other countries in the region. Independent producers have thus become a key resource within specific national markets to achieve a successful presence locally and to reach a larger, regional market. This has produced a struggle for talent in which dominant national television networks have tightened their grip, leading to lawsuits against actors and producers working with competing networks.22 Therefore, global conglomerates, such as Sony, 21st Century Fox, and Comcast/NBC-U, have had to look elsewhere for professional teams and talent, increasingly establishing alliances, albeit temporary ones, with indies that house professional labor and talent, through joint ventures, coproduction, or takeovers. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.02%3A_A_Brief_Overview_of_the_Latin_American_Television_Industry.txt |
The Ibero-American Observatory of Television Fiction (Obitel) reports that in 2013, there were forty-eight national television broadcasting networks in seven of the most important Latin American markets.23 Despite this seeming diversity, the domination of these national markets still falls in the hands of a few corporations. Out of the forty-eight networks, only twelve play a prominent role in these seven major markets. In most cases, two networks hold more than 90 percent of the national audience share.24 There is a correlation between these networks’ prominent market position and their capability to produce their own prime-time content, in particular fictional programming. Big-budget projects such as telenovelas are mostly produced in-house by these networks, but in today’s competitive environment these networks increasingly hire indies to produce series and miniseries that require innovative approaches.
The pressure to innovate derives from the fact that the multinationals are making use of cable and satellite services to gain a foothold in national markets. By 2013, 55 percent of Latin American households had access to these new services, and in some key markets, such as Argentina, more than 80 percent of households had access. These changes are reflected in ratings as well. Latin American Multichannel Advertising Council (LAMAC) reports that broadcast TV audience share fell from 86 percent in 2005 to 70 percent in 2014, while pay TV audience share rose from 14 to 30 percent in the same period.25 Moreover, the pay TV audience is generally more affluent and therefore more desirable to ad-based and subscription television companies. Given this media landscape, transnational television networks are increasingly making investments to localize programming with the aid of local indie producers.
The new battle to effectively capture “desirable commercial audiences” in a newly populated industrial scenario brought about innovation but also the rehashing and remaking of “proven ideas and formulas” to successfully appeal to national and transnational audiences. In the case of indies working for the television industry, their size, budget, and production capacities are reflected in the kind of television programming they can offer their clients. In today’s television landscape, there is great demand for documentaries, global formats (particularly reality TV), and fictional formats (particularly series), which range from small- to large-scale productions.
Despite the growing number of independent production houses, most have a short life span, and very few produce for the national television networks that can reach mass audiences. “Here, house productions last very few years,” acknowledged an executive producer of Laberinto Producciones in Colombia.26 While his production company is twenty years old, his assertion underscores the rarity of its position. In a similar statement, an executive producer from Argos Communication, Mexico, states that most transnational companies looking to produce a telenovela would have to select from only a few production houses: “Anybody who pretends to have strong presence in Mexico by producing more than a hundred episodes is going to face complications. In Mexico only we can do that.”27 Similarly, an artistic director from Del Barrio Producciones in Peru recognizes that “there are really few production houses that have the economic means to produce at that level.”28 These assertions reveal the challenges these production houses face in the context of their structural industrial relationship with the gatekeepers of distribution: the television networks. A handful of networks have the power to decide what gets produced and distributed nationally and, ultimately, internationally. National TV networks set the terms in contract negotiations, generally offering one of three options: they offer a flat producer’s fee and retain the copyright; they forge a coproduction agreement based on the investments of the respective partners, with each retaining distribution rights for specific territories; or they agree to broadcast a program that is fully financed by the independent production house, which retains the copyright. In spite of these three options, most indies can’t risk producing their own content because a single failure would likely bankrupt the company. Most commonly, an indie producer will pitch an idea to the networks. If the network is interested, it offers the first option and the producer accepts, reasoning that there will be a secure flow of income to keep the studio staff employed. When dealing with successful producers, networks commonly offer “exclusive” conditions through “volume content agreements,” signing the producer for a number of projects. In other cases, a network will sign an agreement with a successful producer that gives it a “first look” at all new projects. Such agreements allow the network access to top material and give it the option to buy a project and shelve it for an indefinite period so that it doesn’t fall into the hands of a competitor.
10.04: Genre and Formats Division of Labor
The key role indies play at local, national, or transnational levels results from their flexible accommodation to television networks’ needs. The networks’ requirements matched with the capacity and infrastructure of local indies largely define the kind of genre and television formats that indies can afford to produce. Channels like Discovery or NatGeo have employed smaller indie houses to produce television formats that do not require large numbers of employees. Documentaries, docureality, and journalism have become an important source of content production for indies, as they permit cheaper formats and shorter time commitments. At the same time, these productions represent spaces for innovation and creativity by the producers and scheduling flexibility for the networks.
Fictional programming, in contrast, requires major commitments from both network and producer, so larger indies, some with international reputations, tend to prevail in this genre. Fictional programming is costly, requires larger infrastructure, and quite crucially, depends on access to top-line talent. This is especially true with telenovelas, which usually involve the production of around one hundred episodes. By contrast, series and miniseries provide a more secure space for indies for a couple of reasons: economically, they are shorter projects that involve less risk and demand for resources; creatively, they offer scope for innovation that telenovelas rarely offer. While series have fewer episodes than telenovelas, the fact that they continue through different seasons can ensure the economic well-being of the indie and its employees for several years. For networks, in contrast, telenovelas help them make optimal and daily use of their studio infrastructure and human resources. Keeping larger projects like telenovelas in-house allows the promotion of network talent and the full exploitation of commercial opportunities. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.03%3A_Independents_in_the_Contemporary_Landscape_of_Television_I.txt |
Even though independents provide creative possibilities free from the constraints of network in-house productions, most still lack the economic, technological, and labor capacity to produce their own content without the financial backing of a television network or a major sponsor. An executive producer from Blind Spot, Mexico, describes this model as one of “interdependency.”29 Even in cases where a production house has the means to produce its own content, the executive producer argues, “they still depend on having clients who want to buy this specific content.” In some way, the producer from Blind Spot is subtly recognizing its condition of dependency on clients and television networks.
Argos is an interesting example of the challenges faced by indie producers, even though it is one of the most important indie houses in the large television market of Mexico. Its prestige as a leading producer of “quality fiction for TV” was achieved initially when Argos worked for TV Azteca (1995–2000) and produced some of the most innovative telenovelas from that decade. Ultimately a battle over distribution rights led TV Azteca to part ways with the indie in 2000.30 That year Argos reached an agreement with Telemundo to produce 1,200 hours of fictional programming, but in this case Argos used its leverage to secure a share of distribution rights in international markets.31 A screenwriter working with the Argos team recalls that Telemundo was willing to make concessions so long as Argos would shape its production to accommodate a specific U.S. Latino demographic as well as Mexican audiences.32 Argos worked exclusively for Telemundo until 2007, then with TV Azteca again until 2010, when it struck a production agreement with Cadena Tres, a rising new network targeting upscale audiences.33 Argos was encouraged to innovate as a way to enhance Cadena Tres’s profile. As one Argos writer described it, “I was expressly asked to be polemical and controversial with a new project called Las Aparicio. It was delightful to have almost total freedom.”34 Seeking to distinguish itself from the dominant networks, Televisa and TV Azteca, Cadena Tres “had nothing to lose and too much to gain.” However, the precarious positon of independents is interestingly exemplified by Argos’s executive producer when he talks about its relation with Cadena Tres. He remembers that the indie, challenged by the economic crisis of the time, proposed a new business model: “First, Cadena Tres provides part of the cost while Argos provides an in-kind contribution [pago por especie] and some economic investment, and then both look for a third partner.”35 Argos invited Colombia’s Caracol TV as a third partner, but the producer recalls that “when two television networks got together [Cadena Tres and Caracol TV], they asked themselves, what do we need Argos for?” So TV Caracol and Cadena Tres moved along on a new production agreement, leaving Argos behind.
Increasingly, there is a great deal of transnational deal-making between networks and independent production houses, much of it motivated by the desire to attract audiences in multiple markets. The key role of Argos is best exemplified by the indies’ collaboration in Telemundo’s La reina del sur in 2011, a coproduction that included RTI (Colombia) and Antena 3 (Spain). La reina del sur featured a transnational flow of characters with a drug trafficking narrative that was filmed in Bogota, Mexico City, Miami, and Melilla, Spain, with production teams in each location, a strategy I call “reglocalization,” based on a “network cities system of production.”36 La reina del sur proved to be one of the most successful telenovelas in recent memory for Telemundo. Later, El señor de los cielos (2013), a new coproduction of Telemundo with Colombia’s Caracol TV and the collaboration of Argos, became the second most popular production in Telemundo’s history, only surpassed by La reina. That telenovela earned an Emmy in 2014 37 and propelled a new collaboration between Telemundo and Argos to produce new successful seasons of El señor de los cielos 2 (2014), 3 (2015), and 4 (forthcoming), but this time without Caracol TV. This series of successes set the stage for a new programming/production strategy that Telemundo has called “super series.”
The new set of corporate relationships that localized transnational ventures producing Spanish-language fictional programming is best exemplified by HBO’s incursion into Mexico in partnership with Argos and by Sony’s into Colombia in partnership with Laberinto Producciones. After prior experiences in Argentina and Brazil, HBO in 2007 announced its strategic partnership with Argos to produce Capadocia (2008) for Mexican and regional markets.38 By 2014, HBO had experience producing with indies across the region, with eighteen television series produced in Argentina, Brazil, Chile, Mexico, and Uruguay. While working for HBO on Capadocia, one screenwriter remembers that she had a lot of creative freedom, which allowed the inclusion of “a lot of things that were vox populi, but nobody expected to see on television, not even on cable.”39 At the same time, “the expectation was to have an international product with high production values, with high-quality scripts.” Because of these expectations, “the budget and investment of the company were considerably higher,” and she added, “they paid us very well.”40 She recognizes that HBO supervised the project but never really interfered in creative decisions or practiced any kind of censorship.
Sony’s presence in the Colombian television market can be exemplified by the production of Los caballeros las prefieren brutas with indie producer Laberinto Producciones.41 Sony’s partnership with the Colombian indie seems to be similar to HBO’s and Argos’s in Mexico, as described by an executive producer from Laberinto Producciones. Sony read the scripts and supervised the project but did not interfere too much with creative decisions. The producer explains, “For Sony it was an experiment that had its own risks, and they tried to understand how it works from a different creative point of view. They saw that some things function differently in Latin America, and they were open to that.”42
In contrast, argues the Colombian producer, when a national network deals with an indie, it tries to interfere as much as possible in creative decisions to ensure that the product meets the network’s goals. For instance, some of the most sensitive decisions in which networks intervene are in casting, to promote their own talent, followed by interfering in narrative strategies to please clients and their target audience. For example, an artistic director from Del Barrio Producciones in Peru argues that “sometimes we approach the network to pitch a story, and we want to make it with specific leading talent. Sometimes the channel buys the story but objects to the leading talent, and we need to change it.”43 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.05%3A_The_Asymmetrical_Relationship_Between_Indie_Houses_and_Tel.txt |
Creative professionals face an array of challenges in their mostly unstable work settings across the region. These involve a lack of steady jobs and an absence of work benefits such as health insurance, retirement funds, paid vacations, and maternity leave. Most workers are nonunionized, and their employers are themselves subordinate to the power of national networks and multinational conglomerates. Both workers and management sense the precariousness of their situation. Their only leverage resides in the growing demand for content and the fierce competition for talent. Given that, the relationship between these rather small business entities and their creative labor is shaped by complex and intertwined conditions of professionalism, emotional attachment, creative styles, and in many cases “family”-style ties and relationships.
There are several sizes and modalities of production indies bring to the audiovisual realm. Many houses produce advertising and corporate and institutional content as a way to maintain a healthy income. They also keep their payrolls trim, commonly anchored by three to ten permanent office employees and a network of freelance creatives. Indies typically rely on the prestige and connections of particular professionals, mostly the producers themselves. So, while a company might be initially founded by a famous actor, director, writer, producer, or venture capitalist, most commonly its ongoing operations are led by a producer who is the owner, co-owner, or CEO of the company.
While varying in size, indies nevertheless tend to have similar staffing patterns. They routinely hire administrative personnel for the office, including accountants, administrators, secretaries, and office assistants. In larger companies, especially those producing fictional programming, there is also a second group of permanent employees, mostly technicians and manual laborers. In some cases, there is an editor on staff and a few other below-the-line professionals who enjoy permanent contracts and legal benefits such as health benefits, retirement funds, vacations, and compensation in case of a layoff. Yet they employ mostly a nonunionized workforce.
When launching a particular project, the company hires above-the-line creative professionals on temporary contracts at higher wages but with no legal benefits. These professionals are nonunionized, and although many are paid quite well, their work stints are unpredictable, punctuated by regular periods of unemployment without benefits. These professionals include writers, directors, actors, photographers, casting experts, art and costume designers, and many others. They constitute a floating labor force that offers its services either to independent producers or to fill the needs of network in-house production teams on particular projects. However, there are some working conditions that vary by country. In Mexico, freelance laborers take care of their individual health expenses and save for retirement. In Uruguay, individuals are hired through a personnel company that provides benefits, while in Colombia workers have benefits via payroll.
In terms of stability, freelancers are in the most vulnerable position, always pursuing the next gig by offering their services to multiple clients in different audiovisual sectors. At the same time, some workers embrace the flexibility and freedom of being able to change jobs and negotiate working conditions. An executive producer from the Uruguayan indie house Microtime argues, “There are professionals who prefer a freelance status because that allows them to manage their lives. They take vacations when they want. They do not want to ask permission from their boss, producer, and that kind of flexibility works very well for them.”44 At the same time, this kind of freedom comes with a downside: according to an Argos screenwriter, “I cannot plan my life more than two or three months in advance. If I have earned money from a project, I need to save and take care until I find the next one.”45 Interviewees explained that forging a good professional reputation is essential to keeping themselves in the labor force. Intelligence, talent, technical and artistic abilities, and work ethic are considered to be essential elements for survival. Work continuity and stability are linked to gaining the producers’ trust.
Interestingly, the small number of highly visible indie productions and their small size tend to create tight relationships within professional circles. Work relations are permeated with emotional ties and a sense of common artistic purpose. Some executive producers from indie houses refer to their permanent staff members and regular freelancers as family, saying they want to take care of professionals who have worked with them for a long time by offering a sense of stability. An artistic director from Del Barrio says it’s “because we know that these workers have families, and they have worked for us too many years. So, yes, there is an emotional element.”46 Thus, just as emotional and pecuniary relations are intertwined, so too is the artistic sensibility that permeates some of the indie shops. This is often described as a common outlook on how television should be made, a shared understanding in the realm of either aesthetics or ideology. As opposed to the conventions of in-house network productions, these creative teams believe that indie shops are the perfect space to make something different. The professional profiles of the leading indie producers seem to permeate the institutional culture of their production houses, creating personal bonds around common artistic goals, including innovation and brand distinctiveness that set them apart from the dominant network studios. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.06%3A_Labor_Struggles_Accommodations_and_Strategies_of_Survival.txt |
The new media landscape, with its myriad windows of video distribution and the changes brought about by digital conversion, appears to offer new opportunities to regional indie producers. It remains to be seen, however, if these opportunities can be translated into real improvements in labor conditions. There are already urgent questions about low pay and unstable labor conditions prevalent in the region.
Moreover, the transformation of the media landscape as a result of larger structural processes offers a complex industrial scenario in which the integration of the region into global capitalism leads to battles between national and transnational corporations. In this scenario, different entities are taking advantage of the structural conditions and flexible labor of indie producers. A visible element that underscores the differences among national television industries has been the nature and role of indie productions within these national markets. However, in spite of these differences, the indie production sphere first emerged in relation to powerful national television networks. Those relations have recently been challenged by the emergence of a new generation of “indies” wholly owned by large conglomerates or independently owned but closely related to them. The presence of transnational companies now competing with national networks or collaborating with them offers opportunities to professionals, but at the same time they are competing in an industrial space that would otherwise have been occupied by local companies. FoxTelecolombia (21st Century Fox), Teleset (Sony), RTI (NBC), Cuatro Cabezas (EyeWorks), Endemol, Zodiak (Di Agostini), and FremantleMedia are striking “volume agreements” with television networks that allow them continuity in production but also show the large professional and financial capabilities of these “indies.”
Beyond these new transnational indies, national and local entities are struggling to survive by deploying a variety of strategies. Their existence and roles go far beyond the realm of television industries, offering a diversity of services to different companies and producing a variety of programming, including films, documentaries, and corporate or educational videos. Consequently, they need to be flexible and scale their workforce to the shifting demands of their clients. As described by the executive producer from Blind Spot, producers need to have the capacity to bring the right people to a specific project, but they also need to have the economic resources to deliver. Prestige, capacity of delivery, and proven success are required elements in this equation. Success in delivery defines some level of continuity for these indies, while failure may lead to their demise.
Similarly, professionals working for indies offer their services to an array of potential clients, opening doors for possible future projects. To lessen the anxieties of job uncertainty, professionals actively work at networking and diversifying their skills. Some of the professionals revealed that they combine short-term and long-term strategies to survive. As the artistic director from Del Barrio Producciones explains, working for advertisers pays well, while producing television series is not as profitable; however, waiting for a paycheck from an advertising firm takes several months, while working for television offers a monthly paycheck.
The defining feature of independent production houses is that they are separate from the corporations that own the means of content distribution: the television networks. This definition is also at the center of their vulnerability, which has been reframed by professionals as a space of opportunity. While lacking the stability of permanent jobs, these professionals are motivated by notions of innovation and creativity as well as specific ideological convictions and aesthetic commitments. They believe that talent and skill can create success, while their emphasis upon gaining the trust of producers is a prevailing notion that fits into a market-oriented economic approach which requires the illusion of free competition. Within this ideological framework, this free-floating army of professionals seems to conceive of unionized labor and hiring quotas as constraints upon the very specific creative needs of particular projects. The lack of permanent job status is also reinterpreted as a lifestyle choice representing agency and freedom. Following Bourdieu’s explanation of the dynamics of the field of cultural production, these professionals take innovation, socially–oriented narratives, and quality production as their reasons for working for indies as an assumed restricted space within the larger field of television production.47 Paradoxically, the precarious conditions of this sector seem to be precisely the ideological engine that supports professionals’ imagined conditions of freedom, creativity, and innovation. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.07%3A_Conclusion.txt |
1 Toby Miller and Marie C. Leger, “Runaway Production, Runaway Consumption, Runaway Citizenship: The New International Division of Cultural Labor,” Emergences 11.1 (2001): 89–115.
2 David Hesmondhalgh, Cultural Industries, 2nd ed. (Los Angeles, CA: Sage, 2006).
3 In the last five years there has been a reformulation of telecommunication laws in the main domestic markets across the region with respect to quotas on the distribution of national and independently produced content with economic incentives for independent productions; however, the results of such legislation are uneven.
4 Toby Miller, chapter 2 in this volume.
5 Michael Curtin and Kevin Sanson, chapter 1 in this volume.
6 Petr Szczepanik, chapter 7 in this volume.
7 Michael Keane, chapter 16 in this volume.
8 Joseph Straubhaar, “Beyond Media Imperialism: Asymmetrical Interdependency and Cultural Proximity,” Critical Studies in Mass Communication 8 (1991): 39–59.
9 Among the most visible indies are Argos, Adicta Films, El Mall, Lemon, and Canana Film, Blind Spot in Mexico; RTI, Teleset, Vista Producciones, FoxTelecolombia, Laberinto Producciones, BETV in Colombia; Cuatro Cabezas, Underground, Ideas del Sur, RGB, Chris Morena, Pol-ka, and Endemol in Argentina; Del Barrio Producciones, Imizu, Teatro Libre, and Sol Entertainment Producciones in Perú; and Bueno Puerto Producciones, Valcine Producciones, Wood Producciones, and My Friend Entertainment Producciones in Chile. Conspiração Filmes, HBO O2 Filmes, and Mixer Brazil represent the most visible examples of a much larger group of indies.
10 The recognition of the Latin American television context is the product of years of research as part of the Ibero-American Observatory of Television Fiction (Obitel); but this chapter is based on ethnographic work done at NATPE in summer 2012, and a handful of interviews done in Fall 2014 with high executives and above-the-line creative personnel from independent production houses: Blind Spot and Argos (Mexico), Laberinto Producciones (Colombia), Del Barrio Producciones (Peru), and MicroTime (Uruguay). The pool of interviewees from independent production houses is composed of six top executives. Their shared characteristics are active involvement in the production of television programming for television networks in their countries as well as involvement in coproduction or commissioned productions, particularly fiction, with a U.S. or global media conglomerate.
11 Jesús Martín-Barbero, “Memory and Form in the Latin American Soap Opera,” in The Television Studies Reader, ed. Robert Allen & Annette Hill (London: Routledge, 2004).
12 James Schwoch, The American Radio Industry and Its Latin American Activities, 1900–1939 (Urbana: University of Illinois Press, 1990).
13 Noreene Janus, “Advertising and the Mass Media in the Era of the Global Corporation,” in Communication and Social Structure: Critical Studies in Mass Media Research, ed. Emile McAnany, Jorge Schnitman, and Noreene Janus (New York: Praeger, 1981).
14 John Sinclair and Joseph Straubhaar, Latin American Television Industries (London: British Film Institute, 2013).
15 Elizabeth Fox, Latin American Broadcasting: From Tango to Telenovela (Bedfordshire: University of Luton and John Libbey Media, 1997); Elizabeth Fox and Silvio Waisbord, eds., Latin Politics, Global Media (Austin: University of Texas Press, 2002).
16 Joseph Straubhaar, World Television from Global to Local (Los Angeles: Sage, 2007).
17 Joseph Straubhaar, “Beyond Media Imperialism: Asymmetrical Interdependency and Cultural Proximity,” Critical Studies in Mass Communication 8 (1991): 39–59.
18 Guillermo Orozco, ed., Historias de la televisión en América Latina (Barcelona: Gedisa, 2003).
19 Ibid.
20 Nora Mazziotti, La industria de la telenovela: La producción de ficción en América Latina (Buenos Aires: Paidós, 1996).
21 Raúl Trejo, “Muchos medios en pocas manos: Concentración televisiva y democracia en América Latina,” Revista Brasileira de ciencias da comunicacao 33.1 (2010): 17–51.
22 For instance in Mexico, TV Azteca sued Alan Tatcher for working with indie Nostromo in a new reality show for Telemundo in 2006; TV Azteca also sued talent working with indie Argos in a series for Cadena Tres in 2013.
23 There were five in Argentina, six in Brazil, seven in Chile, five in Colombia, seven in Ecuador, five in Mexico, six in Peru, four in Uruguay and fourteen in Venezuela. Guillermo Orozco and María I. Vasallo, eds., Transmedia Production Strategies in Television Fiction (Porto Alegre, Brazil: Globo Comunicação e Participações and Sulina Editora, 2014).
24 In Mexico, Televisa and TV Azteca hold 95 percent of the audience share; in Colombia, RCN and Caracol TV hold 97 percent; in Venezuela, Venevision and Televen 81 percent; in Brazil, TV Globo holds 40 percent; in Argentina, El Trece and Telefe hold 60 percent; in Peru, America TV, ATV and Frecuencia Latina hold 86 percent; and in Chile Canal 13, Chilevision, and TVN hold 70 percent of the audience share. Ibid.
25 LAMAC, “Penetracion de TV de paga,” Latin American Multichannel Advertising Council, www.lamac.org/.
26 Interview with executive producer from Laberinto Producciones, Colombia, November 25, 2014.
27 Interview with executive producer from Argos, Mexico, January 29, 2013.
28 Interview with artistic director from Del Barrio Producciones, Peru, November 27, 2014.
29 Interview with executive producer from Blind Spot, Mexico, November 19, 2014.
30 Telemundo kept the distribution rights in the United States, Canada, and Puerto Rico. Mary Sutter, “Telemundo, Argos to Link,” Variety, October 13, 2000, http://variety.com/2000/tv/news/tele...nk-1117787691/.
31 Joe Flint, “Productora Mexicana dará telenovelas a Telemundo,” Mural, October 2000, 5.
32 Interview with screenwriter for Argos, Telemundo, TV Azteca, HBO, Cadena Tres, August 22, 2006.
33 Nikolas Maksymiv, “Imagen y Argos Anuncian una Alianza,” Noticias financieras (Miami), January 21, 2010, ProQuest (466639295).
34 Interview with screenwriter for Argos, Telemundo, TV Azteca, HBO, and Cadena Tres, November 26, 2014.
35 Interview with executive producer from Argos, Mexico, January 29, 2013.
36 Juan Piñón, “Reglocalization and the Rise of the Network Cities System in Producing Telenovelas for Hemispheric Audiences,” Journal of International Cultural Studies 17.6 (2014): 655–671.
37 Sara Bibel, “Telemundo’s ‘El Señor de los Cielos’ Wins First-Ever International Emmy for Non-English Language U.S. Primetime Program,” TVbythenumbers, November 25, 2014, http://tvbythenumbers.zap2it.com/201...rogram/332190/.
38 Hernán Casciari, “HBO con acento latinos,” El país, February 2009, http://blogs.elpais.com/espoiler/200...to-latino.html.
39 Interview with screenwriter for Argos, Telemundo, TV Azteca, HBO, and Cadena Tres, November 26, 2014.
40 Ibid.
41 Marie A. De la Fuente, “Sony Will Sell Telenovelas at Mip,” Variety, March 29, 2009, http://variety.com/2009/film/news/so...ip-1118001742/.
42 Interview with executive producer from Laberinto Producciones, Colombia, November 25, 2014.
43 Interview with artistic director from Del Barrio Producciones, Peru, November 27, 2014.
44 Interview with executive producer from Microtime, Uruguay, December 1, 2014.
45 Interview with screenwriter for Argos, Telemundo, TV Azteca, HBO, and Cadena Tres, November 26, 2014.
46 Interview with artistic director from Del Barrio Producciones, Peru, November 27, 2014.
47 Pierre Bourdieu, The Field of Cultural Production: Essays on Art and Literature (New York: Columbia University Press, 1993). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/10%3A_Complex_Labor_Relations_in_Latin_American_Television_Industries/10.08%3A_Notes.txt |
Just as Hollywood production frequently departs the greater Los Angeles area for less expensive shooting locations worldwide, Hollywood studios have also expanded their interests globally, investing in everything from Bollywood studios to telenovela-producing corporations.¹ In this sense, the Los Angeles-based film and television industry is indeed a multilevel “global Hollywood,” as Miller and his colleagues convincingly illustrate in their so-titled book.² Accordingly, the individuals who make up global Hollywood’s workforce are both geographically diverse (in “runaway production” locations from New Orleans to Prague) and industrially diverse, working on domestic film and television productions as well as major international projects.
• 11.1: Introduction
Overview of the goals of the chapter: to examine how the deliberate pursuit of informality that characterizes Nollywood shapes issues of precarity with local labor.
• 11.2: The Context – Nollywood the Place
The decentralized nature of Nollywood, and the proposal of two places that might be considered the location of Nollywood: Lagos and Alaba Market, both known for having a generally informal and undocumented culture.
• 11.3: Implications of Opaque Distribution
Labor implications of the informal nature of Nollywood distribution, including the high level of power held by marketers, the dominance of guilds rather than unions, and the inability of government regulation to improve labor conditions.
• 11.4: Precarity, Repeat Collaboration, and Industry Entry
Examination of the factors contributing to the precarity of work in Nollywood for a non-marketer: short project durations that require workers to quickly find their next job, lack of recourse to collective bargaining or legal protection, and the heavy reliance on preexisting personal connections to break into the industry.
• 11.5: Conclusion
Ways in which the informality that characterizes everyday life in Nigeria is ingrained in Nollywood, as well as efforts from creatives in the film industry, government, and foreign interests to limit this informality.
• 11.6: Notes
11: Labor in Lagos Alternative Global Networks
Just as Hollywood production frequently departs the greater Los Angeles area for less expensive shooting locations worldwide, Hollywood studios have also expanded their interests globally, investing in everything from Bollywood studios to telenovela-producing corporations.1 In this sense, the Los Angeles-based film and television industry is indeed a multilevel “global Hollywood,” as Miller and his colleagues convincingly illustrate in their so-titled book.2 Accordingly, the individuals who make up global Hollywood’s workforce are both geographically diverse (in “runaway production” locations from New Orleans to Prague) and industrially diverse, working on domestic film and television productions as well as major international projects, all of which increasingly rely on Hollywood capital. Global Hollywood’s workforce, then, may include the labor on a Universal Studios movie shooting in Prague, labor on a Bollywood movie that is partly funded by a subsidiary of Sony Pictures, and labor on a telenovela produced by a company with ownership links to NBC. While the dynamics of employment differ among locations, as other chapters in this book illuminate, interest and investment from Hollywood require significant transparency in distribution and management at the very least. This generally means a corporate structure with a few behemoth companies dominating production and neoliberal governance: local and national policymakers and large private companies working in tandem to establish the sort of regulations, policies, and practices, including reliable copyright and contract enforcement, that are attractive to foreign direct investment (FDI) and formal domestic bank investment.
Entertainment production worldwide also exists outside Global Hollywood’s networks. In this chapter, I critically examine the relationships among labor, distribution, informality, and power in one such industry: the massively popular southern Nigerian movie industry known as Nollywood.3 Its productions dominate screens and mediascapes across sub-Saharan Africa and throughout the global African diaspora, though exact numbers about its production output and income are challenging to discern.4 Counting and demonstrating sales are not just relevant to demonstrating an industry’s importance for academic study or popular journalistic pieces. Rather, opacity (and the accompanying general inability to codify formal sales figures) is a defining part of Nollywood’s structure and strength,5 shaping nearly every part of the industry’s day-to-day operations and practices. In particular, this opacity reinforces the power of the film distributors known as “marketers,” who leverage their gray-market knowledge to control the Nollywood marketplace. While an industry predicated on personal relationships may appear to risk breeding disorder, Nollywood is in fact quite organized, a result of the marketers’ self-governing practices and the industry’s guild-based infrastructure. Consequently, Nollywood remains largely disconnected from formal global networks of labor organization, financing, and distribution,6 but is nevertheless a nexus point for its own set of global flows and linkages.
Using a series of onsite observations and interviews with practitioners in the Lagos-based industry, my analysis reveals how global concerns about the precarious nature of local labor are shaped in this context by the particular brand of informality that characterizes Nollywood. If we take precarity and informality to be linked, we can see Nollywood as a particularly informal industry with a particularly precarious workforce, marked by limited recourse for labor grievances. I assess Nollywood’s informality as a phenomenon forged out of a very specific place: Lagos, a rapidly growing, often overflowing megacity and an alternative media capital, a hub for global flows and connections that utilizes few of the formal dominant networks that mark Global Hollywood.7 In this way, this chapter grounds the reality of local media labor in the specificities of the actual places where that labor works. In short, the structure of Nollywood reflects the specific architecture and shape of Lagos.
I would also like to be specific about what I mean when I discuss industrial informality in Nollywood. Film industries from Hollywood to Bollywood can be said to feature informal elements at many levels of production, especially in relation to labor practices like recruitment. And Nollywood features some formal elements in its production inputs and distribution outlets.8 The distinction between formality and informality, then, is not a dichotomy. Rather, it is a continuum with no industry falling fully at either extreme. Additionally, the question of what exactly informality is and whether it should be celebrated has been subject to much scholarly debate, particularly in the context of media distribution studies, and Nollywood has been at the core of many of these arguments.9 There has been concern over the potential to exoticize and “Orientalize” Nollywood via an overabundance of focus on the informality in the industry.10
Accordingly, to demystify the discussion of Nollywood’s informality, I wish to be very clear about what exactly I mean by informality in Nollywood, and the ways in which it is a conscious choice in a global power play. This study understands the basic intersecting constituents of Nollywood’s industrial informality to be 1) not documenting sales or most other distribution figures in any publicly accessible/scrutinizable fashion, 2) not utilizing legal contracts for employment or other business relationships, 3) not using agents or other formal inputs such as accredited schools for talent recruitment, 4) not pursuing copyright violations via legal frameworks, and 5) privileging undocumented financing and distribution networks and spurning alternatives. It may be noted that four of these five elements begin with the word not. This is because our understanding of informality as an industrial feature worthy of mention exists only because of the existence of formality in these areas in other industries. While there may be some level of informality in other global movie or television industries, the dominance and intersection of these elements in Nollywood’s day-to-day functioning render the industry predominantly informal as opposed to the fragmented informality that characterizes the global media industries that compete for Hollywood’s production, coproductions, partnerships, or investments on the international stage. And point 5 underlines the conscious and active choice by Nollywood’s marketers to utilize informality as a means to maintain power and thwart challengers. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.01%3A_Introduction.txt |
“Where is Nollywood?” asked a neophyte on a public Nollywood message board. The mirthful and mocking responses were plentiful. “Nowhere!” said many, while others were more specific, citing places where one can, indeed, see Nollywood at work. One answer is Surulere. This is the neighborhood in mainland Lagos where many producers, directors, and other creative professionals maintain offices and live. A more specific answer might be O’Jez’s, a bar in Surulere’s National Stadium that serves as a meeting point for socializing and making business deals. With the exception of O’Jez’s, chance encounters with Nollywood elites are a rare occurrence in Surulere. Offices, workspaces, and production sites are unmarked; the streets are largely residential. Jonathan Haynes, writing on the geography of Nollywood, notes that the small amount of capital per entrepreneur means that large spaces marking flashy movie industries—studios, theaters, large office complexes—don’t exist here, as Nollywood functions largely behind small unmarked doors.11 It’s a massive industry that remains hard to see and hard to quantify.
Another answer to the message board query could be Alaba Market, a vast sprawling electronics market on the outskirts of Lagos, which also serves as Nollywood’s distribution nerve center. Journeying there in a taxi, one emerges from the city’s densely populated urban maze into a dusty spread of low-lying disconnected buildings speckling the landscape before arriving at the market itself. Alaba is a city unto itself, with streets, churches, banks, and apartments, all low, dusty structures built from inexpensive materials. The market, according to a rough and unsourced estimate from over a decade ago, may be the epicenter of 75 percent of West Africa’s electronics trade, may house 50,000 merchants, and may net \$2 billion each year.12 At Alaba, one can purchase anything from new flat-screen televisions to used generators to, of course, movies for home viewing.
Alaba’s location is a logical one for the largest Nigerian (and West African) electronics market, directly between two sources of product importation: one formal (the Apapa port) and one informal (the Benin border at Seme).13 The peripheral location isolates Alaba from government officials, allowing it to thrive on formal neglect. The market can spread as far as it would like without running into anything that the city would consider important enough to protect or regulate. The only efforts at delimitation are internal, and the market’s infrastructure is mostly self-made. Merchants have private radio-wave towers to ensure mobile phone service, and operate private generators to ensure power. In their study of Alaba as urban form, architect Rem Koolhaas and his colleagues reference a statistic that, even though it may lack veracity, gives an idea of the scale and atmosphere of the market: that Alaba has the highest concentration of generators in the world. Alaba’s self-governance has also included private development of a parking lot, local secretariat, fire station, and local library.14
Despite its fragmented connection to formal trade and governance, Alaba has forged its own global network and emerged as a central hub in the circulation of electronics in West Africa, as well as in Nollywood’s own circulation networks. The market mirrors Lagos itself, a global megacity that is often said to be growing “off the grid.” Possibly home to 21 million,15 with less than a third connected to public water supply,16 Lagos may be on its way to becoming the third largest city in the world, depending on how you count and who is counting. In understanding Lagos, the fungibility of its population estimates speaks to the culture of Lagos at large: mostly informal, undocumented, and difficult to officially count for those who make their living counting such things (and it is worth noting that counting such things has significant financial implications, as population and business figures directly affect applications for everything from loans to grants).
Both Alaba and Lagos are central locations in the production and distribution of Nollywood titles. The growth of their infrastructures serves as not just context but also metaphor for the logics that guide industrial operations in Nollywood. The next two sections of this chapter will detail those specific conventions, focusing on the informal networks that structure Nollywood’s industrial organization and labor processes. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.02%3A_The_Context__Nollywood_the_Place.txt |
While Nollywood produces movies, it is not technically a “film” industry. Movies are not shot on celluloid nor are they usually intended for big-screen projection in cinema houses. Instead, Nollywood movies are shot on video and largely viewed in private or in small public screenings. Though there are constant televised screenings of Nollywood movies, some informal public screenings in video parlors,17 and a growing trend of some higher-end titles utilizing a splashy initial release and short run at expensive cinema houses,18 the vast majority of profits in Nollywood come from physical direct-to-consumer sales. Most Nollywood movies are financed and distributed by one group of people, known as “marketers.” Despite the name, they serve multiple roles for each movie: executive producers, marketers, and distributors. Essentially small-scale entrepreneurs with experience in the gray- and black-market electronics trade, marketers leverage their knowledge of Nigeria’s informal, undocumented marketplaces and open-air bazaars to structure their business dealings in the movie business. While many creative workers in the industry bemoan the lack of everyday cinema houses, affordable to the poor and lower middle class and contributing to a local cinema-going culture,19 the marketers flourish in an environment that leaves them in control of most authorized distribution.
Once finished movies have been pressed at the disc replication plant, the marketers package them with the appropriate graphics and release them to the markets on the next Monday. These movies flow through distribution hubs and subhubs, usually first focused on places like Alaba (the Lagos market) and Onitsha (the city in the Igbo-dominated southeast that is home base for a large number of production companies). Copies are then sent to nearby cities for sale at their markets, and they fan out to smaller hamlets from there, much the way electronics fan out from Alaba across the nation. These distribution networks are held together through trust, personal connections, and informal exchange as opposed to legally binding contracts, a prominent feature in any informal economy.
“Piracy” is part of both the heritage and the current functioning of the video distribution system,20 although I will refer to it here as unauthorized distribution in order to remain value neutral. Despite the marketers’ public proclamations that unauthorized distribution is decimating the industry, they themselves are not fully operating on the legal side of copyright law in all of their business dealings. Instead, one might say that they operate in a gray area, obtaining certain rights for movies and then overstepping them and hiding profits. The core professional experience of the marketers comes from a background in electronics trading, including unauthorized distribution of foreign movies. As Brian Larkin has illustrated,21 their success in distributing the movies they produce comes in no small part from using the same distribution networks they forged years before to distribute unauthorized copies of Hollywood, Bollywood, and Asian action movies, and is augmented by their ability to operate behind closed doors and out of the sight of any potential regulators.
The current process of financing an average Nollywood movie is inextricable from the informality of its distribution. Again, while movie industries worldwide seek diverse funding sources (of varying degrees of legitimacy and reliability), the avoidance of transparent distribution in Nollywood delimits the potential to attract bank loans or other formal investors, as investors tend to require confirmed sales figures and reliable sales projections. The marketers’ often antagonistic relationships with the government over taxes and other documentation issues mean that loans and grants upon which other film industries can rely are not part of the landscape for most moviemakers in Lagos. Instead, individual marketers tend to finance their own productions. Those in charge of distribution—the marketers— are the only ones who can make financial decisions and calculate risk, a confidence that comes from their exclusive control over (informal) distribution networks and associated knowledge of the (opaque) marketplace. This oversight establishes a level of collective power among the marketers that is intentionally difficult to usurp and that repels investment attempts by outsiders.22
Moreover, these marketers zealously guard their power by policing the informal relationships that enshrine their authority, which is particularly important as we analyze the structure of the industry. The strength of their informal networks trumps most attempts at formal takeover, and this strength is derived from informality: opacity in sales figures and distribution networks, and informality in industrial organization. Informality, however, is not the same as describing the industry as a disorganized, chaotic collective. In the absence of governance by legal institutions or the centralized formal power of major corporate studios, control is enforced by a mass of small enterprises whose internal organization helps preserve their collective interests, best represented in the marketers’ guild (FVPMAN).
FVPMAN is just one of the many guilds constituting Nollywood’s internal infrastructure.23 Guilds, each boasting an elected national leadership, represent almost every aspect of the industry from marketers to makeup artists. Guilds provide, in essence, internal governance for the industry, standing in for legal contracts and labor regulations. While they appear centralized and visible, akin to unions, guilds are neither transparent nor formal, neither registered with nor regulated by the government, and subject to no external oversight. For instance, in the absence of legal contracts, guilds are meant to solve disputes. In theory, a grieved party takes his or her grievance to the guild’s leadership, who work to solve the dispute with the offending party’s guild. In practice, however, a dispute between an elite and an underling will rarely result in a disruption of the status quo, and there is no recourse to formal legal litigation as a corrective. Nollywood sets are full of empty complaints about labor practices, from unpaid labor to unsatisfactory working conditions.
While most guilds deal primarily with internal labor issues, the collective power of the marketers’ guild, FVPMAN, is immense. In the past the guild has made attempts to space out movie releases to counteract periodic “gluts” in movie releases. FVPMAN also has cut down on production at times to address the same issue. Attempts to dethrone the marketers have consistently ended in failure, whether the attempts come from blocs of creative workers or government authorities, both wishing more control over the industry. For instance, Nollywood stars are a central mechanism through which movies are branded and sold to the public. Marketers rely on their images to help secure financial success. Nollywood stars are thus widely recognizable and glamorous, though fame secures most of them only a modest fortune. To make a consistent living, most stars must work frequently, and those demanding extravagant fees can fall out of favor with the marketers. FVPMAN acts quickly to shut down productions or blacklist actors if the organization feels the marketers’ power is being chipped away by an overentitled celebrity. Heightened budgets and salaries in a bigger-budget nonmarketer branch of Nollywood, now known as “New Nollywood,” have yet to evoke industrial change: workers must work so frequently that New Nollywood’s limited slate won’t sustain them, and the marketers still set the terms for the bulk of the industry.
For such a young industry, Nollywood is subject to constant speculation on the shape of its future. Various plans to shift that future in one direction or another come from both within and outside the industry. New plans to “formalize” in one way or another are near constant, including cinema construction schemes and various licensing initiatives put forth by the government and guilds. Some of these new ideas die before they are born, while others persist but only affect a small subset of the industry during their tenure. As with most cultural industries, the real power of the industry is centralized with those who control distribution. In Nollywood, this still means the marketers. Locating power in creative industries is key to understanding their functioning, and understanding the marketers as the nexus of power here is key to understanding Nollywood’s persistent informal infrastructure. This informality can be a source of industrial strength to reinforce dominant power structures, and thus integration into formal global networks is unlikely to serve the best interests of those in authority (here, the marketers). This informality also marks the experience of labor and the nature of production and distribution in Nollywood in a manner that exceeds anything seen in industries that rely on theatrical release, official legal contract enforcement, and relative transparency in distribution statistics. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.03%3A_Implications_of_Opaque_Distribution.txt |
Creative work and precarity in labor tend to go hand-in-hand, as the individual case studies in this volume collectively illustrate. In creative industries worldwide, creative workers tend to form new teams for each project.24 This can range from the culturally temporary work structure in the technology sector, in which workers expect to work at companies for only a few years before shifting job title and project,25 to the extremes of per-project team reformation in the creation of movies, television shows, and songs. On this end of the creative industries, workers are largely freelance and must constantly find new work as they coalesce for the completion of a single project.
In movie industries based on theatrical release logic, being hired on a film may mean several months of work on a single project—perhaps even a year or more— depending on one’s role in the process. Nollywood instead operates on an industrial logic favoring rapid, inexpensive production. Instead of pouring money into a single title and laboring over an extended time frame, most Nollywood production moves at a brisk pace, as can be seen in other straight-to-video industries worldwide.26 This is an extreme form of precarity, as a living is forged in Nollywood from working every week or every day, going from production to production. As informality in distribution entrenches power in the hands of the marketers, it also heightens precarity for Nollywood’s nonmarketer labor force, as they lack recourse to legal protections and collective bargaining.
Nollywood production is marked by velocity. It is not unusual to shoot ten scenes in one day, with two to three weeks of shooting per movie a common scenario.27 With one week of preproduction and one week of editing and packaging, a video film can go from inception to sale in four weeks, though three months is a much more common scenario, and New Nollywood titles take even longer. An indemand worker can easily shoot two movies in one month. With modest pay from any individual movie, workers make a living mainly through quantity, and some can be found working nearly every day, ending one movie project to begin another.
Movie industries worldwide are mostly marked by freelance labor agreements, with crews coming together in new and different formations for project after project. Researchers from Caves 28 to Currid 29 have noted the weight this gives to informal relationships that may bridge friendship and business: if you must rely on your reputation to secure employment in a business marked by whom you know, social relationships are a core motor of your career and professional development. Despite the prominence of guilds in Nollywood, people usually hear about jobs from either someone they have worked with before or the recommendation of a friend. While this is not unusual in other movie industries, the lack of formal recourse to government labor regulation, talent agents, or managers marks Nollywood as particularly informal. Furthermore, the sheer number of projects that Nollywood workers must pick up to support themselves—many more than in movie industries marked by theatrical release of higher-budget projects—means that the informal is even more important to workers as a means through which to ensure their continued financial stability. Because of this quantity-based logic in production and labor, repeat collaborations are particularly common, and maintaining trust among individuals in those working groups is an essential survival tactic.
This trust is not easy to come by: breaking into the industry can be challenging. Entry into Nollywood is often based on family, ethnic, or other preexisting ties. There is little formal training in the industry, and many workers learn along the way. Some efforts to institute training programs via moviemaking schools have begun, but these have not become a reliable mechanism for feeding talent into the industry. Gathering places, like O’Jez’s, the bar and restaurant in the National Stadium, can provide another opportunity to network one’s way to the top. Industry events such as premieres, awards ceremonies, and elaborate birthday banquets are the type of invitation-only places where major business deals are negotiated. A bar like O’Jez’s, which is open to the public, is less likely to yield dramatic results. It does, however, provide the chance to see and be seen, and people make an effort to hold personal meetings there, in order to be observed by others in the industry. In such instances, industry aspirants may find themselves only a few degrees removed from a critical phone number; personal introductions are common occurrences in these locations, helping novices connect with senior players in the industry.
Apprenticeships are often a low-cost entry point into the industry, especially for producers and technical crew. It is commonplace across Nigerian industries for a “big man” to train a number of “boys” to work under his mentorship,30 and it is no different in Nollywood. This tradition is thought to ensure personal loyalty. One midlevel producer I spoke with, for instance, runs a production company whose in-house editor is the producer’s former barber, someone the producer trained for the position. This way, the producer says, “I know he will always be loyal to me.” Other industry workers teach themselves. One postproduction special effects artist I interviewed, for instance, learned his craft from free online tutorials on special effects software, such as Video Co-Pilot, Cinema 4D, and After-Effects. His training was a side pursuit based on personal interest while he was enrolled in another field of study at university. University education is common among the creative arms of the industry (producers, directors, editors, and so on) though not a requirement. Marketers are usually not university educated, as they often work in the marketplaces from a very young age and gain their knowledge from those experiences. Some of the rancor and distrust between marketers and directors is based on that point alone.
Guilds structure the industry. While crew and big-name stars are hired through personal connections, nonstar actors are usually enlisted through auditions. Auditions tend to be formal, well-attended affairs, with members of the leadership of each guild expected to attend and make sure everything is operating in a respectable fashion. For guilds with many members and limited work—for instance, the Actor’s Guild—guild membership also forms the framework through which creative workers look for and find work. The power of the guild, regardless of who is in charge, stands in for a legal system in industry disputes, and it can also serve as a mechanism for resisting external interference in the industry status quo, be it from governmental or foreign interlopers. And it is the marketers’ guild, FVPMAN, where industry power is concentrated, in an opaque and cohesive collective of small- to medium-size distributors. The core of government efforts to control Nollywood has been aimed at interrupting this bloc and promoting the emergence of a few corporatesque national distributors that could be more easily controlled. This governmental strategy of control (actively promoting the emergence of a few corporate giants as opposed to a diversity of dispersed small distributors), implemented in many government bids for control over media industries worldwide (perhaps most notably in early U.S. radio development), has thus far failed in Nollywood, as the cohesiveness and collective opacity of the marketers, schooled by years of actively avoiding government notice, has made them challenging consolidation targets.
We can thus see informal, undocumented transactions as the building blocks of the industry, structuring its organization and labor processes. And we can see extreme precarity as characteristic of Nollywood labor. While informal connections and trust also form the basis of the working relationships of most other contract-based creative industries, including most of those highlighted in this volume, what is distinctive about Nollywood is that this is the only currency. There is no recourse to formal political or legal systems or institutions, like talent agencies, to help structure industry operations. At once a source of immense strength for those who control the distributive mechanisms, the southern Nigerian film industry’s self-governed informality remains an ever-present challenge for workers who lack equal footing and for those looking to codify it with standards based on structured global networks, from which the industry remains disconnected. Such tensions have and will continue to shape Nollywood. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.04%3A_Precarity_Repeat_Collaboration_and_Industry_Entry.txt |
The functioning of Nollywood as an industry is inseparable from its location in Lagos, in Nigeria, and from a place of disjuncture with dominant formal global entertainment industry production and distribution networks. Alaba’s rise to regional centrality in the outskirts of the urban core mirrors Nollywood’s rise in a state of disjuncture with formal media industry networks and domestic government oversight. In both settings, a functioning industrial order emerged from an architecture that would be inhospitable to corporate formality. In Alaba, we can see individual fixes through the individual mobile phone towers and the sea of personal generators blanketing the previously barren landscape on the side of the highway from central Lagos to Benin. In Nollywood, we can trace this thread throughout the industry. Financing, for example, is usually done by the eventual distributor in the absence of reliable sales estimates or accountability. Production relationships are built on trust, not contracts, and entry to the industry is rarely through formal schools, as apprenticeships acquired through personal connections rule.
Another commonality is that both Alaba and Nollywood share deceptively organized governance. While both have been mostly ignored by Nigeria’s and Lagos’s actual government,31 both are indeed governed: self-governed. In Alaba, we can see this through the libraries, firehouses, and schools built by the massive collective of small merchants housed in the market. These merchants are held together by the urge for self-preservation as well as the Nigerian tradition of group organization. In the same way, we see Nollywood’s marketers (some of whom are the very same small-stall owners of Alaba) controlling the industry with the firm hand of confident self-organization. They maintain star salaries at a manageable level, control gluts, create stars, and maintain distribution networks that rapidly disseminate new cultural products to the most remote of Nigeria’s hamlets. Unlike the precarity defining the work of most of Nollywood’s labor, the opaque organization of Nollywood’s marketers means they enjoy relative stability, as they themselves control the industry. Although they are threatened by “illegal” distribution practices, they are also strengthened by them, particularly those of their own genesis, and they have recourse to their non-movie side businesses, including electronics trading.
At the same time, it is important not to overromanticize the informal. While the marketers are happy with the current system, those on the industry’s creative side as well as foreign and government forces have made and continue to make significant efforts to delimit the industry’s informality in favor of an industrial structure with room for bigger budgets, theatrical screenings as a norm, and wider global recognition—in short, an industry in which they could achieve their artistic visions while still selling to their core domestic audiences. Yet we can see the current brand of informality that marks Nollywood as closely linked to the environment from which it was born: an environment that encourages small-scale enterprises with opaque business practices, meant to avoid notice by government officials. This informality thrives particularly well in areas characterized as both the urban and the global margins, even as they may be central in their own alternative networks. In this way, we can see the specificities of the local ingrained in the everyday realities of media labor in Nigeria’s internationally popular movie industry. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.05%3A_Conclusion.txt |
1 Amelia Arsenault and Manuel Castells, “The Structure and Dynamics of Global Multi-media Business Networks,” International Journal of Communication 2 (2008): 707–748.
2 Toby Miller, Nitin Govil, John McMurria, Richard Maxwell, and Ting Wang, Global Hollywood 2 (London: British Film Institute, 2008).
3 This chapter does not directly compare Nollywood to Hollywood. To compare the two is to suggest that Nollywood is a less-funded imitator of Hollywood, when it is more accurate to view Nollywood as on its own trajectory, growing out of and flourishing in a different position in the global economy. However, in the context of this edited volume, I engage in a general comparison of Nollywood to more formal global media industries worldwide, as part of this book’s charge to illuminate how global concerns about labor issues play out across a diversity of contexts.
4 It has become popular in recent years to cite a UNESCO statistic that Nollywood is second only to Bollywood in number of titles produced per year, but scholars have questioned the significance of such a statistic in comparing theatrical and home-viewing based industries. See Carmela Garritano, “Introduction: Nollywood—an Archive of African Worldliness,” Black Camera 5.2 (2013): 44–52; and other articles in that Nollywood special issue of Black Camera.
5 See Brian Larkin, “Degraded Images, Distorted Sounds: Nigerian Video and the Infrastructure of Piracy,” Public Culture 16.2 (2004): 289–314, for a full argument on informal distribution as a source of industrial strength.
6 Despite recent efforts in formalization in cinema, online, and satellite distribution, the core of industry profits are still in physical copies sold in networks of domestic open-air markets.
7 Jade Miller, “Global Nollywood: The Nigerian Movie Industry and Alternative Global Networks in Production and Distribution,” Global Media and Communication 8 (2012): 117–133.
8 See ibid. for an analysis of Nollywood’s formal inputs, such as cameras and sound equipment.
9 See Ramon Lobato, “Creative Industries and Informal Economies: Lessons from Nollywood,” International Journal of Cultural Studies 13 (2010): 337–354.
10 Matthew Gandy, “Learning from Lagos,” New Left Review 33 (2005): 36–52; Alessandro Jedlowski, “Nigerian Videos in the Global Arena: The Postcolonial Exotic Revisited,” Global South 7.1 (2013): 157–178; Nyasha Mboti, “Nollywood’s Aporias Part 1: Gatemen,” Journal of African Cinemas 6 (2014): 49–70.
11 Jonathan Haynes, “Nollywood in Lagos, Lagos in Nollywood Films,” Africa Today 54.2 (2007): 131–150.
12 Figures, to be taken with many grains of salt, are derived from Rem Koolhaas, Harvard Project on the City, Stefano Boeri, Sanford Kwinter, Nadia Tazi, and Hans Ulrich Obrist, Mutations (New York: ACTAR, 2000).
13 Ibid.
14 Koolhaas et al., Mutations. And counting comparative global generator density by neighborhood is perhaps the epitome of figures that could never be accurately counted.
15 See Elizabeth Rosenthal, “Nigeria Tested by Rapid Rise in Population,” New York Times, April 14, 2012, www.nytimes.com/2012/04/15/world/africa/in-nigeria-a-preview-of-an-overcrowded-planet. html. This statistic is at the high end of estimates; the population may well be lower.
16 National Bureau of Statistics of Nigeria, Annual Abstract of Statistics, 2012, 115–117, www.nigerianstat.gov.ng/nbslibrary/nbs-annual-abstract-of-statistics/nbs-annual-abstract-of-statistics.
17 Those who made the movie see no direct profits from this.
18 Jonathan Haynes, “New Nollywood: Kunle Afolayan,” Black Camera 5.2 (2013): 53–73; Moradewun Adejunmobi, “Evolving Nollywood Templates for Minor Transnational Film,” Black Camera 5.2 (2014): 74–94.
19 Connor Ryan, “Nollywood and the Limits of Informality: A Conversation with Tunde Kelani, Bond Emeruwa, and Emem Isong,” Black Camera 5.2 (2013): 168–185.
20 See Larkin, “Degraded Images, Distorted Sounds.”
21 Ibid.
22 Those advocating loudest for more formality are those who would likely lead the industry were the marketers to lose control. This small group of big-name Nollywood producers and directors make movies that emerge from a largely separate self-financed system known sometimes as “New Nollywood.” This chapter, however, deals not with “New Nollywood” but with the bulk of the industry, which produces the majority of titles and employs the vast majority of workers.
23 It shouldn’t be all that surprising that the institution of guilds has proven so popular in structuring primarily informal Nollywood. Nigerian society is full of organizations and leadership positions. It is not uncommon for people to spend their little spare time going from meeting to meeting: church governance groups, church committees, groups of those originally from the same village, and so on. It seems that anyone who is anyone (and many who are, in effect, nobodies) holds or has held a leadership position in some organization or another. Ascendency to leadership in any organization is afforded a high degree of respect and importance, and leaders of even the smallest of these organizations are usually hailed by their title in public.
24 Richard Caves, Creative Industries: Contracts between Art and Commerce (Cambridge, MA: Harvard University Press, 2002).
25 Annalee Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Cambridge, MA: Harvard University Press, 1996).
26 Ramon Lobato, Shadow Economies of Cinema: Mapping Informal Film Distribution (London: British Film Institute, 2012).
27 A popular cinematographer suggests that ten to fifteen days is the bare minimum for a movie shoot, while a “good” movie will take twenty-one or more days. Shooting has been taking longer in recent years.
28 Caves, Creative Industries.
29 Elizabeth Currid, The Warhol Economy: How Fashion, Art, and Music Drive New York City (Princeton, NJ: Princeton University Press, 2007).
30 In Nigeria, a wealthy, successful man is usually referred to as “big,” and underlings and servants are often referred to as “boys,” no matter their age.
31 With the exception of sudden dramatic overtures, such as former president Goodluck Jonathan’s disjointed multimillion-dollar funding scheme late in his tenure or the censorship board’s ill-fated efforts to restructure distribution. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/11%3A_Labor_in_Lagos__Alternative_Global_Networks/11.06%3A_Notes.txt |
In this chapter, we examine the conditions of precarity in porn work, situating those conditions in the context of a changing industry and a political economic moment in which uncertainty is the most stable feature. Though we can link precarious conditions to their social contexts, we do not suggest that such conditions are inevitable or historically neutral. As Chuck Kleinhans insists, “Precarity is not a necessary result of [global political economic] changes. Rather, it is a deliberate policy and aspect of neoliberalism in its relation to the labor force.”¹
• 12.1: Introduction
Overview of the chapter goals: to examine the precarity of labor faced by workers in the porn industry, and the ways in which these workers navigate and resist the precarity they confront.
• 12.2: Contours of the Industry
Mapping the structures that fall under the umbrella of the porn industry, including those who actively take part in producing adult films, the other industries that aid in film production and distribution, organizations that support or oppose porn, and the satellite industries such as erotic dance and escort that provide additional income streams to the performers.
• 12.3: Conditions of and Responses to Precarity
Sources of precarity in the adult film industry, including the lack of labor regulations regarding pay, discrimination, and occupational health, performers' status as independent contractors, and the risk of being blacklisted for speaking out. Some of the ways in which performers have confronted this precarity, including diversifying industry streams and developing unconventional means of collective resistance.
• 12.4: Do-It-Yourself Ethics, Class and Boundary Work
The tension in production dynamics between stable performance work at mainstream studios, and having the freedom to express personal brands and escape degrading stereotypes by self-producing porn at the cost of greater economic precarity.
• 12.5: Notes
12: Creative Precarity in the Adult Film Industry
In this chapter, we examine the conditions of precarity in porn work, situating those conditions in the context of a changing industry and a political economic moment in which uncertainty is the most stable feature. Though we can link precarious conditions to their social contexts, we do not suggest that such conditions are inevitable or historically neutral. As Chuck Kleinhans insists, “Precarity is not a necessary result of [global political economic] changes. Rather, it is a deliberate policy and aspect of neoliberalism in its relation to the labor force.”1 Porn workers’ precarity emerges out of an industry struggling in the wake of global recession, rampant piracy, and a hostile legal environment, but precaritizing policies are not a necessary response to these socio-political conditions. Instead, deliberate policies make porn workers precarious—policies ranging from independent contractor laws that excuse employers from labor regulations and proscribe union organizing, to formal and informal anti–sex worker codes that render sex workers especially vulnerable to both state and employer abuse, and of course, the mundane but not inevitable rules of the wage relation under capital.
We are equally interested in creative precarity—the resourceful ways porn workers resist, navigate, and exploit the precarity they confront. We suggest that taking seriously these forms of resistance can deepen our understanding of precarious labor in creative fields and in the world of work more broadly. Why? Because the conditions of precarity that appear to be recent historical developments in other industries have long shaped porn work. The “new gig economy [emphasis added],” brought on by “massive changes that have generated the expansion of precarious employment,” is not, for instance, so new for porn workers, who have long pursued diversified income streams to get by.2 The adult film industry is not exceptional, then, but it may be predictive. Workers’ struggles there speak to the conditions that increasingly characterize labor in the current political economy, and we are well served to pay attention to the strategies they deploy in confronting them.
We must begin by mapping the adult film industry, because so little is known about it, and what is thought to be known—such as the oft-repeated claim that it is a \$10–12 billion industry—turns out to be completely made up though almost never challenged. The establishment of porn studies as a scholarly discipline— with the inauguration of the journal Porn Studies in 2013, the growing number of university courses offered and dissertations undertaken, the availability of more archives and collections for historical research, and the efforts of academics and industry professionals to engage in productive conversations about the current and future shape of the industry—helps make this mapping possible. The Feminist Porn Book: The Politics of Producing Pleasure is the first collection, for example, to bring together writings by feminists in the adult industry and essays by feminist porn scholars.3 But even as space opens up for academic discussions of pornography, the casualization of the professoriate and the erosion of the academic freedom ensured by tenure bring their own precarity to researching controversial areas such as the adult industry. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/12%3A_Creative_Precarity_in_the_Adult_Film_Industry/12.01%3A_Introduction.txt |
Though increasingly diffuse, the adult film industry occupies a central role in the spatial and political economy of California’s San Fernando Valley. We focus our inquiry there in the interests of space and precision, while also attending to the growing production centers of Las Vegas, San Francisco, south Florida, and globally, Brazil and Eastern Europe. A small roster of production and distribution companies, including Manwin, Bang Brothers, Brazzers, West Coast Productions, Evil Angel, Wicked Pictures, Larry Flynt Productions, Playboy Enterprises, and Vivid Entertainment, dominates the adult film industry landscape. But as production costs rise and potential profits from large-scale productions decrease,4 small, boutique production companies producing niche content increasingly populate the adult film industry. In using the term industry, we do not suggest a monolithic, static, or internally consistent body. Instead, we mean to indicate the dynamic networks of workers, management, and institutions that take part in the production process of adult film, all of which are affected by regulatory policies such as Measure B, the 2012 Los Angeles County mandatory condom law that saw more than a 90 percent drop in adult film production permits issued.
Studio executives, investors, producers, talent agents, directors, crew, performers, postproduction editors, and distribution and marketing staff are key players in adult film production. Common institutions connect these actors: trade publications distribute industry news and host annual trade and award shows; the industry’s trade organization, the Free Speech Coalition, lobbies on its behalf and, since the 2010 collapse of what had been an industry-run health clinic, sets the terms for recommended sexually transmitted infection (STI) testing panels and exposure protocols; the Adult Performer Advocacy Committee (APAC), since the fall of 2013, provides worker education such as the Porn 101 video and brings performers together to advocate on their own behalf in discussions of testing protocols and other informal policies; and private but industry-specific testing clinics clear performers for work. Other institutions and actors, while not of the porn industry, are intimately connected to it: multinational software development firms design web platforms and process credit card payments, real estate agents coordinate filming locations, beauty service providers specialize in readying performers for work, publicity firms cater to performers and adult businesses, and nonprofit organizations such as the Aids Healthcare Foundation build political identities and funding bases through their relationships with (or stark opposition to) the industry. In describing the contours of the “industry,” we think it is important to include organizations and institutions that could not exist without the adult industry, such as for-profit “porn addiction” therapies, religious antiporn initiatives like the XXX-CHURCH, which sends its preachers on the college circuit to debate with porn stars (“Jesus loves porn stars!”), and the antiporn feminists who spend extraordinary amounts of time and energy fighting not only the adult industry but those who think it merits study rather than blanket condemnation (Stop Porn Culture).
We also understand the “industry” to encompass the satellite industries—including erotic dance, webcam, escort, and novelty—that enjoy a symbiotic relationship with the adult film industry. This relationship has three dimensions: first, income streams from satellite industries economically sustain adult film performers, securing a reserve army of performer labor for whom the film industry is not financially responsible. Were such income streams not available, it would be difficult if not impossible for performers to maintain themselves amid the vicissitudes of demand, filming schedules, industry, and other factors that mean a performer might work twenty days one month and two the next.5 Second, many performers describe the increased earnings they can draw from satellite industries by marketing themselves as “porn stars” as a primary reason for taking on porn performance.6 Dominic Ace, an adult industry publicist and photographer who has worked as a roadie for performers on feature dancing tours, explained it this way: “You’ve got web sites, you’ve got Skype shows, you’ve got privates [escorting], you’ve got fan clubs, you’ve got custom videos, appearances, feature dancing, Verified Call [a service that connects fans to performers via cell phone], a ton of different revenue streams. . . . You don’t make money doing scenes, a scene is a marketing tool [emphasis added].”7 Talent agents for film frequently recruit in erotic dance clubs and on webcam sites, and adult actresses report having begun careers in these fields, later moving into the film industry. Finally, production companies and agents who sign performers to exclusive contracts may be, depending on the specific terms of the contract, entitled to a percentage of workers’ earnings in satellite industries.
Porn workers push the boundaries of the industry to meet their financial needs, as well as satisfy desires for autonomy, flexibility, and work-life balance. Porn performer and single parent Raylene explained that her average take from three to five hours of webcamming work was comparable to her film performance rate, but webcamming allowed her to have greater control over her schedule and working environment: “I was able to work alone, in my house, during school hours, and then, you know, have the rest of the evening with my child and make a better living at home than when I was in front of the camera.”8 Those performers who prefer satellite industries to adult filmmaking describe taking just enough film gigs to maintain their “brands.” In line with Dominic Ace’s description of scenes as “marketing tool[s],” performer Venus Lux noted, “When you’re in porn, especially transsexual porn, it’s not a money making thing. It’s for the fame, that’s it. The chain reaction of the fame means you can eventually get money.”9 Management too is keenly aware of the industry’s reliance on satellite industries. Christian Mann, a longtime board member of the industry’s trade organization and general manager of distribution giant Evil Angel, compared the porn industry’s increasing reliance on alternative profit streams to similar trends in the mainstream music industry. “The reality is,” he wrote, “albums don’t make money anymore. Record stores are gone, right? So the saving grace for the music industry has been concert tickets.”10 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/12%3A_Creative_Precarity_in_the_Adult_Film_Industry/12.02%3A_Contours_of_the_Industry.txt |
Performers’ deft cobbling together of various income streams and tactical manipulation of their personal “brands” give the lie to the idea of porn workers as passive victims.11 This is not to say that porn workers do not confront modes of work organization that constrain their autonomy and working conditions and threaten their well-being. It is, instead, to center on the creative ways in which they resist such conditions of precarity. Similarly, porn workers do not simply react to top-down management; just as often, changes in management style represent capital’s desperate responses to workers’ manipulating the system in ways management never anticipated. One performer explained, for example, that it is possible to identify clauses of an exclusive performer contract that correspond directly to work-arounds developed by previous contract stars. Listing individual clauses, she named each one after the performer who discovered a new way to assert power in the workplace. If this is a sobering reminder of management’s tireless drive to constrain worker resistance, we might also remember that workers are often one step ahead.
We now take a step back to sketch the conditions of precarity confronted by porn workers. We will then return to an exploration of the ways they can be understood as creatively precarious. Like other industries in advanced capitalism,12 the adult film industry more and more relies on a flexible, itinerant, and deskilled workforce. While the pool of porn performers was a small and close-knit one in the 1970s and early 1980s, today’s seemingly endless supply of eager new performers limits current workers’ ability to negotiate the terms of their labor with agents, producers, and directors. Internet piracy and employers’ increased interest in casting amateurs further depress wages and job opportunities. The growing popularity of amateur aesthetics also decreases the demand for professional camera operators, directors, editors, and scriptwriters.
The porn industry operates free of most external labor regulations governing pay, employment discrimination, occupational health, and benefits. This is in part due to the industry’s liminal legal status. Industry-specific regulations typically focus on age record-keeping requirements 13 and obscenity prosecutions,14 rather than working conditions. Regulations governing occupational health, wages, and employment discrimination are clumsily borrowed from noncomparable industries, such as nursing in the case of blood-borne pathogens. Recent attempts at passing industry-specific workplace health legislation have proven unsuccessful. The Safer Sex in the Adult Film Industry Act passed in Los Angeles County in November 2012 has remained largely unenforced. In the face of overwhelming dissent from performers, anemic legislative support, and concerns about funding and enforcement, AB 1576, the proposed statewide legislation mandating a whole range of rigorous health and safety requirements, including condom use and employer-provided STI testing, failed in August 2014. Workers and management oppose external policy on the grounds that it would undermine what they maintain is the industry’s robust and effective self-regulation, which includes twice-monthly STI testing for performers and industry-wide filming moratoriums in the event of a positive HIV result. Indeed, workers suggest that that the industry’s testing system has suffered in the wake of the 2011 downfall of the Adult Industry Medical Foundation (AIM), which served as an autonomous and centralized testing and treatment clinic. Significantly, outside organizations campaigning for greater state involvement in the porn industry’s health protocols were instrumental in AIM’s closure.
Adult industry workers’ precarious legal status is solidified by their designation as independent contractors. To a large extent, independent contractor law is organized explicitly to excuse employers from their responsibilities to workers. Employers can, fully within the bounds of the law, pass on to workers a broad range of production costs, including STI testing, wardrobe, makeup, and transportation. Workers have little legal protection from discrimination in hiring or pay disparity. Rates for black women performers are a fraction of those of their white counterparts, for instance,15 plus-sized performers too are underpaid, and male performers can be blacklisted based on rumors of their having had same-sex sexual encounters. Independent contractor status means that workers cannot legally unionize and that they have fewer legal protections in the event of retaliation against even informal organizing efforts.
Independent contractor status also affects porn workers in ways that extend beyond the letter of the law. Employers and workers alike make a host of often inaccurate assumptions about legal entitlements based on what they assume being an independent contractor entails. Though producers are legally required to secure production insurance, few do, and this gives workers little recourse in the event of on-set injury or infection. Even in uninsured workplaces, workers are entitled to make workers’ compensation claims against their employers but rarely do. Standard industry rhetoric maintains that the nature of porn work makes identifying the precise cause of (and hence the party responsible for) a work injury difficult, but employers escape financial responsibility even for those injuries that are plainly traceable to a particular set. In one extreme instance, veteran performer Prince Yahshua sustained significant injury to his penis during a scene. The injury required \$120,000 in surgery and follow-up care that left Yahshua out of work during his months-long rehabilitation. He covered these costs out of pocket save for a \$20,000 check the production company sent of its own accord. When asked why he chose not to file a workman’s compensation claim, Yahshua pointed to his independent contractor status. He added, “It worked itself out,” noting that he has since continued to work consistently in the industry.16 Other workers who reported having been injured on set suggested that paying medical costs out of pocket was a small expense in comparison to the wages they would surely lose had they filed a claim.
In spite of the various ways independent contractor status can increase profit for employers and vulnerability for workers, most workers do not identify establishing employee status as a priority, and many have found ways to make the independent contractor status work for them. Performers find tax and legal benefits associated with incorporating their own names and brands—a number of performers are their own LLCs—which is not possible for employees.
On a more abstract level, performers report that they prefer the idea of working for themselves, perceiving that this affords them greater autonomy as they negotiate schedules, wages, and work tasks. Independent contractor status may give workers more freedom to seek out alternative income streams, another way performers can be understood to be creatively precarious.
Adult film performers are skilled at diversifying income streams, a strategy that has become increasingly important as both performance rates and casting opportunities in film diminish. In addition to the satellite industries we previously outlined, performers maximize their incomes by creatively monetizing quotidian moments of their lives: they sell their used underwear, make money while sitting in Los Angeles traffic by charging fans for a cell phone chat, and command fees for opening their birthday parties to the public. Performers also make marketing opportunities out of the mundane, sharing Twitter photos of their morning showers, fitting DVD signings into family vacations, and engaging fans as they watch favorite sports teams. Though these opportunities could be read as discomfiting evidence of the market’s encroachment into even the most intimate spaces of workers’ lives, they could also be said to represent workers’ creative strategies for negotiating precarity. Part of what we find so instructive about porn work is that both things are undoubtedly true.
The Amazon “wish list” is a nearly ubiquitous feature on performer web sites and social media. Performers invite fans to buy them lingerie, sex toys, and cosplay gear, but also novels, records, and daily essentials such as vitamins and shampoo. Performers self-consciously use wish lists as a means to supplement unpredictable earnings and, sometimes, to compensate for payment they feel production companies wrongly withhold. Gay porn performer Conner Habib offered this explanation to his Twitter followers: “Why is it okay for porn stars to have wish lists? [Because] we don’t get royalties even though studios get our images forever.”17 Other performers have suggested that they find it hard to be too concerned with piracy when sales only enrich production companies. Were residuals and royalties standard practice, performers might make more of an effort to encourage fans to “pay for [their] porn,” as the industry slogan goes. As it stands, it may be more efficient for performers to leave antipiracy advocacy to employers and focus their marketing efforts on the alternative income streams for which their porn performances serve as advertisement. Performers are acutely aware of the areas in which they have power, and they manipulate them brilliantly.
Facing the threat of retaliation and legal barriers to formal organizing, porn workers devise creative methods of not only individual but also collective resistance. We caution against a view of labor organizing that recognizes only those forms of action legible in law and mainstream union movements. Apart from various unsuccessful attempts to join the Screen Actors Guild, porn-worker organizations have not, for the most part, sought to replicate a labor union model.18 Instead, they focus on mutual assistance, information sharing, and education. Club 90 in the early 1980s served as an education and support group and inspired an off-Broadway play in which Club 90 members performed.19 Led by Nina Ha®tley 20 in the late 1980s, the Pink Ladies’ Social Club served as a support group but also a space in which performers shared material information about rates, working conditions, and which bosses were best to work for. Under Har®tley, a trained nurse and veteran performer, the organization provided health information, educating performers about which sex acts posed the greatest risk of sexually transmitted disease transmission, safer sex methods, and the signs of sexually transmitted infection. Ha®tley has continued to play a key role in industry organizing, and held a leadership role in the Adult Performer Advocacy Committee.
With a series of on-set HIV transmissions in 1997 and 1998, performers again came together to emphasize health in their organizing efforts. Founded by former performer Sharon Mitchell, the Adult Industry Medical Foundation (AIM) served as a centralized testing and treatment clinic and provided a space for performer education, offering the video primers Porn 101 and 102 to new performers curious about how to negotiate rates, STI risks, consent, and financial matters such as the importance of paying your taxes in a state where it is legal to have sex on camera. The Erotic Entertainers Guild (1997) and Adult Performers Union (2003) focused on establishing a wage floor and continued to push for performer-centered healthcare protocols. These organizations have been short-lived, due in part to industry management’s consistent harassment of the workers involved. Ha®tley explained, for example, that even Pink Ladies’ Social Club, hardly a militant organization, drew management retaliation: “We were instantly branded as lesbian unionizers and barely worked for six months.”21
Fear of management retaliation may partially explain more recent groups’ special efforts to distance themselves from labor unions and any suggestion of labor-management conflict. The Adult Performers Association (2011) made explicit that “everyone in the industry will benefit from our research and efforts [emphasis added],”22 but its leaders, Nica Noelle and January Seraph, were nonetheless subject to harassment and threats.23 The Adult Performer Advocacy Committee (APAC, established in 2013 and still operating) has similarly positioned itself as a voice for performers, but one not in conflict with industry management. Reviving AIM’s educational tradition, APAC produced an updated version of Porn 101, introducing new performers or those just thinking about going into porn to topics ranging from sexual health to contract negotiation. APAC has met with greater institutional support, with the porn industry’s trade organization (the Free Speech Coalition) initially offering meeting space and legal counsel and its trade magazines disseminating APAC’s press releases. This level of support may owe to APAC’s leadership, which includes top performers in the industry, many of whom also hold management roles.
Citing parallel features, including competition among workers, the transience of the workforce, the reality that workers hold multiple positions simultaneously, and management’s concentrated power, industrial relations scholar Gregor Gall suggests that craft organizing of the sort Dorothy Sue Cobble describes in waitresses’ unions might allow for organizing in the porn industry.24 An additional challenge of organizing porn work is the frequency with which those involved shift between management and worker roles. In addition to pursuing various satellite industries (and making new subindustries of their own), porn workers resist precarity by shifting between the roles of manager and worker. After a short time in the industry, most performers will have at least dabbled in management, producing content for their own web sites or clips stores, working as directors for established production companies, or starting production companies of their own. This fluidity challenges the strict class divisions that have been central to state, activist, and academic approaches to labor organizing. That performer groups can consist of worker-managers does not nullify their organizing work, but it no doubt affects the organization’s perspective and priorities. Seeking a purer organization untainted by management interests misses the point, though, because the potential to shift between worker and manager roles is indispensable to workers seeking control over labor processes. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/12%3A_Creative_Precarity_in_the_Adult_Film_Industry/12.03%3A_Conditions_of_and_Responses_to_Precarity.txt |
In addition to resisting the vulnerabilities precarity brings, some porn workers describe precarity as both a potential job benefit and what allows them to be creative. Shifting between worker and manager roles is one way porn workers respond to precarity not by seeking greater stability but by exploiting flexibility to their advantage. Though some performer-cum-managers, like the iconic small business owner, simply prefer to be their own bosses, autonomous production is also a space in which workers refuse status quo labor practices, casting opportunities, rates, and representational politics. Worker-produced porn also makes managers of workers, generating conflicting interests and riven class positions. The medium trades in tensions that orthodox analyses of creative labor cannot account for.
At the most basic level, self-producing gives worker-managers control over the products in which they are featured. In an industry in which the most successful performers carefully craft their personal brands, what benefits a performer’s brand may be less advantageous to agents, directors, and studio heads. Authorial control can be a powerful tool. Performers may choose to wait to perform anal sex, for example, until they can command the highest rate possible, and many also perceive that slowly doling out new types of scenes to fans helps to ensure career longevity or, in industry speak, to avoid getting “shot out.” Agents, however, prefer performers who “do everything” right away, as this ensures more bookings (and thus commissions) in the short term. With a self-replenishing reserve army of labor, agents have little interest in counseling performers for longevity. Directors and producers too have ready access to new talent and are most invested in the current production’s profits. In this context, performers can choose to self-produce the sort of content, such as a first anal scene, that promises higher sales. Leading up to these productions, workers can use performances for other production companies both to gather start-up capital and to advertise the self-produced content from which they profit most. Worker-producers may also use self-production as a long-term planning strategy to the extent that continued sales can generate earnings for years, money that contract performers (who do not receive royalties) will never see.
Though mainstream porn is home to a proliferation of small production companies led by current and former performers, do-it-yourself (DIY) ethics are most strikingly embodied in amateur, independent, and queer and feminist porn. We now focus on these forms to consider the ways such small-batch production simultaneously responds to, perpetuates, and refuses precarity. Web technology has radically changed the landscape of the porn industry, making not only content but also production hyperaccessible. C’lick Me: A Netporn Studies Reader explores these shifts, foregrounding the role of DIY ethics in contemporary Internet pornography. Rejecting any static social meaning of pornography, the anthology’s editors recognize the ways porn producers and users (and where the two meet) modify pornography’s meaning through their interactions with it. Netporn “can contain a critique of commercial work ethics and gender roles,” they suggest.25
Performers take on self-production to create alternatives to available work. Those who do not fit the metrics of physical attractiveness currently in vogue may find better luck producing their own content and creating a niche around their personal brand. Sites such as Suicide Girls and Burning Angel initiated the alt porn genre in the early 2000s to feature tattooed and pierced bodies that, while overwhelmingly white, cisgendered, and thin, did not fit into available porn genres at the time. Queer porn production emerged from the desire to include bodies invisible in mainstream porn, but also had more expressly political aims. Frustrated by the homogeneity of alt porn, Courtney Trouble developed No Faux, now Indie Porn Revolution, the first site to market itself as “queer.” Unable to find work in alt porn as a plus-sized performer, Trouble took to self-production in part to make space to explore her own desires on film. Imagining that others might desire such a space as well, they 26 wanted to create “something that’s truly representative of underground communities and give people a place where they can explore their desires on film.”27 Those who may find a home in queer porn include transgendered and gender queer performers unwilling or unable (read: without surgically altered gender-conforming bodies) to work in mainstream “tranny” porn, plus-size performers who do not conform to the BBW (Big Beautiful Woman) genre’s own strict rules, those with visible disabilities, and some people of color. In her essay on the practice of directing and producing feminist pornography, director and author Tristan Taormino insists that pornographic representations are entirely bound up with production practices. As a feminist pornographer, she works to “capture some level of authenticity, a connection between partners, and sense that everyone’s having a good time. Think of it as organic, fair-trade porn.”28
Amateur porn presents another space that privileges “authentic” self-expression. Trading on the idea of porn as a mode of self-expression, amateur sites and film distributors seek amateur producers who, as Farrell Timlake, the owner of the largest amateur porn distributor, put it, “want to be doing it for the exhibitionist thrill.” Timlake describes Homegrown Video’s scenes as an “authentic” alternative to “paint by numbers porn.”29 We read this too as political. As with other forms of DIY porn, amateur emphasizes the experience of the performer as much as that of the consumer. Again, they may be the same people.
For others, mainstream work is available but requires performing in scenes they feel are degrading or otherwise politically problematic. Roles for black performers are extremely limited, for example, and those available often require workers to perform exaggerated tropes of racialized sexuality. These roles are also poorly compensated, black performers earning a fraction of the rates their white counterparts do.30 Historian Mireille Miller-Young describes self-produced porn as a way for black women performers to assert control over the images they portray. At the same time, self-produced ventures need buyers to survive, so black women performers weave together mimetic performance of expected tropes with portrayals that refuse these roles. For black women porn site producers, she writes, “netporn proffers an intensely politicized space where the line between exploitation and empowerment, pleasure and peril, community and alienation is totally blurred.”31
We find that blurriness compelling. That DIY porn is as much about process as profit contributes to ongoing discussions in media industries scholarship about the dialectics of precarity and creativity. Workers sometimes seek out precarious conditions to enable greater creative expression. With the exception of those self-producers, such as some black women performers who choose DIY in part because it can offer better pay, DIY porn overwhelmingly pays less than mainstream. Those amateur distribution companies that pay at all offer \$500–\$1,000 for a film, to be distributed among all those who participated. Queer production companies pay \$200–\$400 flat rates for a scene, regardless of performers’ gender presentation, race, body type, or the type of sex they perform. Mainstream rates vary widely along these lines, but a standard rate for female performers is \$800– \$1,200.32 Mainstream productions typically employ a host of crew and support staff, whereas DIY productions are drastically pared down. There is no need for scriptwriters, after all, in “unscripted” sex. Films designed to appear more authentic require less postproduction labor.
From a sex-work organizing perspective, DIY porn might be understood to reinforce the idea that sex work is unskilled. More broadly, we are well aware of the widespread management strategy of replacing professional with amateur labor. But DIY, a medium workers initiated precisely in reaction to “professional” pornography, pushes against this critique too. To the extent that focusing on “authentic” sexualities stabilizes them as natural 33 as it frames them as unproduced (that is, unlabored), DIY may serve to stabilize identities as it destabilizes economies. Though DIY production entails greater economic precarity, is that such a bad thing among those for whom stability is personally and creatively toxic? This is, of course, a familiar coupling in the political economy of late capitalism. It puts in relief a set of tensions we cannot and do not wish to smooth over. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/12%3A_Creative_Precarity_in_the_Adult_Film_Industry/12.04%3A_Do-It-Yourself_Ethics_Class_and_Boundary_Work.txt |
1 Chuck Kleinhans, “‘Creative Industries,’ Neoliberal Fantasies, and the Cold, Hard Facts of Global Recession: Some Basic Lessons,” Jump Cut: A Review of Contemporary Media 53 (2011), www.ejumpcut.org/currentissue/kleinhans-creatIndus/text.html.
2 Martha King, “Protecting and Representing Workers in the New Gig Economy: The Case of the Freelancers Union,” in New Labor in New York: Precarious Workers and the Future of the Labor Movement, ed. Ruth Milkman and Ed Ott (Ithaca, NY: Cornell University Press, 2014), 150–170.
3 Tristan Taormino, Constance Penley, Celine Parreñas Shimizu, and Mireille Miller-Young, eds., The Feminist Porn Book (New York: Feminist Press, 2013).
4 Interviewees estimate that, whereas a producer could expect to quadruple his or her investment in the late 1990s, one might double an investment today. Transcripts for all interviews cited in this piece are in Heather Berg’s possession. For more on the broader project of which they are a part, see Heather Berg, “Labouring Porn Studies,” Porn Studies 1.1–2 (2014): 75–79. For a recent update on the economics of porn production, see Alexander Poe, “Seven Directors Focus on the Current Condition of Porn,” XBIZ: The Industry Source (August 20, 2014), www.xbiz.com/news/183848.
5 None of the thirty-three female performers interviewed made a living off film work alone.
6 Gay porn performer Christopher Daniels, for example, explained that the “sole reason” he took on porn performing is because he learned that escorts who are also porn performers get more bookings and can charge on average \$100 more per hour than their nonperformer counterparts. Christopher Daniels, interview by Heather Berg, Los Angeles, April 9, 2014. Female performers in “straight” porn who pursue escorting or erotic dance can command at least double the earnings of their nonperformer counterparts. Tara Holiday, phone interview by Heather Berg, February 22, 2014.
7 Dominic Ace, interview by Heather Berg, Reseda, CA, November 8, 2013.
8 Raylene, interview by Heather Berg, Reseda, CA, October 29, 2013.
9 Venus Lux, phone interview by Heather Berg, June 30, 2014.
10 Christian Mann, “Christian Mann, General Manager, Evil Angel Productions,” in Distribution Revolution: Conversations about the Digital Future of Film and Television, ed. Michael Curtin, Jennifer Holt, and Kevin Sanson (Berkeley: University of California Press, 2014), 121–131.
11 Antiporn feminist writing is rife with constructions of porn workers as passive and un-self-aware. In Catherine MacKinnon’s telling, for example, Linda Lovelace “was pornographed,” Playboy’s consumers masturbate “over the positions taken by the women’s bodies [emphasis mine],” and pornography is “sex forced on real women . . . women’s bodies trussed and maimed and raped and made into things to be hurt and obtained and accessed.” Catharine MacKinnon, Feminism Unmodified: Discourses on Life and Law (Cambridge, MA: Harvard University Press, 1988), 128.
12 See, e.g., Cristina Morini, “The Feminization of Labour in Cognitive Capitalism,” Feminist Review 87.1 (2007): 40–59, doi: http://dx.doi.org/10.1057/palgrave.fr.9400367. Andrew Ross, Nice Work If You Can Get It: Life and Labor in Precarious Times (New York: New York University Press, 2009).
13 Sexual Exploitation and Other Abuse of Children: Record Keeping Requirements, 18 U.S.C.A., 2000.
14 See Constance Penley, “Collision in a Courtroom,” in Images, Ethics, and Technology, ed. Sharrona Pearl (New York: Routledge, 2015).
15 See Mireille Miller-Young, “Putting Hypersexuality to Work: Black Women and Illicit Eroticism in Pornography,” Sexualities 13.2 (April 2010): 219–235, doi:10.1177/1363460709359229.
16 Prince Yashua, interview by Heather Berg, Canoga Park, CA, February 28, 2014.
17 Conner Habib, Twitter post, October 23, 2013.
18 Though performers have repeatedly sought inclusion in mainstream Hollywood’s Screen Actor’s Guild, their presence has remained unwelcome due to both mainstream’s sex negativity and SAG’s policy of only organizing workers on sets where collective bargaining contracts exist. See Gregor Gall, An Agency of Their Own: Sex Worker Union Organizing (Washington: Zero Books, 2012), 28.
19 Legs McNeil, Jennifer Osborne, and Peter Pavia, The Other Hollywood: The Uncensored Oral History of the Porn Film Industry, 2nd ed. (New York: HarperCollins, 2009), 373.
20 Nina Ha®tley’s name is trademarked; this is her preferred spelling.
21 Nina Ha®tley, interview by Heather Berg, Los Angeles, CA, February 17, 2012.
22 Gall, An Agency of Their Own, 31.
23 Nica Noelle, e-mail interview by Heather Berg, October 14, 2013.
24 Gall, An Agency of Their Own, 31.
25 Katrien Jacobs, Marjie Janssen, and Metteo Pasquinelli, “Introduction,” in C’lickme: A Netporn Studies Reader, ed. Katrien Jacobs, Marjie Janssen, and Metteo Pasquinelli (Amsterdam: Institute of Network Cultures, 2007), 1.
26 They is Trouble’s preferred gender-neutral pronoun, as it is for other gender queer and trans people, including Jiz Lee and Papi Coxxx.
27 Courtney Trouble, interview by Heather Berg, Emeryville, CA, March 18, 2014.
28 Tristan Taormino, “Calling the Shots: Feminist Porn in Theory and Practice,” in The Feminist Porn Book: The Politics of Producing Pleasure, ed. Tristan Taormino et al. (New York: Feminist Press, 2013), 261. For a critique of discourses of authenticity in queer and feminist porn, see Heather Berg, “Sex, Work, Queerly: Identity, Authenticity, and Laboured Performance,” in Queer Sex Work, ed. Mary Laing, Katy Pilcher, and Nicola Smith (London: Routledge, 2015).
29 Farrell Timlake, phone interview by Heather Berg, January 23, 2014.
30 Mireille Miller-Young, A Taste for Brown Sugar: Black Women in Porn (Durham, NC: Duke University Press, 2014).
31 Mireille Miller-Young, “Sexy and Smart: Black Women and the Politics of Self-Authorship in Netporn,” in C’lickme, ed. Jacobs, Janssen, and Pasquinelli, 207.
32 Average rates drawn from performer interviews.
33 See Julie Levin Russo, “‘The Real Thing’: Reframing Queer Pornography for Virtual Spaces,” in C’lickme, ed. Jacobs, Janssen, and Pasquinelli. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/12%3A_Creative_Precarity_in_the_Adult_Film_Industry/12.05%3A_Notes.txt |
This chapter makes a case for precarity as a historical state of being for marginalized men and women of color in the entertainment industries. As a preface to underscore what follows, I want to recount two recent experiences that make explicit the larger stakes I’m concerned with here. First, at the originating conference for this collection, a key debate focused on the gendered division of labor and how debates about “progress” often obscure the ongoing marginalization of women from the screen media workforce.
• 13.1: Introduction
Overview of the chapter goal: to make a case for precarity as a historical state of being for marginalized people of color in the entertainment industries. Includes anecdotes pointing out the ways in which women of color are often erased from conversations about labor precarity, which tend to assume all women workers are white and all male workers are racial or ethnic minorities.
• 13.2: Minority Employment – Dismal Data and Industrial Pushback
The underrepresentation of women workers and of both male and female racial minority workers in modern Hollywood productions, both above and below the line. The minimal and even reversed progress since the 1999 protests of lack of minority representation in TV. Samples of the viewpoints held by industry professionals that desire "talent" as the only hiring qualification and do not consider the underlying structures barring people of color from consideration for jobs in many cases.
• 13.3: Casting Directors – Precarious Limbo Gatekeepers
The economic precarity of most casting directors, and how this confines casting directors to sharing the ideological frames of their employers in order to avoid jeopardizing their careers.
• 13.4: Three Strategies of Navigating Today's Hollywood
Three strategies that performers and creative workers of color use to navigate their precarity of labor in today's Hollywood: blind casting, racially ambiguous performance, and universalist discourse regarding works whose cast and creative talent are predominantly of a minority group. The ways in which each strategy reinforces the current inequalities regarding race in Hollywood.
• 13.5: Conclusion
Summing up the various ways in which creative laborers of marginalized groups face and navigate precarity in the entertainment industry: structural insistance that lack of diversity only affects those with a lack of talent; the limited power of casting directors to oppose the status quo, and the strategies that individual workers may adopt to secure work even if they harm collective methods of resistance.
• 13.6: Notes
13: Strategies for Success Navigating Hollywoods Postracial Labor Practices
This chapter makes a case for precarity as a historical state of being for marginalized men and women of color in the entertainment industries. As a preface to underscore what follows, I want to recount two recent experiences that make explicit the larger stakes I’m concerned with here. First, at the originating conference for this collection, a key debate focused on the gendered division of labor and how debates about “progress” often obscure the ongoing marginalization of women from the screen media workforce. Scholars made resoundingly astute points about the ways women continue to suffer under the tyranny of patriarchy in the culture industries and articulated many powerful ways in which we—scholars and practitioners— might engage in the struggle for change and equality. Yet what was missing in this conversation was what is often missing from conversations about identity politics: explicitly marking out the white racial identity of the women we were discussing. I spoke up, named the exnomination, and filled in the gap. Women do not all experience precariousness and contingent labor in the same way. Some women have more access to opportunities than other women simply by virtue of their racial identity, and while all women certainly suffer under patriarchal labor regimes, some suffer less and some suffer more. My intervention in the conversation, then, was to insist on the importance of intersectional cultural analysis when discussing women and labor in the entertainment industries, and insist that any intervention we discuss must be attuned to those differences. Because in a conversation where, to crudely paraphrase Gloria T. Hull, Patricia Bell Scott, and Barbara Smith,1 all women laborers are assumed white and all racial or ethnic minority laborers are assumed male, we can’t begin to address the precarious creativity of women of color without first making them visible in our conceptions of screen media work.
While attending panels about working in the industry at the third annual Austin Television Festival (ATXFest), I encountered another instance when the conversation erased the specific experiences of women of color in the entertainment industries. At the festival, I listened as successful casting directors, staff writers, and showrunners shared their workaday experiences in the field. In a panel on working as an assistant, four women—three white women and one ethnically ambiguous woman—described how they each got their start in the business. Each woman had an internship that then led to permanent employment. They further explained that they garnered the necessary skills for their profession not through college but through their work as assistants or in online extension courses. Lastly, and most relevant to this essay, when asked about accessing entry-level assistant positions, each panelist agreed that leveraging existing relationships and networks was absolutely crucial to employment in the entertainment industries. Indeed, even the panelists’ own hiring practices reinforced this “truism.” They discovered new talent through alumni networks, family members, and friends. Reflecting on this panel conversation, I found precariousness to be an inevitable function of their career choice. Yet I also found that the panelists enjoyed the privilege of stabilizing some of that uncertainty for others by hiring those who reproduce their identities and social relations, and thus offsetting precarity for those who are most like them. I mention this example not only because professional networks are largely racially myopic, but also because the reproduction of identities and social relations vis-à-vis networking and mentorship directly serves a racially unjust status quo. In short, its superficial innocence masks a much more troubling reality: to assume that access to creative work simply depends first on “whom you know” and then on being “the best person for the job” ultimately obscures the power structures that systematically exclude men and women of color from availing themselves of similar opportunities for networking and jobs in the first place.
Both anecdotes reinforce a major crux of the discussion that follows. First, discursive maneuvers that reframe racially myopic professional networks and practices as an ideologically benign function of the creative industries raise the precarious stakes for laborers of color—they effectively neutralize arguments about systemic discrimination and inequality by displacing structural concerns in favor of questions about skills and talent. You’re simply good enough to get the job or you’re not. Likewise, much like my opening anecdote suggests, this discourse risks framing genuine concerns about parity and progress as the product of a contemporary moment marked by extreme precariousness for everybody rather than a function of the socio-historical circumstances of a group of workers whose precariousness has been an ever-present condition of their existence. When meaningful conversations about diversity are outside the confines of common industrial logic (that is, it’s not a problem that exists), the strategies and tactics people of color deploy to gain visibility, secure employment, and maintain careers as creative laborers deserve sustained consideration.
In this chapter, then, I first establish the stark realities of minority employment in the creative industries before outlining how industry professionals abdicate responsibility for structural problems by reframing the issue as one about skills and talent. Such discourse, I argue, is predicated upon the exnomination of its normative ideological basis. In the second section of the chapter, I draw focused attention to how this discourse affects casting for film and television roles. Here I briefly consider how casting directors reproduce normative identities (and thus limited opportunities for actors of color) in their workaday practices. I then conclude by outlining three strategies racial and ethnic minority performers have adopted to contend with their precarious circumstances, and at what cost. Ultimately, I argue that necessary and meaningful political intervention on behalf of a diverse labor force is displaced by persistent notions of “talent” and obfuscated by the simple need to find work in whatever ways possible.
My analysis draws from interviews with media professionals in industry trade journals, conference panels, social media platforms, and my own fieldwork. I borrow John Caldwell’s notion of industrial reflexivity to reframe the workaday experiences and explanations of these “insiders” as a process of self-fashioning and self-theorizing their own identities and interests within existing structures and categories.2 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/13%3A_Strategies_for_Success_Navigating_Hollywoods_Postracial_Labor_Practices/13.01%3A_Introduction.txt |
The lack of a diverse labor force in both above- and below-the-line talent is not simply anecdotal. In April 2014, the Writers Guild of America (WGA) released their latest “Hollywood Writers Report,” the organization’s study on the state of diversity in the film and television industries.3 The report’s findings prompted much debate, and rightfully so, as the data indicated a dismal state of affairs for film and television writers: for instance, minority television writers had increased their share of employment by only 1 percent, and women remained underrepresented by a factor of two to one among television writers.4 Hollywood’s lack of diversity also extended to directing. In 2014, the Directors Guild of America (DGA) diversity report indicated that of the 3,500 episodes analyzed from more than 200 scripted television programs produced in the 2013–2014 season, 69 percent were directed by white males,5 12 percent by white females, 17 percent by minority males, and 2 percent by minority females—a statistic unchanged from the previous year’s study.6 The numbers are no better in acting, where the Screen Actors Guild (SAG) reported that in 2008, white actors dominated television and film roles (70.7 percent). Rounding out the casting data, African Americans represented 14.8 percent of television and film roles, Latinos 6.7 percent, Native Americans 0.30 percent, and unknown/other 4.1 percent.7 Lastly, official statistics for casting directors are more difficult to secure because they are not represented by organized labor to the same degree as other creative professions. They do have a professional society—the Casting Society of America—whose leadership profile follows a pattern similar to employment data collected and distributed by the guilds. Its twenty-six-member leadership team is all white with the exception of one Latino; its gender split is roughly equal.
Such are the data that characterize the premier occupations within Hollywood’s labor force, capturing the degree to which the great majority of feature films and television productions resist multiculturalism. Indeed, despite some small signs of progress (often disproportionately celebrated with self-congratulatory discourses), the film and television industries have yet to initiate any meaningful measures that might correct the staggering lack of diversity in their labor force. In fact, the last time the industry’s exclusionary hiring practices received serious and sustained public criticism was the fall 1999 television season—more than fifteen years ago—when none of the season’s twenty-three new prime-time series featured a single person of color in a leading role. Civil rights organizations and media advocacy groups threatened boycotts and litigation, publicly demanding immediate action from the networks to rectify the troubling lack of minority characters.8 The public shaming and negative news coverage generated some momentum in favor of minority employment both in front of and behind the camera. Networks immediately began casting people of color in supporting roles across a number of series—a liberal “sprinkling” of multiculturalism to quell the controversy. In a structural attempt at change, many networks created in-house diversity positions—executives charged with the futile task of encouraging television showrunners to increase the number of people of color employed on their productions.
Despite such responses, the momentum produced limited success and short-lived interest. Diversity executives are considered all bark and no bite; without the authority to hire or fire, they lack the power to intervene effectively. They furthermore claim that efforts to diversify personnel require fundamental change at every employment rank within a network, and that change remains a far-off reality.9 Furthermore, in the NAACP Hollywood Bureau’s 2008 report, the organization stresses that, despite some gains, the primary objectives it negotiated during the 1999 talks have been largely abandoned by the networks.10 As Vicangelo Bulluck, former executive director of the NAACP’s Hollywood bureau, posited, “The trend definitely seems to be going in the wrong direction.”11 Indeed, nine years after one of the most public industrial shakedowns, employment data retells the same story each year, which further suggests that even if advocacy groups are still pursuing their diversity agendas, the networks have generated strategies to allow them to opt out.
With less than substantial improvement to its exclusionary hiring practices, the television and film industries have nevertheless become emboldened in their apathy about the lack of diversity both in front of and behind the camera. Report after report citing the dearth of employment for creative labor of color has had little effect on how the major Hollywood players choose to conduct their business. Certainly, it is not in their best interest to admit that racial and ethnic diversity is simply a low priority or an unnecessary distraction. In short, no matter how dismal the employment data, diversity just isn’t a problem for many of those individuals in positions with enough power to do something about it. Instead of direct acknowledgment, they employ discursive stopgaps that redirect conversations about employment into discussions of competence and skill—ironically, as I will outline below, concepts that perpetuate familiar ideological beliefs about racial identity.
For example, in response to coverage of the 2013 WGA “Writers Report,” the anonymous commenter “Heartsick” at Deadline Hollywood expressed frustration at the pressure coming from diversity executives as well as talent agents to racially integrate his writing staff. Describing literary agents calling him to suggest writers of color for his staff, Heartsick recalls asking: “What piece of writing have you read that indicates this person would be right for my show, and the answer INVARIABLY is: they haven’t read the person, they’re just calling to con me into hiring someone based on irrelevant, invidious categories that should have no place in the employment of writers.”12 Heartsick is frustrated with agents who allegedly send him ill-prepared writers of color—a phenomenon he doesn’t attribute to white writers also seeking employment—because they interfere with his ability to identify talent based on how well they “fit” with the creative sensibility among his writing staff, a criterion that in his mind transcends racial difference. One can only imagine how many Heartsicks exist in the Hollywood hierarchy. But here’s the critical point: if diversity was an organic industrial practice implemented in staffing hires based simply on postracial notions of fit, talent, and worth, then by extension Hollywood would be a much more hospitable place for ethnic and racial minorities.
Commentators on Deadline are not the only industry-minded folks maintaining that anonymity is the only way to honestly respond to these shameful data-filled reports. One of the more recent trends on Twitter is the emergence of Mystery Hollywood. The “Mysterys,” as they label themselves, are anonymous industry workers/insiders who claim they hold enough clout in the industry that revealing their personal identities would wreak havoc on their professional lives. Mysterys’ racial, ethnic, and gender identities remain unclear unless their Twitter handles or avatars make explicit such differences. The juxtaposition between how Mysterys occupy the socially mediated space as exnominated white and/or male identities and the manner by which they self-fashion personas as successful entertainment industry laborers using the Twitter platform to “tell the truth” anonymously creates some complicated spaces of navigation for a person of color follower. Consider a small section of a Twitter screed by a Mystery account called “DevelopmentHell Exec (DHE)”: “Am I the only one sick of hearing about the plight of women in the film and TV industry? It’s 2014. Just do something awesome, you’re in. Or how bout just making GOOD FILMS? Women-centric, men-centric, alien-centric, muppet-centric, Wall-E-centric. Whatever. Quality > politics.”13 Similar to Heartsick, DHE’s Mystery account allows him to speak his truth about the manner by which diverse employment is discussed in Hollywood. It also allows him 14 to free himself from the focus on employing different kinds of gendered and racial bodies to instead focus on the abstract and apolitical notion of “good work” that cares not about the body from which that work is produced. For DHE, the data suggesting how far white women and men and women of color lag behind white men in all facets of the industry is representative not of a racist structure but of natural selection, sifting out those who create “quality” work from those who are unqualified for the business.
Regardless of how many popular press articles, pie charts, and data graphs consistently demonstrate that marginalized bodies are not allowed opportunities to prove they can produce quality work, the ideological frames perpetuated by the likes of Heartsick and DHE dissociate the structural racism from common industry practices. Creative talents are rewarded with access and opportunity, regardless of the racial or ethnic identity of the worker. The few minority workers who do enjoy some success function as evidence that the best talent does indeed rise to the top. Yet such discursive logic obscures that Hollywood is an industry built around relationships, networking, internships, and apprenticeships—a classed set of practices from which people of color are systemically excluded. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/13%3A_Strategies_for_Success_Navigating_Hollywoods_Postracial_Labor_Practices/13.02%3A_Minority_Employment__Dismal_Data_and_Industrial_Pu.txt |
By the very nature of the career, casting is an overlooked and underresearched component of the filmmaking process. To claim success, the casting director must identify such high-quality talent that his or her part in locating the actors is effaced in favor of an assumption about the process as organic and natural: the actor “just fits” the role. Put simply, good casting happens when no one notices the casting director’s work. Even casting directors themselves elide the skills and expertise required to do their jobs well—in my conversations, they repeatedly claim they “just know it” when they meet the right person for the part. It’s much more likely that casting practices parallel the sort of creativity described by Keith Negus: “Creative practice is not approached as inspirational and radically new, nor as something that everybody does in a kind of everyday creative way. Instead, ongoing cultural production involves working with recognizable codes, conventions and expectations.”15 In other words, casting is not an exclusively intuitive, inspirational, or mystical act. Rather, it is a learned and socialized professional skill. Instead of knowing the right actor when you see her, casting directors understand that the “right” person must adhere to the standardized codes, conventions, and expectations of the industry they service. Casting directors know how to practice their trade because they were trained by other casting directors; identifying the right person is a learned and learnable skill and constitutes the knowledge capital shared among professional networks. A number of current casting directors who spoke with me were trained by the greats—the Marion Doughertys, the Ellen Lewises— and frequently compare this relationship to graduate education.
Casting is a freelance occupation. Casting directors establish careers—and financial sustainability—from job to job. This precariousness further enshrines and reproduces the standard codes and conventions that define the “right” person for the role. Radical or nontraditional casting techniques jeopardize the trust casting directors must maintain with producers and other professionals in their network, especially for casting directors who are young or new to the profession. Indeed, at an ATX Fest panel on casting, Jen Euston, casting director for Orange Is the New Black, said that it was only after she became an established casting director with ongoing and recurring work that she could walk away from a job because she disagreed with the creative vision of the producers or the network bosses. She described this privilege as an outcome of a long and arduous career—a freedom she earned that isn’t available to everyone in her profession. This anecdote underscores how the precarious nature of casting (indeed, much creative work) keeps most professionals tethered to the same ideological frames as those from whom they must gain employment. Learning casting conventions and reinforcing the status quo increase a casting director’s chances for success, but these requirements limit access and opportunity for those individuals who fall outside established codes.
My ongoing research project has been to track mechanisms the film and television industries have promoted as strategies that occasionally allow individuals—like casting directors—to circumvent the racial myopia of professional networks and practices. After spending time observing and interviewing casting directors about the ways they can or cannot incorporate diversity into their workaday practices, I identified colorblind casting as the most prominent contemporary strategy to improve diversity in the postracial era.16 Colorblind casting is the process of excluding racial identities from character descriptions, a tool to increase the number of racial or ethnic actors in front of the camera by ensuring the role is open to (literally) any body (type). While my earlier research investigated how colorblind casting informed the decisions of casting directors and how the practice affected onscreen representations, in what remains of this chapter I turn to the place of agency for actors as they navigate an industry and its gatekeepers, all operating under race- and gender-blind assumptions disconnected from the systemic obstacles designed to exclude specific individuals and representations from common business practices. Racial and ethnic minority actors are forced to play along with this game to secure employment in an industry that is always already characterized by chance, instability, and insecurity. Accordingly, actors of color are doubling down on their precariousness. As they turn to strategies to circumvent these obstacles, we find not minority groups engaged in collective resistance against systematic exclusion but individual minority actors availing themselves of whatever strategies will increase the odds in their favor, ultimately (and unsurprisingly) establishing a set of practices that not only reinforce normative white ideals by exnominating the racialization conventions of the “right fit” for whatever jobs are available but also reproduce subtle tactics of antiblackness through disavowing racial discrimination as an industrial reality. I explicate this dual process in the discussion below by identifying three strategies that help actors of color circumvent their precarious careers. These strategies are blindcasting, ambiguously raced performance, and universal discourse. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/13%3A_Strategies_for_Success_Navigating_Hollywoods_Postracial_Labor_Practices/13.03%3A_Casting_Directors__Precarious_Limbo_Gatekeepers.txt |
Strategy 1: Blind Casting
The physical embodiment of visual difference rather than a qualitatively meaningful representation of difference, blind casting is an illusion of equality and parity in casting. In other words, it functions as a form of diversity you can count rather than a notion of diversity that accounts for the nature of the roles or content. Blind casting thus operates as a way to increase diversity in physical difference without investing in any associated cultural differences. Colorblind casting logic is useful to guilds like the Screen Actors Guild (SAG-AFTRA). Because the guild has no authority over or investment in the form or quality of the employment (that is, the guild does not regulate content), it cares less about the nature of the role than about counting that role as an employment gain for their members. Colorblind casting logic thus offers an easy method to assuage dual concerns: it makes available more job opportunities for the least employed sectors of the guild’s membership and imbues those job opportunities with an air of “respectability” because of the way colorblind logic evacuates any cultural specificity in its operating logic in favor of normative whiteness.
Yet, underpinned by neoliberal race logic, colorblind casting deploys a universal, “we are all the same” rhetoric that only superficially addresses the issues of diversity, employment, and racial representation in television and film. Its resolution relies solely on visible difference. Blind casting thus forces minority actors who desire employment to input cultural differences and output a standardized form of whiteness. Moreover, that this input/output practice has become so commonsensical makes acknowledging its existence a bit of a conundrum. Colorblind logic holds that race is no longer a meaningful barrier to accomplishments, so pointing out continuing injustice as a consequence of racist structures is now, in fact, a racist act. When I make visible the dissonance that occurs when blind casting places an actor of color in a narrative context that, say, isolates him or her from a larger community of color, I am the racist for making a fuss over what Stuart Hall calls “matter out of place.”17 Rather than identifying how this process erases a community’s socio-historical specificity, the logic requires that we celebrate its evacuation of race as an issue at all. Diversity matters now only inasmuch as the networks (and guilds) can count the representation of visible difference.
Yet identifying “matter out of place” is one way to observe the failure of the blind casting strategy. If blind-casting roles for actors of color ultimately normalizes them to the degree that they become culturally illegible, the same effects can reveal the dangers of such an enterprise. When writers do not consider racial difference and history as part of their character’s backstory, they too often succumb to a set of unintended racially troped pitfalls. Consider the blind-cast role of the Black character Bonnie Bennett—a witch—in CW’s The Vampire Diaries (TVD). Bennett, whose original surname in the book series, McCullough, was changed to Bennett as a consequence of the casting decision, is a central character in the televised series. According to the original material, Bonnie McCullough is a fair-skinned redhead. Yet Bonnie Bennett signifies as an African American teenager. While it is a laudable effort on the part of the network and its executives to diversify TVD’s ensemble cast, the failure to adjust the character’s backstory to account for the long history of racialized imagery of Black women and witchcraft opens the series to a number of accidental pitfalls. In the series, Bonnie belongs to a family of witches who historically served as slaves to the lead actors. Continuing in the tradition, Bonnie’s servitude to her white friends results in her sacrificing her life for theirs. Finally, unlike Bonnie’s female counterparts, who are immersed in teen sexuality and coupling—a vital convention of the teen drama—Bonnie is rarely paired with a love interest. Instead, her sole devotion is to those she serves. Collectively, these tropes raise troubling historical associations with Black representation and further perpetuate the sort of symbolic violence against Black female bodies that blind casting’s postracial ethos is intended to counter.
Despite the pitfalls, blind casting remains a viable option for employment when few other promising opportunities exist for actors of color. Moreover, because the parts are written normatively, many actors themselves celebrate the opportunities as “respectable” alternatives to race-specific casting calls, which often perpetuate troubling stereotypes.
Strategy 2: Racially Ambiguous Performance
A second strategy deployed by underrepresented groups to circumvent the industrial barriers to employment is self-fashioning as a racially ambiguous actor. Recalling the earlier statistic from SAG-AFTRA, racially unknown/other actors accounted for 4.1 percent of all roles in 2008. The category “racially unknown/ other” designates actors who did not select a racial or ethnic identity on the surveys SAG-AFTRA sends to identify the racial and ethnic makeup of its membership. According to interviews with guild representatives, members opt not to self-identify because they fear it will relegate them to the limited roles intended for a particular racialized group. Implicit in this trend is the practice of “passing” among those actors who believe they can be cast in roles with a racial identity other than their own.
For instance, the biracial identities of Jessica Szohr from CW’s Gossip Girl and Rashida Jones from NBC’s Parks and Recreation remain “unmarked” in these texts and others in which they appear. Similarly, Troian Bellisario’s racial ethnicity is ambiguous enough to allow her to pass as just one of the (white) girls in ABC Family’s Pretty Little Liars—even though reading her body suggests there is something “not quite white” about her character. Beyond indeterminate racial identities, racially or ethnically ambiguous performers find themselves cast as multiple races and ethnicities. Blair Redford’s ambiguous look allowed him to be Latino for ABC Family’s Switched at Birth and American Indian for ABC Family’s The Lying Game. While the logic behind this strategy might increase an actor’s employment opportunities by expanding the number of types through which his or her look is interpreted, it also privileges (oftentimes racist) assumptions about the look of a given racial group.
Racially ambiguous performers also amplify colorblindness’s insidious power. As a strategy, it not only makes race something that is “unseen” but detaches racialized bodies from their socio-historical contexts. In other words, the actors function as empty signifiers in that their bodies can be read by audiences in multiple ways, and they can be placed in infinite settings without being tethered to a reality rooted in the socio-historical specifics of their racial and cultural experiences. Furthermore, racial ambiguity allows the network to claim diversity without engaging with the concept beyond superficial (physical) differences.
Strategy 3: Universal Discourse
The final strategy I want to discuss is one that, unlike the first two, is applicable to a variety of laborers in the film and television industry. It concerns distributing and marketing film with predominately Black casts that are also written, directed, and/or produced by Black creative talent. From films like The Best Man Holiday (2013) to Think Like a Man (2012), and Think Like a Man Too (2014) to the About Last Night (2014) remake, publicity and advertising largely frame these films within a universalist discourse—one designed to assure (white) mainstream audiences that the experiences onscreen are both “human” and “relatable” even though the characters may not look like them and elements remain that are, in fact, quite culturally specific.18 Consider the promotional strategy for Think Like a Man. According to a 2012 Vulture article, while Black producer Will Packer devoted a large portion of his marketing and advertising budget to flying the cast to events with large numbers of African Americans in the audience, he deployed an alternative strategy to draw white audiences. Here he relied heavily on “crossover” comedian Steve Harvey—who hosts the daytime game show Family Feud and wrote the film’s source material—as the movie’s messenger. “Packer has deployed Steve Harvey . . . to sell the ‘everybody’s welcome!’ message to the general public, sending him out to tub-thump . . . on CBS’s This Morning and ABC’s The View, shows that Packer explains, ‘don’t necessarily over-index with African Americans.’”19 Moreover, to target white women, Packer and his team stressed the romantic comedy conventions of the film via a television campaign that, according to one former studio marketing head, looked “like classic Romantic Comedy 101. In fact, it looks like a Nancy Meyers movie, with black people. Which is fine. . . . All it has to be is funny, and make it clear that the concept has no race.”20
Yet at what cost comes this universal rhetoric? Extra labor taken on by the actors and producers during these press junkets and promotional events to sell the film as fitting for “all” is expected from marginalized bodies if they desire to reach a mainstream audience. Films with predominately white casts are not expected to sell their films (domestically at least) as universal and relatable because they always already operate within the normative and authentic standards by which we judge the human experience. Similar to blind casting, the burden falls on the person of color to perform his or her “sameness” as a mechanism to ensure that the preferred demographic is not alienated from the production. As Packer asserts, “There is a process to get those audiences. It starts with making a film like this, which is broad, smart, and one where there’s no cultural or ethnic specificity that would not be relatable to mainstream Americans.”21 Black cast films, then, are not a niche production; they are the benign reflection of a large, multicultured world that poses no threat to liberal sensibilities and consumption practices.
Universal discourse underscores the historical precariousness of minorities in the creative industries whose labor always existed under this double bind structure that equates success with being both similar to and different from the normative order. Such a double bind recalls how Clyde Taylor defines the mode of Black film production in early cinema as one of “unequal development,” that is, a phenomenon that exists where there is “an exploitative/dependent relationship that ultimately results in a more powerful society drawing from the less powerful selected goods and resources without regard for what the loss of those resources will mean to the exploited.”22 According to similar logic, universal discourse insists that for creative laborers of color to participate in the film and television industries, they must embrace a rhetoric of sameness that not only elides their unequal professional footing but also encourages them to lose any sense of socio-historical specificity. Yet what must it mean to be a minority worker who, to find employment, must not only cross over to mainstream filmmaking but also disavow elements of his or her own racial identity to remain gainfully employed? Unequal development epitomizes precarity. It considers the minority worker—whose skills are utilized and borrowed, or more specifically, appropriated, for a variety of purposes—always operating on a conditional and probationary basis. Once the current diversity zeitgeist ends, so does the work.
I would extend the universal discourse to branding and would draw attention to how showrunner Shonda Rhimes, most recently described as a “revolutionary,”23 tells the story of how when she looked at an invitation for an award ceremony at which she was to be honored and saw that she was described as the most powerful Black female showrunner, she scratched out the modifiers Black and female.24 While I understand her desire to not be limited or constrained by those modifiers if she, in fact, is the most powerful showrunner of any description, her rationale that white men do not have to name themselves is based in structural power and an inherent specialness that allows them to be ex-nominated. They don’t have to be named because it’s common sense. Thus, though Rhimes’s refusal to take the modifier may for her be an insistence to transcend, the rhetoric ultimately reinforces the very whiteness implicit in industry. Further, while her strategy attempts to upend the “unequal development” of being considered successful only in relation to other Black female television producers—of which there is one: Rhimes—as opposed to being placed in contention with the predominately white male showrunners, by shrugging off those identity modifiers Rhimes reinscribes herself in a universalist posture. A posture that ironically makes racial difference a type of pathology one needs to be cured of, thus reinvigorating the very tenets of unequal development she hopes to quash. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/13%3A_Strategies_for_Success_Navigating_Hollywoods_Postracial_Labor_Practices/13.04%3A_Three_Strategies_of_Navigating_Today%27s_Hollywood.txt |
Throughout this essay I tried to illustrate different ways that various creative laborers at various levels of access navigate precarity—from the anonymous (white) gatekeepers who stress that precarity is not a diversity issue but only an issue for those who lack the necessary skills; to the casting directors who find themselves stuck in a precarious limbo with insufficient power to break the status quo despite unparalleled access to diverse pools of talent; to actors of color whose precarious existence means they must strategically plan to circumvent the system even if those strategies benefit individuals at the expense of collective forms of resistance. Ultimately, the uncertainty of employment forces all these groups into their own strategies and tactics of negotiation. And while employment and maintaining one’s livelihood is the point, the danger of such precarious livelihoods is that oftentimes survival takes precedence over all other factors—including the need for cultural resonance and specificity. That precarity results in the maintenance of a white heteronormative status quo is not shocking, but demands that future research consider the historical and discursive ways creative men and women laborers of color have survived in spite of the uncertainty as a guidepost for understanding issues of labor at all levels.
13.06: Notes
1 Gloria T. Hull, Patricia Bell Scott, and Barbara Smith, eds., But Some of Us Are Brave: All the Women Are White, All the Blacks Are Men: Black Women’s Studies (New York: Feminist Press, 1993).
2 John Thornton Caldwell, Production Culture: Industrial Reflexivity and Critical Practice in Film and Television (Durham, NC: Duke University Press, 2008).
3 Darnell Hunt, “The 2014 Hollywood Writers Report: Turning Missed Opportunities into Realized Ones” (July 2014), www.wga.org/uploadedFiles/who_we_are/HWR14.pdf.
4 Ibid.
5 It should be noted that white males directing decreased by 3 percent in contrast to the year before, while minority males increased by 3 percent, although the study cites that the increase can be attributed to a higher number of episodes directed by Tyler Perry, who solely directed episodes for his three television series.
6 Directors Guild of America, “Employers Make No Improvement in Diversity Hiring in Episodic Television: DGA Report,” September 17, 2014, www.dga.org/News/PressReleases/2014/140917-Episodic-Director-Diversity-Report.aspx.
7 Screen Actors Guild, “2007 and 2008 Casting Data Reports,” www.sagaftra.org/files/sag/documents/2007–2008_CastingDataReports.pdf.
8 Katie M. Kleinman, “Minorities in Prime-Time Television,” November 15, 1999, www.katiekleinman.com/portfolio/minoritiesmedia.php.
9 Jennifer Armstrong and Margeaux Watson, “Diversity: Why Is TV So White,” Entertainment Weekly, June 13, 2008, www.ew.com/ew/article/0,,20206185,00.html.
10 NAACP, “Out of Focus—Out of Sync Take 4: A Report on the Television Industry,” December 2008, http://action.naacp.org/page/-/NAACP...OS%20Take4.pdf.
11 Armstrong and Watson, “Diversity.”
12 Heartsick, comment on the Deadline Team, April 14, 2014, 10:53 a.m., “WGA Diversity Report: Women Writers See Gains in TV, Slide on Film Side,” Deadline Hollywood, April 14, 2014, http://deadline.com/2014/04/writers-...riters-714478/.
13 DevelopmentHell Exec, Twitter post, May 30, 2014, 3:46 p.m., https://twitter.com/DevHellExec.
14 Based on the pattern of Mystery account titles, the assumption is that unless modifiers are in place that suggest otherwise, the owners of said accounts are white males.
15 Keith Negus, “The Production of Culture,” in Production of Culture/Cultures of Production (Thousand Oaks, CA: Sage, 1997), 362–363.
16 Kristen Warner, The Cultural Politics of Colorblind TV Casting (New York: Routledge, 2015).
17 Stuart Hall, “The Spectacle of the Other,” in Representation: Cultural Representations and Signifiying Practices, ed. Stuart Hall (Thousand Oaks, CA: Sage, 2003), 236.
18 For evidence of this trend, look at the press junkets for the films and note how often the actors stress that these are movies “for everybody.”
19 Claude Brodesser-Akner, “Think Like a Man Is the Best-Testing Film in Hollywood—but Can It Win Over Black Men and White Audiences,” Vulture, April 10, 2012, www.vulture.com/2012/04/think-like-a-man-kevin-hart-will-packer.html.
20 Ibid.
21 Ibid.
22 Clyde R. Taylor, “Black Silence and the Politics of Representation,” in African-American Filmmaking and Race Cinema of the Silent Era: Oscar Micheaux and His Circle, ed. Pearl Bowser, Jane Gaines, and Charles Musser (Bloomington: Indiana University Press, 2001), 3.
23 Mark Harris, “The Shonda Rhimes Revolution: Finishing What The Sopranos Started,” Grantland, October 16, 2014, http://grantland.com/hollywood-prosp...s-scandal-abc/.
24 Lacey Rose, “Shonda Rhimes Opens Up about ‘Angry Black Woman’ Flap, Messy Grey’s Anatomy Chapter and the Scandal Impact,” Hollywood Reporter, October 8, 2014, www.hollywoodreporter.com/news/shonda-rhimes-opens-up-angry-738715?utm_source=twitter. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/13%3A_Strategies_for_Success_Navigating_Hollywoods_Postracial_Labor_Practices/13.05%3A_Conclusion.txt |
In this chapter, we pay full attention to the structural conditions and human cost of precarious labor in a particular local instance of the games industry. But at the same time, we attempt to shift the debate on precarity from the existential (the creative individual attracted to industries promising autonomy and meaningful work and finding only casualization, no work/life balance, and poor management) and the totalizing (all work under regimes of neoliberal hypercapitalism is increasingly characterized by precarity; indeed a whole new class—the precariat¹—is posited as emerging) to a focus on analysis for actionable reform.
• 14.1: Introduction
Introduction to the goal of the chapter: analyzing labor precarity in the Australian games industry, in terms of structural conditions and human cost, as a symptom of the emergence of a new global class of worker, the precariat. A brief history of the Australian games industry, focusing on the post-global financial crisis era.
• 14.2: They Still Make Games
Surveying the identities and motivations of several Australian game development studios, with a focus on indie developers. Analyzing how the shared pride in having survived the financial crisis has led developers to desire making original IPs, and the various economic strategies necessary to achieve this goal compared to the lesser precarity of fee-for-service work.
• 14.3: Precariousness as a Function of Policy and Industry Cultures
Ways in which precarity in the Australian games industry arises from government policy, professional advocacy, and poor management within the industry.
• 14.4: Conclusion
Strategies for reform of the games industry, so that innovation and creativity can be a part of stable jobs.
• 14.5: Notes
14: Games Production in Australia Adapting to Precariousness
In this chapter, we pay full attention to the structural conditions and human cost of precarious labor in a particular local instance of the games industry. But at the same time, we attempt to shift the debate on precarity from the existential (the creative individual attracted to industries promising autonomy and meaningful work and finding only casualization, no work/life balance, and poor management) and the totalizing (all work under regimes of neoliberal hypercapitalism is increasingly characterized by precarity; indeed a whole new class—the precariat 1—is posited as emerging) to a focus on analysis for actionable reform.
Significant “creative destruction”2 through the global financial crisis (GFC) led to games industry restructuring and consolidation, including withdrawal of major publisher investment in many dispersed regional hubs of games production. More fundamentally, major platform shifts and new business models started before the global downturn and continue through this contemporary period of slowdown in the world economy. There has been major consolidation at the console production end of the games industry, with more expensive blockbuster or AAA titles, a hollowing out of the midrange games market, and rapid growth and proliferation of casual gaming and mobile applications with unprecedentedly lower production costs and barriers to entry.
What has happened to one such regional hub, the Australian games industry, spatially remote from the centers of publisher power and hubs of creative ferment?
A recent “perfect storm” of factors has arisen to change the face of the Australian games industry. The industry had grown on the model of work for hire producing “catalog fillers” for the major publishers; very little original IP was produced. And while very few AAA titles were made in Australia, games companies had a reputation for quality. However, the business proposition was buttressed by more than a decade of favorable exchange rates, which (literally) underwrote international investment. The industry by 2007 was structured around approximately forty-five midsize small businesses.3 Notable companies included Krome, Pandemic, THQ StudioOz, Creative Assembly, Torus, and 2K.
The global financial crisis saw higher-end production scaled back, a withdrawal by the major publishers from spatially distended supply chains, and a new preference for formally affiliated production companies. At the beginning of 2007, the Australian dollar was 75 cents on the U.S. dollar. During the GFC, the Australian dollar became a “currency haven,” such that by the start of 2012 it was worth US\$1.02, gutting the industry of its price advantage. Of even greater structural consequence for the industry was the simultaneous explosion of apps-based mobile casual games play based on the smartphone platform and later the tablet.
Official statistics tell a stark story of destruction of value. Of the 1,431 reported employees in 2007, only 581 remained by mid-2012, and reported game development income had dropped from A\$116.9 million to just A\$44.4 million.4 The industry’s spatial pattern in 2007 evidenced a significant presence in Queensland and Victoria, with additional studios in New South Wales, the Australian Capital Territory, and South Australia. By 2012, the majority of the bigger studios had closed, and the industry had retreated to be concentrated in Victoria. Those whose doors had closed or who had radically downsized included Krome, Pandemic, THQ StudioOz, BlueTongue, Team Bondi, SEGA Creative Assembly, and Tantalus Media Brisbane. The major studios remaining included Halfbrick (Brisbane), 2K Australia (Canberra), and in Melbourne, Big Ant, Torus Games, Tantalus, and Wicked Witch. According to the Games Developers Association of Australia (GDAA), the main advocacy and professional association for the industry, somewhere between 60 and 70 percent of industry workers had either moved to another industry (many skills, preeminently programming skills, are very transferable) or had left Australia for more resilient industry locations or those better supported by government policy and programs.5
In 2014, the GDAA characterized the industry as composed of two hundred formally registered businesses, of which 92 percent are considered to be independents.6 It defines independent as a typically small-scale enterprise that concentrates exclusively on original IP and self-publishes on the new digital platforms (Apple App Store, Android, Steam). It estimates about eight hundred workers now in the industry. This is a recent history of an industry much reduced in turnover and traditional employment, but which has transformed its revenue base from 80 percent work for hire to 75 percent original IP—an almost complete reversal in the balance between business models.7
But, invoking Joseph Schumpeter, how “creative” has this destruction been? A rigorous critical organizational-studies analysis of the Australian industry advances the argument that severe power differentials between publisher and producer/developer have persisted across this momentous industry restructure and continue to compromise local agency in global supply chains.8 An equally rigorous media-studies argument anatomizing poor labor conditions in the industry globally is nevertheless clear that “the most plentiful and well-paying jobs in the video game industry continue to be those provided by major video game publishers either directly or indirectly.”9 Neither view offers much comfort for the idea that this destruction could be in any way “creative.”
These perspectives, however, contrast with the self-understanding of many of those games workers (whom we have interviewed for the research that supports this chapter) who survived the shakeout or are sufficiently new to the industry to know no other conditions. Culturally and industrially, original IP—and the conditions under which it can be prioritized—tends to be championed by these developers against fee-for-service and as a normative aspiration. Industrially, a dominant narrative in the industry has been the desire to move from fee-for-service (where the company is a price taker and doesn’t control its own destiny) to original IP. Culturally, this aspiration also speaks to many developers’ creative impulse and is actually enshrined in the advocacy and the definition of indie established by the representative body, the GDAA. It is reinforced by normative criteria built into state policy and program funding support.
Given the degree to which higher-end fee-for-service business has dried up, while essentially self-publication on major digital distribution platforms (Apple’s App Store, the Google Play Store, Steam, and so on) has grown exponentially, necessity has become a virtue. Conditions have crafted an industry that is much reduced in terms of turnover and traditional employment but now operates within a disintermediated value chain that radically forces the pace of innovation. Despite much commentary that treats Apple, for example, as basically yet another global corporation “taking their (un)fair share of financial profits,”10 near-global dissemination via the digital platforms on a 30/70 split of income derived represents an ostensibly better deal than the power asymmetries enshrined in dealing with the major publishers.11
Australian companies, in particular Halfbrick after its huge success with Fruit Ninja (2010), made hay while the sun shone in the early days of apps-driven games and became a sort of template for national ruminations on how to succeed in the new environment.12 It is distinctly harder now to capture attention: massively lower barriers to entry create conditions in which it is estimated that more than 1.3 million apps are now available on the App Store with duplication across the platforms, of which around 20 percent are games.13 Mobile games production is markedly less driven by the crunch associated with games development under the dominant business model of fee-for-service work, in which development schedules were driven by milestones at the behest of large international publishers. This has led, Antony Reed suggests, to a situation where the industry has seen much less attrition in last few years. Furthermore, there is arguably a great deal more innovation activity in original IP. Indeed, there is runaway innovation,14 with the rapid shift from games as a product to games as a service driving the mobile apps purchase price points to zero, accompanied by the proliferation of in-app purchasing. And these rapid shifts have in turn been challenged by a return by some to premium mobile app pricing as well as premium pricing for games released through Steam. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/14%3A_Games_Production_in_Australia__Adapting_to_Precariousness/14.01%3A_Introduction.txt |
It is to these identities and motivations—the scripts games developers have written for themselves to adapt to the new conditions—and their relation to business models and production cultures that we now turn.15
Predominantly, we encountered a sense of pride in the fact that these developers were still making games. They had found a way to survive the changes upending the Australian industry. Many emphasized that they were now doing this more on their terms and that the shift from fee-for-service to original IP meant they enjoyed greater creative control and autonomy. In describing this sense of creative control, none of the developers were remotely Panglossian about the precariousness they and the workers around them routinely face. Many recounted the pain of downsizing and seeing fellow workers lose their jobs, with many needing to leave the country for work in the United States, Britain, and Canada. Others told us about their companies coming repeatedly to the brink of closure and yet finding a way to keep the doors open. Nevertheless, this assertion of creative control came through in a comment by Dean Ferguson at 5Lives (a Brisbane-based group of five developers making the Kickstarter-funded game Satellite Reign, 2015): “It’s probably the first time in a number of years where I’ve felt like I’m crafting a game and not simply part of a cog. Before ‘the crash,’ I worked with and formed great relationships with many very creative people, with really well-meaning people, including publishers, but it often came down to pure economics much of the time. It could be a real struggle to just craft something, and while it sounds tacky, a lot of us do this largely for the love of crafting.”16
Morgan Jaffit, director of Brisbane’s Defiant Development, put the case even more strongly: “Australia has a history of terrible work-for-hire projects and shitty lowest-bidder poor-quality games. It not only erodes your studio but I think it kills your soul too.”17 Trent Kusters, founder and director of Melbourne-based League of Geeks, also noted the importance of “having an impact on the medium, and the progression of the medium, and where that is happening. That you as a creator, you’re not just pumping out some crappy title that’s, you know, just going to turn a quick buck. If you want to make things that matter, you need to have a cultural understanding. You need to be involved in that, the discussion of the cultural zeitgeist of game development and games as a medium, and you can see a clear pattern between the people that are right now developing great games and the networks that they move within.”18 The values that these leaders of what has emerged as a profoundly different Australian games sector associate with “indie” game development need to be carefully interrogated; they are in no way opposed to commercial interests or business sustainability. Creative adaptation, experimentation, and opportunity have arisen under conditions of profound uncertainty and precariousness.
Many developers clearly feel there is a great deal more innovation potential—and identity reinforcement—in original IP. On the other hand, viewed from an industry-wide perspective, some companies continue to pursue fee-for-service work to offset the risk associated with free-to-play—and indeed with making original IP games generally. For some developers, work-for-hire remains important to the sustainability of their studios. Therefore, we now posit a typology of approaches to funding and releasing games in the overall ecology of the sector and then briefly profile companies that exemplify this range of approaches.
Along with licensed IP, there are five distinct variations on the exploitation of original IP: subscription, premium payment, free-to-play with in-game monetization, advertising supported, and pay-to-play. The subscription model is consistent with the games-as-service approach, where at the beginning of each period, usually monthly, the player pays to stay engaged with the game. This is typical of games like World of Warcraft, which continues to have a significant player base ten years after launch. The premium model is very much traditional in the games industry and is consistent with the games-as-product approach. The consumer pays for a complete experience with a one-off payment. Such a model is typified in AAA titles, such as the Call of Duty series (2003–) or titles like Minecraft (2011), but a quite different level of premium pricing also applies to variations on free-to-play.
Free-to-play can be adopted in a variety of forms, placing this category in both games-as-product, where you pay to unlock additional content but expenditure is capped—for example, Puzzle Retreat (2013)—or in a games-as-service form, where there is no cap on monetary expense (for example, Clash of Clans [2012] or Kixeye’s VEGA Conflict [2013]). The advertising-supported revenue model leverages advertising as the primary source of income by inserting advertising at regular or semiregular intervals; it is most typical of browser-based flash games. The final model is the pay-to-play monetization model. Typified by the original arcade machines, each play of the game requires an input of credit for the player to progress. The developers that we discuss in this chapter have tended to focus on emerging opportunities of free-to-play and premium payment approaches, especially in the context of the shift toward games-as-service models.
The funding for games development takes a variety of forms, depending on the availability and the scale of the project. Briefly, these sources include government funding, in the form of loans or grants with funds available not just for development costs but also for travel or to engage marketing expertise; crowdfunding through platforms like Kickstarter; the traditional publisher model, where the developer is engaged to produce content at a set fee and with set milestones for delivery, essentially work-for-hire; variations on the work-for-hire approach that may involve undertaking projects such as game installations, serious games, or nongaming apps; and securing donations, where donations are received against the development costs.
As an index of the stakes involved in this challenging innovation space, consider the case of Halfbrick, the company that bet the farm on original IP on mobile game platforms. Halfbrick has continued this approach with recent releases like Fish out of Water (2013), Collosatron (2013), and Bears versus Art (2014). While the company’s recent releases experiment with various approaches to free-to-play and in-app monetization by drawing on analytics and metrics to inform their design, development process, and decisions, they have not as yet managed to repeat the stellar commercial success enjoyed by Fruit Ninja (2010) or the lesser but still substantial success of Jet Pack Joyride (2011). Halfbrick had led the industry in adapting to the shift from work-for-hire to original IP titles for mobile devices.19 In front of the pack when mobile games were all paid for upfront, success has so far eluded the company after the market shift to free-to-play and games-as-service.
Wicked Witch, which was started, like Halfbrick, in the late 1990s, is different. It mixes work-for-hire with original IP development. During the industry decline, Wicked Witch radically downsized, almost closing. However, by continuing fee-for-service work for domestic sports titles that were not subject to the exchange rate crisis, together with developing original IP games for mobile devices, Wicked Witch has managed to rebuild a fifty-person studio. This makes them one of the largest companies in the new ecology. Successful titles include Catapult King (2012), released for both Android and iOS devices. Wicked Witch has also released Whac-a-Mole (2014) for Mattel, a conversion of the classic arcade game for Apple devices, and Jet Run: City Defender (2014), a free-to-play game with in-app monetization, for iOS and Google Play. Wicked Witch CEO Daniel Visser observed that in his opinion the free-to-play model was becoming “a race to the bottom that is so intense that we’re going to end up paying people to play our games.”20 Free-to-play is becoming such a crowded market, with such great potential for destruction of value, that developers need to explore other models, including premium payment titles for mobile platforms.
Melbourne-based League of Geeks exemplifies such an approach. League of Geeks is not banking on chasing the mobile free-to-play market. Since 2011, this group of developers, including designers, programmers, and artists, have come together to make Armello (2015), a game they describe as “a swashbuckling adventure that combines RPG elements with the strategic play of card and board games, creating a personal, story fuelled experience.”21 Structured as a core creative team of four directors and a loose coalition of programmers and artists who contribute collaboratively to the project, they are located in the Arcade in innercity Melbourne, a game development space shared with other companies that has the look and feel of a creative start-up and is supported by the Victorian government. League of Geeks garnered attention in 2014, when they raised \$305,000 from Kickstarter to keep the Armello project progressing. Director Trent Kusters describes League of Geeks as a game development collective rather than a formal studio.22 Kusters left the Australian industry in 2011–2012 to seek work overseas. He said that through this period he felt “disenfranchised” by the big studio developer culture. He worried that in such an environment he could end up being “a little cog in a big wheel, tweaking combat timings on some NPC for, you know, some multimillion dollar game.” In contrast to Wicked Witch, Kusters emphasized the importance of developing original IP, saying that fee-for-service work was “like quicksand.”23 Unlike Australian developers who retain some fee-for-service work to balance the risks associated with an original-IP-only approach, Kusters believes relying too much on fee-for-service can compromise a studio’s ability and commitment to create original IP.
The game development engine Unity was becoming widely available by the late 2000s, offering low-cost but high-quality technology for making games. Combined with digital distribution opportunities through the App Store and Steam, this radically changed the possibilities for making and releasing games. Kusters noted the emergence of online indie developer communities using productivity tools to manage distributed collaborations among teams around the world. Armello relied on a distributed network of developers that Kusters sees as exemplifying his vision of a game developer collective. Some developers were engaged through a points-based system in which they would receive a cut of the profit from Armello based on their contribution to the project. Others worked on the project through an arrangement that combined points with contracted and paid employment. Armello also raised funds to continue development through Kickstarter, but both national and state government funding was critical to Armello’s viability. League of Geeks plans to release Armello as a premium title rather than pursuing a free-to-play approach with in-game monetization. This model of indie development, Kusters says, is about “adapting to the current climate. . . . The market completely shifts underneath us all the time. We just have to be agile. We just have to do what we need to do, and that’s basically how we came up with the model . . . that doesn’t require us to have cash.”24 This is a business model that marshals government backing, deferred points-based payment systems, and crowdfunding to underwrite passionately conceived games that depend on innovation, reputation, and point-of-difference from most standard mobile games product.
Sharing office space in the Arcade complex with League of Geeks, Voxel Agents (a small studio of five or six employees) pursues the opportunities of original IP and free-to-play game releases for mobile devices with successful titles like the Train Conductor (2009) series and Puzzle Retreat (2013). Voxel Agents is tackling the shift toward games-as-service, which requires regular content updates and the use of metrics and data analytics to respond to player behavior. Voxel’s Simon Joslin noted the value of working in a collaborative space such as the Arcade, which permits both formal and informal sharing of knowledge and experience about the rapidly changing video games market.25 This includes access to small specialist firms, such as Surprise Attack, which offers consulting services to developers as they seek to develop effective business models that embrace the demands of games-as-service, particularly expertise in game monetization and effective use of data analytics. Both state and national government support for business development was critical as they experimented with various approaches to the games-as-service model. Joslin noted that while the shift to original IP provided greater creative control, changing business models to games-as-service, especially free-to-play games like Candy Crush Saga (2012), may compromise the craft of making quality game experiences. He worried—as did other developers—that many of the monetization strategies associated with in-app game purchases relied on mechanics that may be addictive. He discussed the ethical and craft implications of free-to-play: “It’s a complex question, a gray area. . . . There are points where I’ve played games and I feel that’s the wrong way to do it. . . . I wouldn’t feel comfortable doing that to my players.”26 He talked about the steep challenge of adapting existing game design knowledge and skills to create engaging and compelling free-to-play titles while making effective use of metrics and analytics.
So far, with the exception of Wicked Witch, we have emphasized GDAA-defined indies in this survey of the precarious but widening range of business models and company and developer identities. But some U.S. company presence remains in the country. Kixeye, situated in Brisbane with a staff of some fifty to sixty, manages the distance from centers of developer culture by being a wholly owned subsidiary of San Francisco–based Kixeye, a developer of online browser-based strategy and combat games, such as Battle Pirates (2011), War Commander (2011), and VEGA Conflict (2013). The studio director, George Fidler, a veteran of the industry, emphasized the fundamental challenge of shifting from a work-for-hire and games-as-product model to a games-as-service market environment.27 He suggested that while the fundamental skills of programming, art, and good design were still crucial, new skill sets and expertise in digital retail now needed to be integrated into the production process and studio culture. Australian development studios still lacked the skills crucial to successfully making the shift to games-as-service. Fidler commented that the work-for-hire origins of many Australian studios and developers meant that they perhaps had not gained the market discipline of focusing on a core competency or a core market. Speaking of the games-as-service shift, Fidler concluded that for Australian developers, “it’d been tough to create those kinds of games early in the cycle, because the expertise simply wasn’t there.” By expertise, he clarified the product manager and producer skills required to combine and balance retention, monetization, and engagement: “We’ve got thousands of game designers in Australia. No problem at all, but we have very, very few experienced product managers, and that’s meant most of the attempts have fizzled out, because if you think of the build-measure-learn cycle, we built, we didn’t quite know what we were measuring, and we learned nothing.”28 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/14%3A_Games_Production_in_Australia__Adapting_to_Precariousness/14.02%3A_They_Still_Make_Games.txt |
In the overwhelmingly nonunionized games industry, advocacy for the sector is largely conducted by professional associations, and support is offered through state policies and programs.29 This section considers the extent to which precariousness is a product of policy, advocacy, and industry self-governance.
A key feature of the games industry is that it is poorly understood by the political class. This is despite its size and growth rates globally dwarfing anything remotely comparable, and is an outstanding example of creative content and use driving technological innovation and take up, not the other way round, as is usually constructed in innovation policy and business strategy. It tends to fall between the three “stools” of cultural policy, industry and innovation policy, while its main interface with the political class and the wider populace is around social and educational policy concerns (violence, game-playing addiction, claims and counterclaims about educational benefits). Inconsistent or nonexistent policy support, particularly compared to other cultural industries, such as film and television, contributes to precariousness. Such policy inconsistency across different countries, as well as policy entrepreneurship or arbitrage between countries in bidding for the services of this high-skill component of the “creative class,” contributes to the hypermobility of games creatives.
In Australia, federal policy and programs supporting the industry had been piecemeal,30 seeking to fit games into the established cultural template developed over decades for the arts, film, and television. They required developers to articulate game proposals as forms of storytelling to measure the cultural significance of the game. The long march toward a realistic balance between cultural and industry policy for the creative sector was accelerated by the industry transformations of the last five to seven years. Government accepted that very little original IP was being created; that Australian developers were locked into a fee-for-service system; that the country was no longer attractive to licensed IP; and that oversees competitive incentives were “luring” talent away from the country. A significant A\$20 million package was developed, the Australian Interactive Games Fund, whose objectives were to promote industry growth and sustainability, support the development of new intellectual property, encourage skills retention and renewal, and maximize the creative opportunities of fast broadband.31 With a change of government, however, the initiative was cancelled with only half the money spent.
At a state or provincial level, the policy rationale for support has been equally uneven, with an equal or perhaps even greater impact on precariousness for the labor force. The state of Victoria has been most consistent in its approach to games, which are recognized as a core component of the state’s industrial and employment base in its information and communication technology sector. Effective advocacy for the sector forestalled a cost-cutting attempt to close down support in 2012. Funding and programs in support of the sector are administered through a mainstream screen agency. The approach in Queensland was exclusively industrial and remained positive while the industry was generating jobs as midsize small businesses proliferated in the pre-GFC period. The collapse of several of the larger companies effectively eliminated games from a standard industry development policy logic as pursued within a department of state development and saw the policy focus narrow to a minor part of the screen agency’s remit. Government did little to arrest the collapse of the industry in the state, and has done little since. New South Wales, the most populous state and the one with the largest slice of GDP, had rarely focused policy and program attention on games, leading to the irregular doughnut shape of the industry’s geography.32 The effect of such policy variability is clear—Victoria has seen strong 15–20 percent growth in each of the last few years, while Queensland has not grown strongly out of the downturn. The mobility and associated uncertainties faced by game workers are often forced on them by the volatility of an industry whose profile with government is equally volatile.
Policy fluctuation and failure contributes to precariousness; so does the industry’s reputation for poor management. Some of its notoriously poor working conditions can be attributed to the immaturity of the industry and the need for self-governance reform. The industry’s still overwhelmingly male-dominated production base needs to change if it is to attract the best talent, improve balance and sustainability, and capture value in a rapidly evolving consumption environment. Women and girls now account for 48 percent of all gamers. The high skew toward men and boys—more than 78 percent in the console core demographic—underlines that women are in the majority in the more casual gaming areas of the market.33 GDAA survey data for 2014 suggests that, of the approximately eight hundred people now working in the industry, some 26 percent are women, and most of these are programmers and artists. This is beginning to align with the IGDA’s most recent survey results, which report 22 percent women employees globally in September 2014.34
Management deficit is by no means confined to gender. Casey O’Donnell’s loving but forensic description of the “secret world of videogame creators” does not spare the industry.35 Tacit knowledge has been poorly converted into transferable knowledge. This is a critical shortcoming because the daunting complexity of bringing together engineers, artists, designers, marketers, and managers in intense iteration can lead to crunches, “intense and extended periods of socially mandatory overtime, and a seemingly perpetual start-up environment for game development companies.”36 There is little industry formalization and representation. Invoking the analytical work of Gina Neff and David Stark, O’Donnell asserts that the industry is in a state of “permanent beta.” Cross-disciplinary collaboration—which causes unremitting creative tension at the level of the firm and poses some of the most challenging project management tasks in contemporary industry practice—is absolutely necessary for the industry’s future. The tendency is for the industry, because of its closed opacity (and, as we have seen, because of its extreme volatility), to continually reinvent the wheel. O’Donnell stresses the great breakthrough by Unity when it made transparent authoring knowledge of great value, for example, for developing country industries.37 All of these factors contribute to working cultures and conditions that see 50 percent leave with up to ten years exposure.38 On the other hand, Australian industry, GDAA claims, is rare in the way it shares knowledge and resources among industry players now that the industry is composed overwhelmingly of indies. This is not typical of companies based in the United States, and was also not common when Australian developers were producing licensed IP, as a result of nondisclosure agreements.
A better articulation of the broader value of the industry to the society and economy can address precariousness. Antony Reed, an industry advocate, asserts that “this industry could make such a huge contribution if only it was understood better.”39 Advocacy, he argues, should seek to raise awareness of, for example, the value of game design input into health and education; the transferability of games skills into mainstream IT or the burgeoning apps industry; and the highly skilled entrepreneurial games workforce, which any country should seek to retain as part of its creative class. This draws on evidence that uncertainty of work in games is mitigated to some extent by the capacity to work outside the sector (due to the high transferability of skills, particularly of programmers). There is also some evidence that companies and individuals manage precarious original IP development with sourcing licensed IP opportunities within the growing domestic apps industry, so-called serious (edutainment) games, and a small range of domestic purchasers of games products and services (sourcing licensed IP domestically can be more sustainable because it is not subject to currency fluctuation). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/14%3A_Games_Production_in_Australia__Adapting_to_Precariousness/14.03%3A_Precariousness_as_a_Function_of_Policy_and_Industry_Cultures.txt |
Deuze, Martin, and Allen stress the importance of mapping what they call “gamework”: “the key issues informing and influencing the working lives and professional identities” of developers in the global computer and video game industry.40 Deuze and his colleagues were writing at a time when the dominant model involved developers working for large studios making games for publishing conglomerates like Electronic Arts. However, as we have seen in the case of the Australian industry, several options for making games and different workplace models confront developers. Some developers celebrate the creative freedom they experienced following the shift toward producing original IP games for mobile platforms, while others caution about the compromises associated with in-app monetization mechanics. The turmoil transforming the Australian games industry exemplifies precariousness. But it also includes adaptive experimentation in studio culture and associated changes in professional developer identity so as to continue the craft of making games in the midst of uncertainty. Analysts who have been very close to the industry and its developer culture, such as Casey O’Donnell, suggest that the current situation presents an opportunity to recapture the industry’s craft basis, the sustaining heart of the developer culture, stressing that gaming is not just a software industry.41 Creative destruction in the Australian games industry has been extraordinarily two-edged. As Gina Neff comments in the broader context of creative labor, “The trick for future media and business revolutions will be to find ways to support venture labor, so that innovative and creative jobs can also be stable and good jobs.”42
To achieve this, programs designed to support the industry need stability and predictability. Turning the public support spigot on and off according to political whim and policy fashion escalates precariousness. Furthermore, the industry needs better management practices. In addition to providing a more welcoming workplace for women and managing the crunch, it needs to learn how and when to cooperate as well as compete, and how to identify and incorporate new skill sets to deal with “runaway” innovation. Advocacy needs to articulate the wider value of the industry to society and economy, and to emphasize viable career structures within it. Precariousness, we have suggested, is an addressable matter—one that governments, the industry as an associative entity, and those who still make games can work on together.
14.05: Notes
1 Guy Standing, The Precariat: The New Dangerous Class (London: Bloomsbury, 2011).
2 Joseph A. Schumpeter, Capitalism, Socialism and Democracy (New York: Routledge, 2006 [1942]).
3 ABS, “8515.0—Digital Game Development Services, Australia, 2006–2007,” last issue April 8, 2008.
4 Ibid.; ABS, “8679.0—Film, Television, and Digital Games, Australia, 2011–2012,” last issue June 18, 2013.
5 Antony Reed, CEO, GDAA, interview with the authors, Melbourne, August 28, 2014.
6 Ibid.
7 Antony Reed, e-mail message to authors, September 30, 2014.
8 Rachel Parker, Stephen Cox, and Paul Thompson, “How Technological Change Affects Power Relations in Global Markets: Remote Developers in the Console and Mobile Games Industry,” Environment and Planning A 46.1 (2014): 168–185.
9 John Vanderhoef and Michael Curtin, “The Crunch Heard Round the World: The Global Era of Digital Game Labor,” in Production Studies: The Sequel! ed. Bridget Conor, Miranda Banks, and Vicki Mayer (New York: Routledge, 2015).
10 Larissa Hjorth, “Games: Mobile, Locative and Social,” in The Media and Communications in Australia, ed. Stuart Cunningham and Sue Turnbull (Crows Nest, NSW: Allen & Unwin, 2014), 281.
11 See John Banks, “The iPhone as Innovation Platform: Reimagining the Videogames Developer,” in Studying Mobile Media: Cultural Technologies, Mobile Communication, and the iPhone, ed. Larissa Hjorth, Jean Burgess, and Ingrid Richardson (New York: Routledge, 2012), 155–172.
12 John Banks and Stuart Cunningham, “Games and Entertainment Software,” in Handbook of the Digital Creative Economy, ed. Ruth Towse and Christian Handke (Cheltenham: Edward Elgar, 2013), 416–427.
13 “Number of Available Apps in the Apple App Store from July 2008 to September 2014,” October 12, 2014, www.statista.com/statistics/268251/number-of-apps-in-the-itunes-app-store-since-2008/; “Most Popular Apple App Store Categories in September 2014, by Share of Available Apps,” October 12, 2014, www.statista.com/topics/1729/app-stores/.
14 Innovation, captured in Schumpeter’s powerful phrase creative destruction, can have extraordinarily two-edged social, economic, and cultural effects. Runaway innovation occurs when change outstrips even its progenitors’ comfort and control. See, e.g., David Lane, “Towards an Agenda for Social Innovation,” European Centre for Living Technology, www.insiteproject.org/wp-content/ uploads/2014/02/Social-Innovation-Manifesto_INSITE.pdf, 2–3.
15 The research on which this chapter draws includes semistructured interviews conducted in Q3 2014 by the authors with sixteen developers from eight development studios, as well as with leaders in the games association and in government program support.
16 Dean Ferguson, interview with John Banks, August 21, 2014.
17 Morgan Jaffit, interview with John Banks, November 27, 2014.
18 Trent Kusters, interview with John Banks, September 24, 2014.
19 For a short history, see Banks, “The iPhone as Innovation Platform.”
20 Wicked Witch CEO Daniel Visser, interview with John Banks, September 23, 2014.
21 “Armello Press Kit,” http://press.armello.com/.
22 Trent Kusters interview.
23 Ibid.
24 Ibid.
25 Simon Joslin, interview with the authors, August 29, 2014.
26 Ibid.
27 George Fidler, interview with the authors, September 10, 2014.
28 Ibid.
29 The Media, Entertainment & Arts Alliance (MEAA), the largest and most established union and industry advocate for Australia’s creative professionals, has no section for games.
30 Christian McCrea, “Australian Video Games: The Collapse and Reconstruction of an Industry,” in Gaming Globally, ed. Nina Huntemann and Ben Aslinger (Basingstoke: Palgrave Macmillan, 2012), 203–207.
31 Screen Australia, “Australian Interactive Games Fund Industry Consultation: Objectives and Context,” www.screenaustralia.gov.au/gamesoptions/category/Objectives-context.
32 Queensland, the major northern state, and Victoria, the major southern state, have been the twin centers of the Australian industry; the largest state located between them (New South Wales) has had a smaller proportion of the industry.
33 Entertainment and Software Association, “2014 Sales Demographic and Usage Data: Essential Facts about the Computer and Video Game Industry,” ESA (2014), www.theesa.com/wp-content/ uploads/2014/10/ESA_EF_2014.pdf; DataGenetics, “Casual Games Demographics,” www.datagenetics.com/blog/december12010/.
34 IGDA, “Press Release: IGDA Developer Satisfaction Survey Results Are Released,” www.igda.org/news/179158/Press-Release-IGDA-Developer-Satisfaction-Survey-results-are-released.htm.
35 Casey O’Donnell, Developer’s Dilemma: The Secret World of Videogame Creators (Cambridge MA: MIT Press, 2014).
36 Ibid., 28.
37 Ibid., 74.
38 Ibid., 161.
39 Antony Reed, interview with the authors, August 28, 2014.
40 Mark Deuze, Chase Bowen Martin, and Christian Allen, “The Professional Identity of Gameworkers,” Convergence 31 (2007): 336.
41 Casey O’Donnell, “This Is Not a Software Industry,” in The Video Game Industry: Formation, Present State and Future, ed. P. Zackariasson & T.L. Wilson (New York: Routledge, 2012), 17–33.
42 Gina Neff, Venture Labor: Work and the Burden of Risk in Innovative Industries (Cambridge MA: MIT Press, 2012). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/14%3A_Games_Production_in_Australia__Adapting_to_Precariousness/14.04%3A_Conclusion.txt |
People often fantasize about Hollywood’s workforce being composed of innovative people imbued with refined tastes and aesthetics. They are further imagined as being well paid and therefore able to enjoy a rosy bohemian or bourgeois lifestyle, as opposed to other industrial workers. This romantic vision of the hip Hollywood creative may be apocryphal, but few would deny that the industry has long prospered because it has been able to foster and harness the creative energies of its employees.
• 15.1: Introduction
Introduction to the goal of this chapter: to investigate the values and attitudes of workers in game-related companies in Asia, and how they diverge from the traditional Western model of creativity.
• 15.2: Problematizing Creativity
Problematizing the model of the "creative class", developed in the 1990s from analysis of American creatives. Considering how varying socio-political contexts of creative industries result in different conceptions of creativity.
• 15.3: Subcontracting and Subcontracted Creativity
Forming a framework for cultural labor in globalized creative industries that considers how such labor in developing countries is often subcontracted from and subordinated to these global industries.
• 15.4: Three Major Production Regions in Asia
A brief overview of the historical and modern status of the games industry of three regions of Asia—Southeast Asia, Korea, and China—in relation to the global games industry.
• 15.5: Modes of Creative Labor
Developing a framework for analyzing globalized labor in creative industries, including subcontracted labor, with two major axes: creative dependence and tolerance for creativity.
• 15.6: Three Modes of Cultural Labor
Analyzing the cultures of the game industries in the three major Asian production locales, to identify the dominant mode of cultural labor in each: progressive artists in Korea, skilled enthusiasts in Southeast Asia, and contented bourgeois in China.
• 15.7: Conclusion – Creative Industries With or Without Creativity
Summary of the ways in which creativity is a relative concept, with reasons from globalized economic hierarchies to internal national ideologies.
• 15.8: Notes
15: Redefining Creative Labor East Asian Comparisons
People often fantasize about Hollywood’s workforce being composed of innovative people imbued with refined tastes and aesthetics. They are further imagined as being well paid and therefore able to enjoy a rosy bohemian or bourgeois lifestyle, as opposed to other industrial workers. This romantic vision of the hip Hollywood creative may be apocryphal, but few would deny that the industry has long prospered because it has been able to foster and harness the creative energies of its employees. Moreover, Hollywood has served as a model for other creative industries in the United States, including gaming, animation, software, and information technology (IT).
This paper offers an alternative perspective on creative labor by investigating the values and attitudes of workers in Asia. The data of this study is based on my face-to-face interviews with workers in different kinds of game-related companies in China, South Korea, Malaysia, Singapore, and Vietnam, as well as observation in their working sites from 2011 to 2013.1 These locales vary from large-scale factory-like game enterprises with over one thousand workers to small companies operated by a few personnel; they include online or video game companies, game distributors, and production houses that focus on animation, character design, or programming for online, mobile, and web games. The interviewees include workers of all levels: owners, artists, programmers, distributors, and promoters. These interviews cover various modes of creative labor in East and Southeast Asia. By comparing the lifestyles of these Asian workers with their U.S. counterparts, this chapter suggests that “creative labor” in East and Southeast Asia does not conform to the model described above, due largely to different industrial practices and cultural contexts as well as different experiences with processes of globalization. Given these divergences, the term cultural labor is more apt and comprehensive, indicating the ways video game production varies around the world. Moreover, this essay highlights distinctions within this Asian region, noting different attitudes, practices, and working conditions.
Three modes of cultural labor are theorized: in Korea, progressive artists, who are innovative in developing their entrepreneurship; in Southeast Asia, skilled conformers, who are “the arms” of the Western giants; in China, the contented bourgeoisie, who are skillful but less creative under state censorship. The presence of these emerging forms of cultural labor in Asia challenges the ethnocentric view of creative labor that has largely been shaped by North Atlantic tradition.
Previous studies of creativity suggested that research on creativity has been limited by ethnocentric boundaries in a world of cultural pluralism, implying that the traditional Western model of creativity is not appropriate in Southeast Asian countries. Maharaj Krishna Raina investigated the labor and lifestyles of Southeast Asian creative workers, concluding that they varied greatly from their Western counterparts.2 Beth Hennessey proposed that the concept of creativity is not applicable across nations, suggesting that creativity is constituted by both apparent and embedded values of different nations and is dependent on social agreements about what precisely constitutes creativity.3 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.01%3A_Introduction.txt |
The study of creative labor or the creative class attracted substantial academic interest in the 1990s. Richard Florida might be one of the first researchers to describe the emerging occupational, demographic, psychological, and economic profile of the American creative class.4 In his view, it is a privileged group that not only excels in creativity but leads a modern bohemian lifestyle. Moreover, this unconventional artistic existence is often associated with material comforts, cultural capital, and above-average working conditions. Problematizing the assumption that workers in creative industries are by default creative opens up a new imaginative space for conceptualizing this kind of labor.
David Hesmondhalgh and Sarah Baker’s book on creative labor is perhaps the most recent and comprehensive work to synthesize and problematize the concept of creative labor in the United Kingdom.5 They challenge the assumptions of autonomy, well-paid work, and the high quality of life of the cultural laborer. Moreover, recent debates about the concept of creativity have suggested that it is highly variable and contextually embedded.6 Hence the assumption that being creative is natural when technology and capital are in place is problematic. Similarly, in a study of artists and administrators for digital game companies in a small city in Canada, Laura Murray and others showed that they were often involved in contractual relationships with the audience and hence were swayed by the latter’s feedback, a relationship that challenged notions of autonomous creative genius.7
Creative industries in the United States, which usually involve exporting cultural commodities, are a crucial driver of economic revenue and account for a large portion of the country’s GDP. Other nation-states share these priorities, whereas for some countries, factors such as political interest and the vested interests of autocrats are more important. Yet even in countries like China, where political priorities prevail, cultural exports are seen as a way of exercising soft power and developing the economic power of their media and film industries.8 Under such circumstances, creative labor might benefit from top-down support for their industries, even though key elements of creativity, such as free expression, cultural tolerance, and the marketplace of ideas, may be stringently limited. The intriguing question is whether we redefine or requalify this type of creativity. If so, what specific characteristics might define alternative notions of creative labor?
In an effort to more accurately profile those working in the creative industries, I use the term cultural labor instead of creative labor. Whether it is creative or noncreative depends on the specific sociopolitical context. In China’s political environment, cultural labor is not “creative” enough to construct a virtual game world that would enable universal suffrage and voting. Yet an employee of Netease, a major online gaming platform, explains that Chinese game planners are often smart enough to bypass and outmaneuver constraints imposed by political leaders in order to launch and operate popular games.
In other words, the socio-political contexts in which creative industries are developed and sustained, and in which creative laborers work, produce, and are reproduced, result in different conceptions of creativity. When explained in terms of maneuvering around boundaries, creativity might be very limited; however, when understood in terms of entrepreneurial strategies aimed at navigating both market and political structures, they comprise a broader scope of cultural labor. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.02%3A_Problematizing_Creativity.txt |
A comparative perspective on creativity labor should also take into account the new international division of cultural labor in relation to creative industries, which is a result of cultural globalization.9 Florida, in the new edition of his book, also mentioned the global effects of the creative class,10 acknowledging that cultural globalization would inevitably create differences between people working in the center and those working on the periphery. The difference is a reflection of the dependence of new creative satellite cities or nations on the global media hub or transnational creative industries. As a value chain of cultural industries, it is a strategy for big transnational companies to search for the cheapest locations to “manufacture” cultural products using low-cost, labor-intensive processes in developing countries. However, for creative industries, the relocation of production is not simply a direct transference of manual labor from the established economy to developing countries. Game industries, for example, involve highly skilled labor in programming, computer graphics, and artwork. Thus contracting companies from the United States have to ensure that the skills and techniques in programming games are adopted by the contracted companies, and the aesthetics of the artwork have to be consistent throughout development. It is quite common that multiple subcontractors are used, but this doesn’t mean that all those working on the project share a collective vision. The contracted labor will most likely feel complacent about the arrangement; they often feel diffident about adopting the aesthetics of the giant lead companies. For example, Disney often subcontracts for artwork and Nintendo for animation, using many small- to medium-size teams for numerous aspects of a production. Moreover, the exploding popularity of online gaming among Asian consumers has further stimulated an expansion in the number and variety of companies operating in the gaming industries.
Despite the energetic growth of the gaming sector, Asian cultural labor is largely subordinated to global creative industries. The latter set the overall agenda for developing popular titles and prescribe aesthetics for the subcontractor to follow. In other words, global companies subcontract their version of creativity to game developers who must uncritically accept guidance from above. This involves standardization of both the professional knowledge needed to produce the work and the values and aesthetics needed to appreciate and legitimize the production. To understand the complex workings of market-submarket and prescribing and prescribed creativity requires a theoretical model that takes into account both the value-free cultural work and the potential commodification and fetishization of mass cultural production.
Reconciling critical theory, neo-Foucauldian, and liberal-democratic approaches, Banks’s critical framework on cultural work is useful here for examining the de facto nature of creativity and cultural labor across the neoliberal and capitalist markets dominated by multinational enterprises. In the context of Western cultural economies, he argues that there are always tensions between autonomous production on one side and corporate functionality of production and governmental prescription on the other, and this creates a spectrum of arguments from the discourse of moral, empowering cultural labor to the subordinated, alienated workers, with some alternative and moderate discourses in the middle.11
Putting the global context of cultural production of game industries in Banks’s framework, we can assume that while there are relatively free autonomous cultural workers in the major market of Western Hollywood production, there are also subcontracted cultural workers struggling in many other parts of the world, where the conditions for subcontractors vary by firm, location, and job. To understand the uneven and diverse terrains of game labor, it is important to critically survey differences in the condition of cultural labor among geographic locales without assuming that their lives, wages, working conditions, and (dis)empowering possibilities are equal. What follows is an overview of three important production locales: Southeast Asia, South Korea, and the People’s Republic of China. As we will see, government policies, market dynamics, and cultural specificity all have a significant impact on working conditions and on the attitudes and values among game company employees. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.03%3A_Subcontracting_and_Subcontracted_Creativity.txt |
The Dependency of Southeast Asia
Many local game developers in Southeast Asia, mostly small- and medium-size companies, were start-ups in the 1990s or 2000s. They sustained themselves through contracted, skilled work that they accepted from international game companies seeking to outsource projects. Although some companies produce small-budget games, such as mobile and flash games, they also produce quality art assets for global titles. They are perfect subcontractors for cross-platform games that operate on consoles and PCs, as well as mobile devices.
Several global game companies have chosen to set up studios in Southeast Asia because skilled talent there can be hired at lower pay rates. Such companies include Electronic Arts, Lucasfilm, Koei, Gevo, and Ubisoft in Singapore; Codemasters Studios in Malaysia; Square Enix and Gameloft in Indonesia; Activision, Bioware, Bungie, and Eidos in the Philippines; and Gameloft in Vietnam. In addition to serving as satellite production hubs, these operational outposts are sometimes used to distribute products in Southeast Asia as well.
Korea as a Rising Global Exporter
In East Asia, countries like Korea and Japan have a long history of colonial dependence, but they are not satisfied with being culturally colonized nations. In recent years, the concept of exporting culture has taken hold, particularly in Korea because of strong government support. Nowadays, Korea is the major hub for creative industries both in Asia and globally, particularly the games industry. One major breakthrough in addition to financial support is that state policies encourage a sufficient supply of cultural labor to support the industry.
The year 1997 was a watershed year, when the Korean government announced a scheme to boost the local games industry, including a plan to set up a Korean Game Promotion Centre (KGPC). This not only meant that governmental policy was focused on developing the Korean games industry, but it also set in motion a series of related state initiatives that attempted to reverse one-sided importation of culture from the U.S. and European markets. Early efforts involved the reorganization of governing bodies. For example, in 1998, the games industry had officially been grouped with cultural industries, implying that games were not only entertainment but also part of the national cultural arena. The launching of KGPC in July 1999 and its Japan branch two months later marked major steps in the government’s plan to nurture its national games industry. The mission of KGDI (Korean Game Development and Promotion Institute, later renamed the Korean Game Industry Agency [KGIA]) was to establish a strategic platform between the government and the games industry. The strategy included providing the industry with overseas market information, infrastructural support, and subsidies for research and development. In this way the government provided industry practitioners with timely technological and marketing advice.
The formation of the Ministry of Culture and Tourism (MCT) in 1998 and the transfer of the cultural assets policy portfolio, including the games industry, to MCT’s jurisdiction served to enhance industry expansion. The launch of Cyber Korea 21 one year later accelerated the development of the online games industry, particularly by speeding up broadband technology development, rapidly increasing the number of Internet users in Korea, and enhancing national Internet education. Moreover, the subsequent launch of the Korean Creative Content Agency (KOCCA) provided these industries with additional support, including equipment rental, investment, technological training, international marketing advice, and research support for medium- and long-term development, as well as developing strategic partnerships with overseas buyers and suppliers. To nurture talent for the booming games industry, the Korean government started the Games Academy in November 2000. A games investment association and a games investment valuation association were established in December 2000 and June 2001, respectively, to nurture talent and obtain venture funding and investments to meet the needs of the rising games industry. Such outcomes were indicative of the visible success of a state-led industry policy.
The critical mass of cultural labor formed under the state-driven model unquestionably adhered to the philosophy of the state, although in fact it is neither a free-market nor a neoliberal model, but is instead similar to the Hollywood model for cultural export. Consequently, the mind-set of workers in the Korean gaming sector is very close to the values and lifestyles of their American counterparts, largely because the United States is regarded as Korea’s benchmark of cultural exportation.
China's Global Expansion
Like Korea, China changed from an importer of online games (initially from Korea) to a major international player and exporter. According to the official statistics released by Ministry of Culture, in 2010 the total annual revenue of mainland China’s online games industry reached \$5.7 billion. It also became a substantial exporter of online games. In 2011, Chinese game companies generated \$360 million in overseas sales revenue from thirty-four games that performed especially well in Southeast Asia, North America, Korea, and Japan. Such success has encouraged China’s game giants to extend their overseas reach by acquiring major game titles and companies. In 2011, Tencent’s acquisition of Riot Games, the developer of League of Legends, for about \$230 million is typical of the global expansion of China’s game companies. League of Legends is the most popular PC game in North America and Europe, with an average of 27 million gamers daily.
However, the industry developed much faster than the regulations did. Before 2004, there were no regulations or cultural policy to drive or control the industries.12 In subsequent years, Chinese authorities introduced a series of regulations: the Regulation on Digital Publication in 2007, the Regulation on Publishing of Digital Publication in 2008, and the Administration of Software Production in 2009. A censorship system was promulgated in August 2010, when the state delegated such censorship to the provincial level (People’s Congress Decision on the 5th Batch Cancellation of and Delegation of Approval to Level of Management and Controlling Unit). In the name of protecting minors, most of the regulatory guidelines focus on controlling violence and indecency. However, my interviews with the committee responsible for censorship revealed that ideological controls are fairly common, thus hindering the import and publishing of foreign games in China.
Despite these stringent controls, the authorities have not hindered private investment and game support. Instead, as incentives, tax reductions are given to game companies that export, and the provincial and local authorities set up technology areas or cultural clusters to cater to the needs of game companies. The challenge for game companies, as I will describe, is the general lack of talent and the high cost of recruiting teams to develop games.
15.05: Modes of Creative Labor
The conditions in which creative goods are produced affect labor. In the U.S. context, Richard Caves explained that the production of creative goods is largely susceptible to basic economic properties, including commitment and devotion of labor, skills needed, product differentiation, time and demand, and costs incurred.13 However, in non-free markets and nondemocratic states, many other factors shape, foster, or dictate the conditions of how cultural labor is produced, trained, and socialized. As explained earlier, on the social level, in Asia the state plays a prominent role in driving creative industries that require appropriate types of workers. Even more directly, the state plays an active role in training and nurturing cultural labor for emerging industries. To a certain extent, cultural workers adhere to the values and worldview prescribed or controlled by the state. Even if they are not totally synchronized, in areas like Asia, workers’ innovation under the current neoliberal market is always driven by, intertwined with, and managed by the state and multinational enterprises.14
On the institutional level, the specific features of creative industries structure the know-how, attitudes, and type of creativity needed. This is not just a matter of employing talent. When cultural workers invest their time in creative industries, they internalize the norms and roles imposed by these corporations. The entire range of cultural production, from unpaid digital laborers who participate in the production of blogs and free content to paid, highly skilled cultural labor, is “channeled” and “structured” within capitalist processes of consumption and production perpetuated in multinational corporations.15
In summary, both macro and medium organizations have set boundaries for cultural labor, and hence shaped their values, lifestyles, and even political ideologies. I am not arguing that cultural labor is deprived of free will. Nor do I posit that such labor lacks “creativity,” understood in the Western context. However, rarely can the cultural labor in a particular locale go beyond the ideological boundaries of the workplace and the political or regulatory system of the regime. This chapter formulates a framework that takes into account two major dimensions: creative dependence and creativity tolerance.
In the tumultuous drive toward cultural globalization, subcontracting activities have become a key issue in creativity dependence. The less jobs are subcontracted by the creative industries, the more autonomy companies have to develop their own creativity. On the contrary, the more creative industries depend on subcontracting work, the more labor must have the professional skills and standards that the company requires. In other words, the creativity of the subcontracted labor can only be a replication or derivation of the subcontracting global companies.
The cultural, economic, and political context plays an important role in determining the motivations of cultural labor.16 In addition to financial incentives, the so-called atmosphere, or “people climate,” plays an important role in attracting talented workers to creative enterprises.17 This is further affected by the ideology of the state, which shapes the overall tenor of public attitudes toward innovation and diversity.18 From the perspective of political economy, stronger politico-economic control diminishes the creative expression in the cultural products. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.04%3A_Three_Major_Production_Regions_in_Asia.txt |
The two dimensions of creativity dependence and creativity tolerance yield four possible quadrants or formations of cultural labor. The fieldwork in Asia yielded at least three modes of cultural labor.
Progressive Artists
Cultural labor in Korea works in a democratic, free-market environment where the business model of games depends on both internal consumption and export. The conditions of the market, the democratic system, and the business models of the games companies are similar to those in the United States. Koreans used to be the leading players in the games market in Asia and still contribute to many leading games on the market. One of the most important games is Aion: The Tower of Eternity, a MMORPG released by NCsoft, a major Korean game developer. It was first released in Asia in 2009, and by May 2009 had acquired 3.5 million subscribers in the region, where China is the largest market under its operator Shanda Interactive Entertainment. Aion was later localized for Western markets, including North America, Europe, and Australia.
My interview with a former NCsoft programmer who was involved in developing Aion revealed the conditions of Korean cultural labor. The venue of the interview was picked by the interviewee: the backyard of a stylish coffee shop in the Bochun area, a district featuring well-preserved traditional Korean houses where artists reside. During the interview, I saw visitors going back and forth taking photos. The interviewee was well aware of the overly romanticized lifestyle of cultural labor in the area. Because of their strong programming skills and sophisticated tastes, which are in high demand, they are able to pursue a relatively uninhibited lifestyle. The interviewee is now a graduate student at Seoul National University writing a thesis on the Korean creative industry. He has confirmed that he will return to work at NCsoft after completing his graduate degree. Coincidentally, another programmer I met was also a graduate student, at Yonsei University. As he explained, he became interested in games when he was in high school. His mother promised to pay his tuition at a private game college where students studied game programming, if he passed his college entrance exams with flying colors. He did, which allowed him to study game design and work for a game company while also studying as an undergraduate student at Yonsei. The money he earns as a game developer is considerable and allows him to pay the tuition at Yonsei, which is one of the most expensive private universities in Korea.
The information revealed in the interviews suggests that these cultural workers are part of an educated elite in Korea. They can be regarded as middle-class workers with a flexible schedule and relative creative autonomy. They readily switch between studies and work, and living expenses are not a financial burden. Hence, they resemble artists enjoying a high standard of living. Moreover, they are able to live in downtown Seoul, which is unaffordable for many university graduates. The interview with another Seoul National University student who wanted to enter the industry revealed that the expected average annual salary was around two to three million Korean Won (US\$20,000–30,000), which is not a particularly high salary, but the increments could be very high, depending on performance.
I deliberately discussed politics in Korea with them, particularly asking them about President Park Geun-hye, the first woman president in East Asia and the daughter of Park Chung-hee, president of South Korea from 1963 to 1979, when South Korea was a military dictatorship. Interestingly, the interviewees did not seem concerned with politics and soon changed the topic to the history of gaming in Korea, game development, and the support provided by KOCCA. This does not mean that they were not reflective or critical, but it is a significant shift from their parents’ generation, when educated young Koreans were passionate about politics. Unlike the old days under authoritarian rule, they have a strong belief in the current electoral and democratic system, which allows individualism and generally supports creativity. Subscribing to civic nationalism, they highly value the authorities’ rapport with the games industry.
Skilled Conformers
Cultural laborers working in the Southeast Asian region, including Singapore, Malaysia, and Vietnam, can be considered skilled conformers. In the major cities in these countries, cultural workers take pride in subcontracted jobs from Hollywood and elsewhere. In response to our question of why Singapore was chosen to inaugurate a Disney game in Asia, the CEO of Infocomm said that for Disney, “it is difficult to have control in other countries. [Singapore] is the only place that they will feel that everything will be accomplished according to their plan.”
Asian cultural labor perceives China as the most profitable market in Asia aside from the United States and Europe (the CEO of the distributor of World of Warcraft in Singapore concurred). If these small- and medium-size Asian companies could break into the Chinese market, American game giants would entrust them to be the distributors to the Chinese market. The dependence on the American market, in their view, is not imbalanced; instead their view aligns with the government’s point of view, and they feel privileged to be the “Asian arm of Hollywood.” They believe that their creative industries have greatly contributed to making Singapore an international cosmopolitan city on par with the professional standards of Hollywood. In 2014, the Singaporean government aimed to attract Disney and Lucasfilm to locate their Asian headquarters in a futuristic horseshoe-shaped building, the Sandcrawler, in Fusionopolis, the new cyber area of Singapore. In a joint effort, the National University of Singapore and the Media Development Authority launched the Singapore Hollywood Attachment Programme, which networks with top IDM (Interactive Digital Media) institutions in the United States to place Singaporean students. These are conscious efforts by Singapore to keep up with the West. In short, they want to follow, conform to, and defend the Western standard of creativity.
On the other hand, the philosophy of cultural labor is synchronized with the official ideology of the state. Executives and employees of major game companies in the region express a strong sense of social contribution and harmony, values they share with the authorities. Infocomm, the Singapore game distributor mentioned earlier, launched many offline activities to attract online game players to socially beneficial causes, including their blood donation event. Virtual gifts were given to players who participated in the event.
According to the CEO of a major Malaysian game company, Codemasters Studio, both the company and its employees embrace global standards while also maintaining local values. Codemasters, which has been in operation for seven years, is like many other companies in Southeast Asia. It started by producing CGI for Pachinko and now focuses on subcontracted game art for overseas clients in North America, the United States, the United Kingdom, Europe, and Japan. The CEO of the company explained that he is proud of the fact that most employees were trained in Malaysia but they are all familiar with creative products from around the world. Given that awareness, he notes, “When you talk about style we are quite flexible. I will say that it is our strength actually. So we don’t really favor any type of style, but we are quite flexible. When we actually talk to people like those from the UK, or from the U.S., we communicate quite well. Moreover, I think the main strength is our communication.”
The strength of these Malaysian laborers, as they see it, is that they embrace Western aesthetics. In fact, most feel “superior” in that they share among themselves values that are not local but cosmopolitan. In other words, they regard Western values as more modern, trendy, and worldly than indigenous Malaysian values. As we found in our fieldwork, most of these workers display Western pop culture decorations—including posters, toys, and games—in their offices, work spaces, and production sites to demonstrate their artistic, avant-garde, modern values. Their work spaces provide a stark contrast to other offices in Kuala Lumpur.
Contented Bourgeois
Cultural labor in China is characterized by contented bourgeoisie. China’s domestic cultural market is as large as the American market, and in Asia, China competes with Korea for exports. There is always a perceived shortage of talent in the market. The entire cultural labor market in China can be explained by the literal Chinese translation of Yu Xin’s (513–581) ancient Chinese expression “Crouching tiger, hidden dragon.” The essence is that China is full of talent that remains unseen and undiscovered. The fact is that in major Chinese cities and universities, this hidden talent has to be actively sought by major game companies. They prefer to appear as “curling roots”—crouching tigers that are content with politics and society as long as their lives are settled. Thus if game companies want to expand, they must provide incentives to attract workers from leading universities and other game companies.
The head of a game engine programming team at Perfect World Company—one of the largest online game companies specializing in MMOPRG in China and an IPO company on the NASDAQ—described the trajectory of his work experience. He started with a Taiwanese company and produced the strategy game Three Kingdoms. He then was brought in by Perfect World to lead a game production team. When asked about his views on developing games, he offered a very pessimistic response, saying that it is very difficult to break out of the corporate culture to launch a small start-up. Although his team is exceptionally talented, with the capacity and know-know to develop a full-fledged game engine on its own, the path is full of obstacles. The biggest problem is financing. He mentioned that at least two phases of capital investment are required to develop a game. On average, RMB 12 million is required in the first year, mainly for salary, and in the second year, a special bonus or commission has to be distributed to the team. The high investment makes it impossible for him to start his own business. Hence most people seek refuge in big companies, where the corporate culture is pragmatic rather than imaginative or spiritual. The guiding principle is the contractual relationship, which specifies incentives based on commissions or stock options. When asked how long he would continue to work for this game company, he responded, “Right now I am quite distressed. The day drags—it’s okay to stay in a big company. People resigned because [the job] is too demanding and life is too stressful. I am now twenty-six. I am too old; I can’t move.”
However, in terms of salary relative to living standard, these workers in fact lead affluent lives. Based on our estimation of his monthly salary, which is at least RMB 30,000, the interviewee could live luxuriously. However, he rented a small, old apartment and worked long hours, and he said that he did not have much of a personal life in the city, since he devotes most of his time to his job.
We heard similar concerns from the vice president of Perfect World. The interview was conducted in a canteen on the ground floor of a complex located in the Shangdi district of Beijing, an area designated for high-tech industries. Despite the building’s postmodern appearance, office decorations hardly reflected the bohemian values of the so-called cultural laborers. Against the plain gray walls and floor, the tables were packed close to each other, and people sat back-to-back on low-budget chairs in a closed, noisy, stuffy environment. In contrast to this dull, monochromatic interior, the food was extraordinary! The entire interview was conducted in a quiet corner of the canteen. I was on my way out after finishing the interview when I came across a senior programmer sitting at a crowded table outside the canteen. He seemed socially detached and relatively subdued. I would say that he and his team, who sat around him, were quite content with the bland office environment.
This visit to Perfect World immediately deflated my fantasy about these cultural workers. It seems that on both organizational and individual levels, the lifestyles of these workers were among their lowest priorities. My visits to the game production sites of the giants, Perfect World and Qilin, as well as to smaller-scale game companies, reconfirmed my presumption that these workers were concerned with sales figures, not lifestyle. In contrast to high-tech companies like Yahoo! and Gamania, which equip their offices with coffee bars and play corners against walls of highly contrasting orange and gray, these companies have large-scale offices with endless partitions and small cubicles where a programmer or artist sits.
It could be said of cultural labor in China that despite the “creative” nature of the industries, the programmers, artists, and marketers in these game companies resemble industrial workers in their tastes, aesthetics, and lifestyles. Their offices and hubs, sometimes called creative clusters, resemble factories where games are produced on computers and servers instead of machines on an assembly line.
The blind pursuit of wealth amid a chaotic market is typical of China. From the perspective of cultural laborers, because they are constantly exploited by the system, what matters to them is the immediate financial reward, not a fancy lifestyle that might not be sustained for long. While the individualistic pursuit of desire is on par with those in the creative industries in the United States and Europe, other factors differ. On the individual level, the assumption of a democracy of economic agents is not valid because China is not a state that allows free choice.19 In the market, if competition or entrepreneurial actions drive knowledge, then that knowledge, hence creativity, does not exist because the market, particularly the cultural market (for example, media, books, and films), is dominated by a very few wholly or partly state-owned companies.20 Even when creativity exists, it is distorted. In this unregulated “market,” both investors and entrepreneurial consumers seek to navigate an array of choices to maximize their benefits without the baggage of intellectual property and copyright as in the West.21 As many interviewees expressed, once a game of considerable popularity is published by a small company, major players simply clone it and improve its quality). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.06%3A_Three_Modes_of_Cultural_Labor.txt |
The three modes of cultural labor discussed in this study reveal key differences in the working conditions and lifestyles of game company employees in Asia. Of course, the reality is more complicated than the theoretical constructs. The positions of cultural labor, the changing nature of the creative industries, the political atmosphere, and the degree of urbanism in which the creative industries operate could change the nature and relative positions of the three modes of cultural labor.
The findings showed that the concept of creativity is relative. As long as the political economy of transnational corporations is robust and stable, the global division of labor remains. However, there must be locales in which creativity is highly valued and protected for global cultural production; at the same time, there must be dependent satellite locales in which creativity is less valued, as they are serving the center of production. If creativity industries are ideally meant to foster cultural diversity, social inclusion, and a wider development pathway—as clearly indicated in UNESCO’s 2013 report on the creative economy—the Asian cases indicate that conditions on the ground are more complicated.22 Given the existing economic structures, the creative economy in some Asian countries is deemed dependent and secondary, and their creativity, if any, is derivative of the transnational corporations.
Besides the impact of global hierarchies, we have also seen that the internal dynamics within a nation greatly affect the conditions of cultural labor. Where governance is top-down and state-driven, cultural workers subscribe to ideologies of development ranging from capitalist democracy and neoliberalism to socialist economy, cultural nationalism, religious economy, and state corporatism. For example, for China and Korea, national cultural policy basically dictates the development of creative industries, the content produced, and the products exported, whereas in Malaysia, Vietnam, and Singapore, cultural workers are still immersed in the illusive gaiety of prescribed creativity. The different philosophies of cultural policy nourish two very different modes of cultural labor. Cultural policy also varies according to the regime, the regional economy, and the relative competitiveness of creative industries in the region. Taken as a whole, these forces and influences encourage us to consider the cultural valences and the theoretical implications of concepts such as precarity, creativity, and creative labor.
15.08: Notes
This work was fully supported by a grant from the Research Grant Council of Hong Kong Special Administrative Region (Project no. 4001-SPPR-09).
1 Data collection was conducted from 2011 to 2013. The research team interviewed seventy informants in China, South Korea, Malaysia, Singapore, and Vietnam. All are practitioners in the game industry, including online games, mobile games, handheld games, console games, computer games, and arcade games.
2 Maharaj Krishna Raina, The Creative Passion: E. Paul Torrance’s Voyages of Discovering Creativity (Stamford, CT: Ablex, 2000).
3 Beth Hennessey, “The Social Psychology of Creativity,” Scandinavian Journal of Educational Research 47.2 (2003): 253–271.
4 Richard Florida, The Rise of the Creative Class and How It’s Transforming Work, Leisure, Community and Everyday Life. (New York: Basic Books, 2002).
5 David Hesmondhalgh and Sarah Baker, Creative Labour: Media Work in Three Cultural Industries (New York: Routledge, 2011).
6 Todd Lubart, “Cross-Cultural Perspectives on Creativity,” in The Cambridge Handbook of Creativity, ed. James C. Kaufman and Robert J. Sternberg (New York: Cambridge University Press, 2010), 265–278.
7 Laura Murray, S. Tina Piper, and Kirsty Robertson, Putting Intellectual Property in Its Place: Rights Discourses, Creative Labor and the Everyday (New York: Oxford University Press, 2014).
8 Michael Keane, Creative Industries in China: Art, Design and Media. (Cambridge: Polity, 2013).
9 Toby Miller, Nitin Govil, John McMurria, Richard Maxwell, and Ting Wang, Global Hollywood 2 (London: British Film Institute, 2008).
10 Richard Florida, The Rise of Creative Class, Revisited (New York: Basic Books, 2012).
11 Mark Banks, The Politics of Cultural Work (New York: Palgrave Macmillan, 2007).
12 Anthony Fung and Vicky Ho, “Cultural Policy, Chinese National Identity and Globalization,” in Global Media and National Policies: The Return of the State, ed. T. Flew, P. Iosifidis, and J. Steemers (New York: Palgrave Macmillan, 2016).
13 Richard Caves, Creative Industries: Contracts between Art and Commerce (Cambridge, MA: Harvard University Press, 2002).
14 Maureen McKelvey and Sharmistha Bachi-Sen, eds., Innovation Spaces in Asia: Entrepreneurs, Multinational Enterprises and Policy (Cheltenham: Edward Elgar, 2015).
15 Tiziana Terranova, “Free Labor,” in Digital Labor: The Internet as Playground and Factory, ed. T. Scholz (New York: Routledge, 2013), 33–57.
16 Murray, Piper, and Robertson, Putting Intellectual Property in Its Place.
17 Florida, The Rise of Creative Class, Revisited.
18 Florida, The Rise of the Creative Class.
19 Pitman Pott, “Belief in Control: Regulation of Religion in China,” in China Quarterly 174 (2003): 317–337.
20 Ibid.
21 Lucy Montgomery, China’s Creative Industries: Copyright, Social Network Markets and the Business of Culture in a Digital Age (Cheltenham: Edward Elgar, 2010).
22 UNESCO, Creative Economy Report 2013 Special Edition: Widening Local Development Pathways (2013), www.unesco.org/culture/pdf/creative-economy-report-2013.pdf | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/15%3A_Redefining_Creative_Labor__East_Asian_Comparisons/15.07%3A_Conclusion__Creative_Industries_With_or_Without_Creativity.txt |
Job security is under unprecedented threat in many developed nations as a consequence of the mechanization of work.¹ In addition, rising production costs are seeing the relocation of production to low-cost locations. This is a well-known story. Emerging economies are achieving substantial growth by providing cheap labor and preferential investment policies. For China, already an economic powerhouse, a foreign country’s insecurity is a their security: the “made in China” phenomenon manifests in products that are designed elsewhere and fabricated in China. Much of this outsourced production involves components. Economists call this “tradein-tasks,” “unbundling,”² or OEM (original equipment manufacturing).
• 16.1: Introduction
Introduction to the goals of the chapter: unbundling the concept of creative precarity and analyzing the relationship between creativity and knowledge capital, focusing on China.
• 16.2: Contextualizing Precarious Creativity
Contextualizing creative precarity in China, with an introduction to the concepts of “knowledge-how” and “knowledge-to.”
• 16.3: A Chinese Political Economy Framework
Considering the social and political context of labor in China, and how this distinguishes the meaning of precarious creativity in China from that in other countries.
• 16.4: Knowledge Capital
Definition of knowledge capital, and how it relates to both knowing-that and knowing-how.
• 16.5: Media – The Game Changes
Recent changes in the Chinese media industry, including a rise in workers trained in multinational companies transferring their knowledge to Chinese companies, the commercialization of broadcasting industries, and the growth of private-owned production companies which are still subject to state censorship.
• 16.6: The Cultural Innovation Timeline Reconsidered
An argument for the concept of creative precarity in China being not simply a negative indicator, using the cultural innovation timeline (standardized production, imitation, collaboration, cultural trade, production/management consolidation, peer/creative communities).
• 16.7: Concluding Remarks – The (Precarious) Elephant in the Room
How creative precarity in China is distinguished from creative precarity in other developed nations, by censorship and the existence of an additional dimension of knowledge capital (“knowing-to”).
• 16.8: Notes
16: Unbundling Precarious Creativity in China Knowing-How and Knowing-To
In the industrially developed countries they run their enterprises with fewer people and with greater efficiency and they know how to do business. All this should be learned well in accordance with our own principles in order to improve our work.
—Mao Zedong, On Contradictions, 1956
They say low wages are a reality
If we want to compete abroad.
—Bob Dylan, “Workingman’s Blues #2”
Job security is under unprecedented threat in many developed nations as a consequence of the mechanization of work.1 In addition, rising production costs are seeing the relocation of production to low-cost locations. This is a well-known story. Emerging economies are achieving substantial growth by providing cheap labor and preferential investment policies. For China, already an economic powerhouse, a foreign country’s insecurity is their security: the “made in China” phenomenon manifests in products that are designed elsewhere and fabricated in China. Much of this outsourced production involves components. Economists call this “trade-in-tasks,” “unbundling,”2 or OEM (original equipment manufacturing).
In this chapter, I attempt to unbundle precarious creativity, a concept that is somewhat ambiguous and misconstrued.3 I look at the relationship between creativity and knowledge capital. Knowledge capital is a currency that is much sought after in the PRC and in some respects overseas players are temporary custodians: the relationship of knowledge capital to “precarious creativity” is therefore worth exploring.
I begin by contextualizing precarity in China’s workforce. Following this, I explore the idea of knowledge. I discuss the distinction between “knowing-that” (propositional knowledge) and “knowing-how” (the acquisition of abilities and skills). I then turn to the question of how knowledge capital and precarious creativity apply to China’s media and cultural industries, specifically animation and television. In the final section, I explain variants of precarious creativity by drawing on a heuristic called the “cultural innovation timeline,” which shows how many policy makers and commentators see the gap between China and its competitors closing. I argue that it is closing because employment is mobile and because knowledge (know-how) is being transferred. But it is also closing because the world is coming to China, not because China is going to the world. In the conclusion, I examine censorship, the “elephant in the room.” The precariousness of expression in China affects all cultural and media workers, Chinese and foreign. Finally, I argue that the value of know-how is augmented by “knowing-to,” a disposition that emanates from cultural and political contexts and constitutes a crucial modality of knowledge capital for persons looking to operate successfully in the Chinese market. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.01%3A_Introduction.txt |
The notion of precarity and the neologism precariat have emerged in academia over the past decade to account for the way the labor market is reorganizing in many developed economies in response to flexible forms of capitalism; for instance, an increasing number of jobs are listed as casual without fixed incomes or benefits. Skills learned in schools and universities, such as reading and theorizing, are losing value in occupations that rely on on-the job learning.4 When this argument is extended to art, design, and media sectors, we are informed of the condition of “precarious creativity.” On the surface, this coinage conjures up a dark side to creativity. In contrast to a wide-ranging consensus among educators, psychologists, and business leaders that creativity is positive and aspirational, there are now negative externalities to consider, among which is the apparently transient nature of employment in many creative sectors. Many scholars opt to use precarity as a corrective to euphoric claims associated with the creative economy, particularly that it is expanding globally and generating more meaningful jobs.5
When used to refer to cultural and creative labor, precarity normally picks up on the employment insecurity of workers in industry sectors affected by technological convergence—for instance, music, film and TV production, online games, and design—rather than those providing lower-level service jobs in the same industries.6 Of course, creative work itself is difficult to define, and it is beyond the scope of this essay to investigate gradations of creative labor intensity. What can be argued, however, is that creative products and services sold to consumer markets are generally produced by people with specifically acquired skill sets. My own research into media parks and creative clusters in China identified that 95 percent of creative workers had tertiary degrees, mostly undergraduate (54 percent).7 Moreover, such technical and managerial skills can be easily learned or transferred when businesses move offshore, particularly when R&D sharing is part of the market entry equation.8 For instance, AnnaLee Saxenian has characterized the migration of the Taiwanese integrated circuit (IC) supply chain to Shanghai in the early 1990s as “perhaps the greatest transfer of managerial and technical skills in human history.”9
In China, where the nation’s capital stock has accumulated largely by virtue of “sweat industries,” the discourse of creativity juxtaposes productivity gains and labor market transformation. It promises a way to lift masses of people out of polluting, repetition-based industries and move them into value-adding service sectors while at the same time revitalizing domestic cultural and content industries by making them internationally successful; it offers what might be termed “cultural soft power” dividends.10 Indeed, gains in expertise and innovation in new media sectors, which are less burdened by regulation, are assisting the Chinese government in its mission to extend the nation’s soft power internationally. The key factor is knowledge—or more specifically, know-how. The concept of know-how is by now fairly well entrenched in management literature. In speaking of China, moreover, it is worthwhile noting the epistemological distinction between knowledge-that and knowledge-how, as elaborated by Gilbert Ryle in the 1940s.11 Knowledge-that constitutes propositional knowledge, things that we know about the world. Ryle believed that knowing-how is a “higher-grade disposition,” associated with “abilities and propensities” as well as “capacities, habits, liabilities and bents.”12
Many view the challenge in terms of “catching up” with and learning from advanced soft-power nations in terms of acquiring more know-how. While the Chinese government is reluctant to openly identify such “know-how-rich” nations, there is no doubt that most practitioners in the creative industries target Western developed economies as well as Japan and South Korea. These have become China’s “soft power competitors.”13 The acquisition of foreign know-how, in addition to codified intellectual knowledge (knowledge-that), readily obtainable from reports and scholarship, offers a key that can unlock secrets of innovation.
For a Chinese person, the chance to work in a foreign company may be the means of acquiring both crucial know-how and know-that. But how long the worker stays with a company depends on salary, job satisfaction, and career expectations. The inclination to change occupations can be explained in terms of “compensating differentials,” that is, the coexistence of monetary and nonmonetary elements of employment.14 People undertake jobs for a variety of reasons: in many creative industries, some work for less or work long hours because they enjoy the work they are doing and the people they associate with. Moreover, in a market like China, where there are plenty of job openings in new media sectors, workers can experience significant mobility. Taking knowledge gained, including IP, elsewhere is therefore another variant of precarious creativity. In short, the Chinese government hopes that the transfer of international knowledge together with an understanding of markets and consumer preferences might contribute to the rise of Chinese media influence. Whether this rise signals the receding influence of international media in China is a moot point. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.02%3A_Contextualizing_Precarious_Creativity.txt |
There are several ways of understanding precarious creativity in China. The first is to recognize that China, like many other countries, faces new opportunities from information abundance. Technology is having an impact on traditional patterns of life, as distant friends and potential customers are connected instantaneously through apps like WeChat and Taobao. Second, rapid urbanization has significantly altered the demographic pattern of Chinese society. One study estimates that China will have more than two hundred cities of over one million inhabitants by 2025.15 Urbanization changes the mobility of the workforce as more people are drawn to opportunities in big cities. Third, the One Child Policy, instituted in 1978 to curb population growth, has skewed population demographics, giving rise to a generation without siblings.16 Fourth, recent liberalizations in the household registration system (hukou),17 have increased people’s ability to change employment. In tandem with unprecedented mobility and technological change, skill shortages are appearing in the workforce, a problem that is bound to continue over time as a result of the One Child Policy, with fewer young people transiting into the labor market.
Industries need labor. In China the term industry has an ever-present relationship with economic modernization. Policy documents emanating from Beijing, particularly the five-year economic and social plans that underpin the allocation of key government resources, emphasize the industrialization of welfare, manufacturing, education, and even culture.18 Whereas the English word industry comes from the Latin industria and refers to “diligence, activity and zeal,” the dominant term in China until recently was gongye, literally the “activity of physical labor.”19 The use of the body, more than the mind, reminds us of the agrarian base of Chinese society until the mid-twentieth century. The sustainability of the Chinese economy from a so-called feudal agrarian system prior to the Chinese Revolution in 1949 to the socialist commune system of the late 1950s was founded on manual labor. The ensuing rise of export-led manufacturing in the 1980s and 1990s entailed further separation of mind and body, resulting in an intensification of production lines throughout the country.
By the turn of the century, this “new factory system” was well entrenched, drawing migrant laborers into working conditions that were often unsafe and exploitative. Migration to cities led to increasing social fragmentation and exacerbated informal employment.20 Laborers, predominantly male, toiled in urban construction projects from high-rise buildings to ostentatiously named “cultural and creative clusters,”21 while female workers offered housekeeping (baomu) duties for urban residents or serviced the bodies of the middle class in thousands of massage parlors. Sweatshops proliferated on the fringes of cities, taking in work from overseas clients. As Loretta Napoleoni comments, “In the second half of the first decade of the twenty-first century China becomes the center of the global assembly line, the pieces produced at lower costs in neighboring countries and put together in Chinese factories.”22
However, it is difficult to equate precarity in such labor-intensive sectors with media and cultural industries. In the latter, we see widespread transfers of knowledge capital that can translate into social mobility. Indeed, the zones of attraction and influence for China’s creative classes are distinct from the labor-intensive Special Economic Zones (SOEs), which have led to a proliferation of sweatshops and global assembly lines. Beijing and Shanghai in particular draw creative migrants into their cosmopolitan orbits.23
While the precariousness of creative work is the subject of a number of important studies,24 precarity in China’s cultural and creative industries requires us to be cognizant of social and political context. I will return to this point in the conclusion. In most usages, precarious creativity refers to unstable employment in occupations that generate symbolic goods and services—for example, design, VFX and film, and software. Most international depictions relate to market economies where the hand of government is at a distance.25 In a country where freedom of expression is constrained by politics, the term precarious creativity implies something quite different. The hand of government is very visible. Even when it is less evident, for instance in design, fashion, and music, there is usually a need to appease a government official somewhere. This situation reflects the organization of cultural production under socialism, still the prescribed ideology in Chinese schools today.
16.04: Knowledge Capital
The importance of knowledge to creative industries is on the surface uncontroversial. Most people accept the proposition that the creative industries are knowledge based. Knowledge is a cognitive capacity, conventionally understood as expertise. In conventional media and cultural sector value chains, expertise is valued as an input into content generation, delivery, and sales (that is, marketing). Knowledge capital, sometimes referred to as intellectual capital, includes the workforce (human capital), demands and preferences of audiences and consumers (customer capital), and systems, products, processes, and capabilities (structural capital).26 While this is invariably knowing-that, it embodies know-how.
Knowledge capital can be sticky: it is often difficult to transmit or export. The term absorptive capacity describes the capacity to absorb knowledge that presents as “spillovers,” the latter term implying unintended consequences of actions. Both spillovers and absorptive capacities exist because an organization or an individual cannot capture all the benefits resulting from inventive activity. A good example is business precincts where there are convivial spaces in which people meet and share ideas informally. Localized spillovers and cultures of interaction frequently occur when participants are close to the knowledge source. Such clustering allows the exchange of tacit knowledge: that is, people may become smarter by interacting with each other. Creativity, learning, knowledge networks, and innovation occur because of skilled labor markets and movement of people. Media capitals like Hollywood and Mumbai, and technology hubs such as Silicon Valley, provide evidence of how knowledge capital is shared.27
In the past few years, a rise in collaborative production opportunities in China together with the construction of cultural parks and media bases has led to significant transfers of knowledge capital: this includes human capital, customers, and structural capital as well as technological know-how. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.03%3A_A_Chinese_Political_Economy_Framework.txt |
One senses a strong belief among China’s creative workforce in the rising power of Chinese media and cultural production. At the moment workers are in demand in China; pay, conditions, and job satisfaction exceed many other service occupations. While data from the National Bureau of Statistics is not fine grained in term of occupational categories, it does indicate that salaries in “culture, sports, and entertainment” occupations are higher than other service industries, not a surprising finding considering the amount of investment, both domestic and foreign, that has taken place over the past several years. In addition, the largest increase in salaries is in the category “information transmission, computer service, and software.” In 2009, the average wage for workers in this area was RMB 58,154 (approx. US\$9,500); in 2013 had become RMB 90, 926 (US\$14,800).28 Discussions of bonuses paid by domestic technology and games companies like Tencent can be found online.29 Successful projects can deliver dividends to creators that often exceed monthly salaries.
Job opportunities are increasing, allowing workers to move from place to place, from job to job. The animation and gaming industry, despite its reliance on outsourcing contracts, is a case in point. Cities such as Beijing, Shanghai, Suzhou, Hangzhou, Wuxi, Guangzhou, Shenzhen, and Dalian are competing to be the “animation capital” of China, and there is a relative shortage of “talent.” Video gaming is indicative of demand. Data shows that salaries in the games industry are highest in Beijing, where the category “technology R&D” (jishu yuanfa) dominates; in Shanghai and the Guangzhou-Shenzhen region, salaries are less than Beijing and more focused on “product design” (chanpin cehua).30 As I discuss below, while talent is frequently nurtured in foreign companies, workers are hard to retain because there is somewhere else to go and companies willing to pay more.31 Shaun Rein, author of The End of Copycat China, says:
A lot of Chinese feel like they can’t make it to the top of their organizations in multinationals. They’re moving to the Chinese private sector where companies have a lot of money for research and development. And there’s no bamboo ceiling. So why be the country head of R&D for 3M when you can be the global head of R&D for a private Chinese company? It’s happening that a lot of multinational companies now have to realize that their biggest competitors are often people that they trained directly over the past decade.32
This scenario now plays out in creative sectors. Many international companies have established offshoring operations, mostly in animation, software, design, and film production. In these environments, workers acquire knowledge capital through learning, sometimes in a “master-apprentice” system. Skills are molded by watching, listening and imitating, “learning-by-doing,” illustrating Polanyi’s tacit knowledge, essentially the acquisition of know-how. The foreign business introduces new ways of thinking about design while the locals provide cheaper labor. But cheap labor is not without transaction costs. Skills (and knowledge) are transferable, and many workers see no need to be loyal to the foreign master. When workers walk out the door, actual codified knowledge in the form of patterns (IP) might be lost. In the games industry, we observe a similar phenomenon.
In addition to a strong demand for graduates from China’s communication universities in the new media sectors, there is another side of development that allows us to further unbundle the nuances of the term precarious creativity in China: this is the commercialization of broadcasting industries. As mentioned above, prior to the 1990s, media production was regarded as a public service. The term generally used in this regard is public institution (shiye). The largest shiye is China Central Television (CCTV), a cultural mothership that for most observers of China symbolizes the hegemony of the state. Ying Zhu has eloquently described CCTV’s turn toward the market in Two Billion Eyes. The introduction of contract labor with higher pay, as opposed to ongoing employment, signaled a move toward the kinds of outsourcing practices common in most international media industries. When the broadcaster launched its flagship current affairs program, Oriental Horizon (Dongfang shikong), in the early 1990s, production work was contracted out, with less than 10 percent of workers remaining on CCTV’s payroll.33 The contracted workers became central to CCTV’s talent identification. As Ying Zhu notes, CCTV poached talent from independent production companies, thereby refreshing its workforce. On the other hand, as her study points out, many of the best people at CCTV have left and moved into independent and digital media sectors.
Similarly, at Beijing Television (BTV) competition is coming thick and fast from digital media. In a focus group interview I conducted with several senior personnel, one person commented, “Jobs were relatively high paid a decade ago, but staff members have received no pay increases for eight years.”34 The same person said that his department recently recruited about a dozen workers. The quality was high, many being overseas returnees with postgraduate qualifications and media experience. He noted, “The expansion of digital sectors is providing ‘talents’ with better working conditions and pay, and as a result there is a lot of mobile human capital. Mobility is accelerated and as a result the turnover rate is very high. This [competition] might force us to raise salaries in the future.”
At the time CCTV was renovating its operations in the 1990s, a number of leading production units emerged in television and film. Some of these, including Enlight Media,35 were formerly within the system; in time they would become leading players in the provision of entertainment content, particularly television formats and live events. A new “variant” of precarity soon came into being. Regulations allowing licensing of private companies came into force in 2004.36 With the exception of “foreigners,” any person or enterprise can form a media production company in China as long the State Administration of Press Publicity Radio, Film and TV (SAPPRFT) ratifies the license. The performance of non-state-owned production units, although precarious under conditions of censorship, has helped fulfill quotas of domestic content required to fill schedules, thus thwarting the incursion of foreign media content from the United States, Europe, Hong Kong, and Taiwan, which had spiked in the early 1990s.37 By the time China joined the World Trade Organization in 2001, the number of companies registered as private (mingying) had climbed to over three hundred, most of them plying their trade in TV serial drama production. By 2009, the number of independent production units in the broadcasting sector had exceeded four thousand, with 90 percent of drama production commissioned from such enterprises.38
While private companies have changed the game, it is necessary to add a caveat to the meaning of the term independent: that is, their existence is dependent on the dominance of the state in determining what content is suitable for audiences. Compared with the independent sector internationally, state-owned television stations maintain dominance in contract alliances; for instance, private production units might produce a show that is successful, but the rights are generally owned by the broadcaster. There are other uncertainties built in that make production precarious. Censorship is something that private entities need to be mindful of because production licenses are renewable. Moreover, when a program is successful, it might be replaced if the TV station decides to make its own version. This demonstrates the fragility of the concept of copyright in Chinese media industries. To maximize revenue and remain solvent, many production companies seek out and produce advertising content. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.05%3A_Media__The_Game_Changes.txt |
The question of how knowledge capital is circulated and deployed in generating successful cultural and media products leads me to reconsider the utility of the term precarious creativity. Rather than simply being a negative indicator, precarious creativity is adding to the knowledge capital of China’s creative industries. In other words, many workers are moving from low-cost production to higher-end production and from low wages to higher salaries. People are moving from traditional broadcasting to new media, where the salaries are better. In the process, some are identifying ways to innovate and internationalize rather than relying on the domestic market. To see how this plays out, I explore the cultural innovation timeline, a concept I have used elsewhere to explain the uneven development of China’s creative industries. Essentially the cultural innovation timeline depicts how production moves from low to higher value offerings. In describing these processes, it underscores the centrality of knowledge capital and the role of cultural intermediaries.
The base level of the cultural innovation timeline is standardized production; for instance, deterritorialized production and outsourcing of call centers gives low-cost locations an opportunity to be included in global trade networks. Factories are established because land is made available for cultural and creative industry projects, often with the help of local governments.39 Human capital in most of these instances is unskilled. Workers are paid low wages, and they work long hours. While there may be some status working in a design sweatshop or an animation outsourcing company, there is no desire to innovate; this is just a job. That is not to say, however, that there will be no learning on the job.
The second level of the timeline is imitation. Without the capacity to experiment or expend resources in content development, many producers of content follow the path of copycatting. China has a global reputation as a copy-nation, but again this needs to be put into the context of precarity. In the main, copying is a safe way of proceeding; if something has made money, it is reasonable to try it again, or tweak it a little bit. Shaun Rein says copying exists because you don’t have to pay high upfront costs. While an offender might get fined by the government, this is usually cheaper than paying for the rights.40 In addition to the economic dividends, there are cultural reasons for copy culture that would require a lengthy exposition.
In fashion we see clear evidence of this process. The fashion industry is populated by copyists; in responding to the question “How do you keep reinventing?” Ralph Lauren once said, “You copy. Forty-five years of copying. That’s why I’m here.”41 In China today, fashion and textile manufacture coexist in a symbiotic relationship. Workers toil to produce textiles while the country’s leaders exhort people to build an “innovative nation.” The worker on the Chinese-owned production line in Shaoxing, the capital of the textile industry in Zhejiang Province, is unlikely to be concerned with the idea of an innovative nation; work is a means to put food on the table, hopefully providing a stable income. This is the ultimate sweat industry: long hours, cramped conditions, and low wages. Workers produce capital through duplication; in this context creativity is redundant.
In many foreign-owned design workshops, creative capital is configured differently. According to Tim Lindgren, an independent designer who took his production to Shanghai several years ago because costs in Australia had escalated to the point where he could not employ staff, “It’s hard to keep high quality staff: they learn on the job and then leave to start their own business or find higher pay.”42 The mobility of workers in this situation is understandable. For the foreign design enterprise, it entails a search for staff replacements, not always so easy in a cross-cultural work environment. According to Lindgren, the other side of this dilemma is that patterns and designs (copyright) are lost, often appearing as high-priced garments in local markets.
Collaboration follows imitation, as content producers seek out alternative ways to capture value. Two examples of collaboration are pertinent to knowledge capital: film coproduction and TV formats. Coproductions in film are an interesting vehicle of knowledge capital transfer. They are categorized in three ways: the first is joint production (lianhe shezhi, or hepai), in which domestic and foreign parties make a joint investment of capital, services, or materials, and jointly share the benefits and risks of such “codevelopment.” The second is known as assisted production (xiezuo shezhi, or xiepai). This is where a foreign party makes an investment to produce in China: equipment, apparatus, sites, services, and so on are provided by the Chinese party, which receives a fee for services. The third model is entrusted production or commissioned production (weituo shezhi, or daipai). Here the Chinese party is “entrusted” by the foreign party to produce content in China.
Joint productions are considered domestic productions. In discussion with producers in Australia, I have observed a preference for codevelopment over assisted production—that is, producers are endeavoring to make stories that will sell in the Chinese market, taking advice on how to proceed from locals. There is a sense of optimism, some believing that Chinese audiences will learn to appreciate stories that have new ingredients. Codevelopment suggests that the Chinese side is interested in cocreating with foreign entities rather than just supplying low-cost production services. To date this road is littered with failure. It is precarious from a market sense as well as reputation. If producers edit their stories to appease the Chinese government, the danger is that critical acclaim in international filmmaking communities will diminish. Yet the road most traveled is likely to be documentaries about the wonders of China, not the treatment of its dissidents.
In effect, different risks exist depending on whether a project is a joint production or a commissioned one. East Asian businesses in many cases have a better appreciation of the precariousness of working with Chinese scripts; for instance, Chinese screenwriters and producers invariably have a well-developed sense of how to self-censor. Chinese content is imbued with allegory, parody, and oblique references, one of the reasons it encounters audience resistance when exported.
While coproductions have helped Chinese television improve its markets and have injected new ideas, China’s TV industry has struggled to export its brand. In television China’s comparative advantage in overseas sales comes from adaptations of the four classics of Chinese popular literature (sida mingzhu),43 as well as historical serials about emperors, eunuchs, and court intrigues. This advantage has conspired to produce a glut of second-rate productions; even home audiences have turned away in large numbers, precipitating edicts from the SAPPRFT to rebalance production slates toward contemporary stories. In short, economic success in the home market does not equate to success abroad, and critical success abroad (as in the case of art house cinema) does not necessarily translate into economic success at home.
In the case of TV formats, a transaction is made between the domestic licensee and the format holder or distributor. According to the managing director of a leading international TV format distributor, Chinese television stations generally want more knowledge than other international partners. The TV format distributor in question not only sells the copyright of the TV format to the local TV station or production company but also provides direction on program localization, consulting on production, and even direct participation during the production and postproduction periods. This represents a difference from the format licensing business elsewhere; in China the foreign party is regarded like a consulting firm.44 The Chinese want the know-how.
Following collaboration, the fourth level in the cultural innovation timeline is cultural trade. By the first decade of the millennium, the impetus to move programs out of China into international regional markets had increased. Coincident with the cultural trade impetus was the consolidation of production and management, first in media conglomerates and later in media bases or clusters (the late 1990s, early 2000s). This consolidation constitutes level five of the timeline. I have earlier mentioned the virtues of clustering and the spillover effects that can accrue. Of course, this is the ideal. Clustering has a checkered history in China’s reform period. In many instances, clusters function to attract business investment (zhao shang) more than attracting creative talent (zhao chuang).45 Moreover, the fact that China has hundreds of clusters in which outsourcing is the bread and butter indicates that there is still a market for semiskilled labor.46 The massive injection of government funding in clusters is driving competition for talent and investment. Places are competing for creative talent and investment on a scale unprecedented in China, with local governments providing preferential policies, tax breaks, and free rent. In effect, precarity has a broader context in China. While employment is volatile and profits are uneven across the broad spectrum of commercial creative industries, the high level of government subsidy in constructing zones and parks has created a bubble.
In previous work on cultural and creative industries, I have described level six as constituting “peer communities” and “creative communities.”47 This is the online world, with more than 650 million participants. The activities of online producers are indeed precarious, so much so that much content is predominantly parody. It is posted, reposted, and then taken down, often by persons employed by the government to monitor unhealthy commentary. It is this meaning of precarious creativity, I feel, that characterizes China more than debates about job losses in the “creative industries,” which are invariably construed as symptomatic of neoliberalism, a move that arguably succumbs to its own kind of reductionism.48 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.06%3A_The_Cultural_Innovation_Timeline_Reconsidered.txt |
A great deal of government investment has targeted the cultural sector in China over the past decade. As a result of market openings, many foreign players are lining up to take advantage of the “world’s biggest audience.”49 The technological gap between China and the developed economies is closing fast because of the transfer of knowledge and the movement of human capital.
I have argued that the concept of precarious creativity requires rethinking if it is to apply to China. Of course, the conventional usage of precarity as depicted in much of the literature does apply, especially in manufacturing sectors, where sweatshops operate with impunity. Much work in the creative industries in China is project-based, and we observe a marketplace for talent. In this latter sense, the key point is the mobility of workers in and across media sectors, and from foreign companies back to Chinese digital companies. While many foreign companies are struggling to retain talented workers, the new media challenge is significant, extending to state-owned media enterprises. As suggested by CCTV’s poaching of talent from the independent production sector, even though the work may not be long-lasting, skills are in demand. Conversely, the example of BTV shows that digital media, from games to mobile media, is bent on securing the best “talent” and paying more money.
Despite the massive market for culture in China, government regulation undermines attempts to be taken seriously internationally as a soft power competitor. When competing for the hearts and minds of international audiences, two main challenges confront creators of film, television, and animation content. There are other elements of precarity that stymie China’s outward-bound ambitions. The first is the challenge of credibility. An emphasis on historical revisionism and a propensity toward melodrama, while acceptable in the PRC market, fail to transfer into commercial success abroad. This in turn points to a second problem. There is a lack of understanding within China of how to make content that might be successful overseas and actually assist in reinvigorating “brand China.”50 Hence the demand for foreign know-how.
Another ubiquitous aspect of precarious creativity is really the “elephant in the room.” What is the point of talking meaningfully about creativity in China if its existence is made perilous by censorship? In this context I want to add another dimension to our understanding of knowledge capital, namely “knowing-to.” Whereas knowing-that and knowing-how provide ways to ascend the cultural innovation timeline, knowing-to comes into play at important times; for instance, a person might wish to push the boundaries of creative work, or a foreign producer might seek to promote a film coproduction in China. Knowing-to becomes an important modality of knowledge capital. Knowing-to manifests in four circumstances: first, anticipating outcomes (understanding the effects of an action or a policy); second, timeliness (making one’s move at the right time); third, context (working in a way that takes account of others’ political obligations and guanxi);51 and fourth, “understanding weightiness” (knowing the relative weight of policies and regulations).52 Knowing-to combined with knowing-that makes for good business in China. However, this does not guarantee good content, just survival.
Finally, it is worth considering how knowing-to applies to the Chinese leadership’s attempt to rebrand China as a “strong cultural power” (wenhua qiangguo), the latest rhetoric emanating from cultural industry think tanks. In China, precarity extends beyond employment; if someone expresses a view in writing that directly challenges the government or infers that a member of the Chinese political elite is corrupt, this person’s employment may be terminated—and this may have consequences for personal liberty and the welfare of the person’s family. The European Enlightenment view that creativity is about asking difficult questions, challenging authority, and destabilizing norms does not sit well with the government. The discourse of creativity is based on a harmonious vision of progress, captured in the soporific idea of a Chinese Dream, one in which all Chinese citizens are presumed to participate. That means 1.3 billion Chinese dreams. The problem in this rhetoric is that dreaming by definition is difficult to control. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.07%3A_Concluding_Remarks__The_%28Precarious%29_Elephant_in.txt |
I acknowledge the support of the Australian Research Council in enabling this research to be undertaken. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research for this paper was funded through the Australian Research Council Discovery-Projects DP140101643 Willing Collaborators: Negotiating Change in East Asian Media Production.
1 I use the term developed nation or economy rather than West or Western. A developed economy in this sense is where tertiary and quaternary sectors dominate the economy.
2 Robert Baldwin, “Globalisation: The Great Unbundling(s),” Economic Council of Finland, 2006, available at www.tinyurl.com/2ol2n8.
3 For instance, see G. Hearn, R. Bridgstock, B. Goldsmith, and J. Rodgers, eds., Creative Work beyond the Creative Industries (Cheltenham: Edward Elgar, 2014).
4 See Nestor Gabriel Canclini, “Precarious Creativity: Youth in a Post-Industrial Culture,” Journal of Latin American Cultural Studies: Travesia 22.4 (2013): 341–352.
5 Boosterist claims of the creative economy proliferate in most countries. Much of this relates to errors in accounting, namely the propensity to count as creative things that are evidently not creative. The first book to advance the cause was John Howkins’s The Creative Economy: How People Make Money from Ideas.
6 For instance, facilities maintenance, hospitality, domestic work. See Terry Flew, The Creative Industries: Culture and Policy (London: Sage 2012), 107.
7 From 251 valid surveys (from a total of 400) in Beijing (Fangjia 46, Shijingshan Cyber Recreation Park), Suzhou Industrial Park and Creative 100 (Qingdao), conducted May 2009 to July 2010. Michael Keane, China’s New Creative Clusters: Governance, Human Capital and Investment (London: Routledge 2011).
8 See Seamus Grimes, “Foreign R&D in China: An Evolving Innovation Landscape,” in Innovation and Intellectual Property in China: Strategies, Contexts and Challenges, ed. K. Shao and X. Fend (Cheltenham: Edward Elgar, 2014), 186–205.
9 AnnaLee Saxenian, The New Argonauts: Regional Advantage in a Global Economy (Cambridge, MA: Harvard University Press, 2006), 201.
10 The term cultural soft power in China is used to refer to China’s attempts to move its cultural and media products into international markets. See Michael Keane, Creative Industries in China: Art, Design, Media (London: Polity 2013); and Michael Keane, China’s Television Industry (London: BFI Palgrave, 2015).
11 Gilbert Ryle, The Concept of Mind (Chicago: University of Chicago Press, 1949).
12 Ibid., 45.
13 For a discussion of soft power competition regionally, see Beng-Huat Chua, Structure, Audience and Soft Power in East Asian Culture (Hong Kong: Hong Kong University Press, 2012).
14 For a discussion, see Jason Potts and Tarecq Shehadeh, “Compensating Differentials in the Creative Industries: Some Evidence from HILDA,” in Creative Work beyond the Creative Industries, ed. G. Hearn, R. Bridgstock, B. Goldsmith, and J. Rodgers (Cheltenham: Edward Elgar, 2014).
15 McKinsey Global Institute, “Preparing for China’s Urban Billion” (McKinsey and Company, 2009).
16 The One Child Policy refers to the policy adopted in 1978 that mandated that each family is allowed one child with the exception of minorities. The policy has undergone revision in the past few years, allowing people who were single children to marry and have two children.
17 The hukou refers to the household registration scheme initiated in the 1950s to maintain population control. It is essentially a work permit.
18 For a discussion, see Keane, Creative Industries in China.
19 The word industrialization is translated as gongyehua.
20 Martin King Whyte, “The Paradox of Rural-Urban Inequality in Contemporary China,” in One Country, Two Societies: Rural-Urban Inequality in Contemporary China, ed. M.K. Whyte (Cambridge, MA: Harvard University Press, 2010), 1–28.
21 Michael Keane, China’s New Creative Clusters: Governance, Human Capital and Investment (London: Routledge, 2011).
22 Loretta Napoleoni, Maonomics: Why Chinese Communists Make Better Capitalists Than We Do, trans. Stephen Twilley (New York: Seven Stories Press, 2011), 41.
23 See Juncheng Dai, Shengyi Zhou, Michael Keane, and Qian Huang, “Mobility of the Creative Class: A Case Study of Chinese Animation Workers,” Eurasian Geography and Economics 53.5 (2012): 649–670.
24 Mark Banks and David Hesmondhalgh, “Looking for Work in Creative Industries Policy,” International Journal of Cultural Policy 15.4 (2009): 515–430; David Hesmondhalgh, David, The Cultural Industries, 2nd ed. (London: Sage, 2007); Andrew Ross, Fast Boat to China: Corporate Flight and the Consequences of Free Trade: Lessons from Shanghai (New York: Pantheon Books, 2006); Kate Oakley, “In Its Own Image: New Labour and the Cultural Workforce,” Cultural Trends 20.3–4 (2012): 281–289.
25 For instance, see the other essays in this volume.
26 Alan Burton-Jones, Knowledge Capitalism: Business, Work, and Learning in the New Economy (Oxford: Oxford University Press, 1999).
27 For a discussion of media capital, see Michael Curtin, Playing to the World’s Biggest Audience: The Globalization of Chinese Film and TV (Berkeley: University of California Press, 2007). For Silicon Valley, see Martin Kenney, ed., Understanding Silicon Valley: The Anatomy of an Entrepreneurial Region (Stanford, CA: Stanford University Press, 2000).
28 From National Bureau of Statistics and China Statistical Yearbook, Beijing.
29 See www.zhihu.com/question/19588383.
30 See www.zhuayoukong.com/122808.html.
31 Dai Juncheng, Zhou Shangyi, Michael Keane, and Qian Huang, ‘Mobility of the Creative Class and City Attractiveness: A Case Study of Chinese Animation Workers,” Eurasian Geography and Economics 53.4 (2012): 649–670.
32 Interview with Shaun Rein, September 15, 2014, available at http://www.creativetransformations.a...chinese-dream/.
33 Ying Zhu, Two Billion Eyes: The Story of China Central Television (New York: Free Press, 2012).
34 Focus group discussion with representatives of Beijing Media Group, including production managers, marketing, and programmers, QUT, Brisbane, September 25, 2014.
35 Enlight media was formed in 1998 by Wang Changtian, a former producer at Beijing Television. For a discussion, see Yuezhi Zhao, Communication in China: Political Economy, Power and Conflict (Lanham, MD: Rowman & Littlefield, 2008).
36 Michael Keane and Bonnie Rui Liu, “China’s New Creative Strategy: Cultural Soft Power and New Markets,” in Asian Popular Culture: The Global Cultural (Dis)connection, ed. Anthony Fung (London: Routledge, 2013), 233–249.
37 Keane, China’s Television Industry.
38 X. Yingdan, “90% from the Private Production in China’s TV Drama Market,” Xinhua Daily, May 28, 2007, www.ccmedu.com.bbs33_45123.html.
39 See Keane, China’s New Creative Clusters.
40 Interview with Shaun Rein, September 16, 2014.
41 Eric Wilson, “O and RL: Monograms Meet,” New York Times, October 25, 2011, www.nytimes.com/2011/10/27/fashion/oprah-winfrey-interviews-ralph-lauren.html?_r=1&.
42 Interview with Tim Lindgren, September 12, 2014.
43 The Dream of the Red Chamber (hong lou meng), The Journey to the West (xiyouji), Outlaws of the Marsh (shuihu zhuan), and Romance of the Three Kingdoms (sanguo yanyi).
44 Interview, Beijing, August 25, 2014.
45 Keane, China’s New Creative Clusters.
46 See ibid.
47 Keane, Creative Industries in China.
48 I have elsewhere argued that the concept of neoliberalism is problematic and is not applicable to China. See, for example, Keane, The Chinese Television Industry.
49 Curtin, Playing to the World’s Biggest Audience.
50 See the discussion by Yingchi Chu regarding horizons of expectation. Yingchi Chu, “The Politics of Reception: ‘Made in China’ and Western Critique,” International Journal of Cultural Studies 17.2 (2014): 159–173.
51 Guanxi is usually translated as “personal relationships”; these may have political implications.
52 For a discussion of knowing-to in traditional Chinese philosophy, see Stephen Hetherington and Karyn L. Lai, “Knowing-How and Knowing-To,” in The Philosophical Challenge from China, ed. Bryan Bruya (Cambridge, MA: MIT Press, 2015), 279–301. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/16%3A_Unbundling_Precarious_Creativity_in_China__Knowing-How_and_Knowing-To/16.08%3A_Notes.txt |
This chapter elaborates the concept of revolutionary creative labor. The Arab uprisings, particularly the conflict in Syria, have given rise to a notion of creative resistance. Various activists, journalists, academics, and curators have used that phrase to celebrate a gamut of expressive practices and forms encompassing graffiti, digital memes and mash-ups, handheld banners, political rap, and others.¹ The wording combines two terms with overwhelmingly positive connotations that evoke human ingenuity and agency.
17: Revolutionary Creative Labor
This chapter elaborates the concept of revolutionary creative labor. The Arab uprisings, particularly the conflict in Syria, have given rise to a notion of creative resistance. Various activists, journalists, academics, and curators have used that phrase to celebrate a gamut of expressive practices and forms encompassing graffiti, digital memes and mash-ups, handheld banners, political rap, and others.1 The wording combines two terms with overwhelmingly positive connotations that evoke human ingenuity and agency. But if creative resistance is to convey anything beyond a nebulous concept of ingenious rebellion, it needs to be systematically explored and situated vis-à-vis notions of activism, creativity, and labor in cultural production. One way to achieve that goal is to theorize processes of artful dissent as revolutionary creative labor.2
In order to develop a working definition of revolutionary creative labor, this chapter draws on a study of the body and activism in the Arab uprisings based on primary materials, most collected in 2011 and 2012.3 In this chapter I pursue the following questions: To what extent does the extreme duress of revolution shift our understanding of creative labor? Is revolutionary creative labor different from other kinds of creative labor? What does revolution add to our understanding of creativity and precarity in cultural production? To answer these questions, I engage with a few key texts. The chapter first zeroes in on the use of creativity in social movement theory, mainly in James Jasper’s The Art of Moral Protest.4 Then it reviews some work in media industries research that addresses precarity and creativity, namely Vicki Mayer’s Below the Line.5 A comparative analysis of “industrial”’ and “revolutionary” forms of creative labor follows. Finally, via brief references to the magisterial compendium provided by Hans Joas in The Creativity of Action 6 and to Lazzarato’s theory of immaterial labor,7 the chapter concludes with a theoretical elaboration of revolutionary creative labor.
17.02: Creativity and Labor in Social Movement and Production Studies A Snapshot
Social movement theorists have rarely discussed activism in terms of creativity or labor. Though creativity is sometimes mentioned in its prosaic meaning and the word occasionally appears in titles of books on social movements, rarely is it systematically theorized or critiqued as a conceptual category.8 Jasper’s The Art of Moral Protest comes closest to a sustained conceptual treatment of creativity: the notion of artfulness is a cornerstone of the book’s “cultural” approach to protest, which intends “to increase [the focus on] explanatory factors . . . to concentrate on mechanisms, not grand theories . . . to give the voice back to the protestors we study.”9 Jasper writes: “Protest movements work at the edge of a society’s understanding of itself and its surroundings. Like artists, they take inchoate intuitions and put flesh on them, formulating and elaborating them so that they can be debated. Without them, we would have only the inventions of corporations and state agencies, products and technologies created to enhance efficiency or profitability.” Jasper then concludes: “In order to understand these innovations, we need ‘moral innovators’ too: the artists, religious figures, and protestors who help us understand what we feel about new technologies.”10 By comparing activists to artists, Jasper anchors artfulness in the socio-political realm of activism, valorizing innovation not in its potential for commodification but for its ability to generate political-rhetorical value.11
For Jasper, artfulness refers to “experimental efforts to transmute existing traditions into new creations by problematizing elements that have been taken for granted.”12 Artfulness articulates biography and culture: beginning as individual creativity, it becomes strategic once shaped by a group, and subsequently it is enacted in protest. Examples include deploying widely familiar and emotionally evocative symbols and grafting new meanings onto existing symbols. Language is a primary vehicle through which activists project, manipulate, and redefine symbols. Having elsewhere in the book compared activists to artists, Jasper writes that “at the most extreme, ideologists operate as poets; they define emerging structures of feeling with new terms and images.”13 Invoking the “immense value we place on individual creativity,”14 Jasper employs the notion of “tactical innovation,” a mainstay in the social movements literature, which emerges at “the interplay of protest groups and their opponents.”15
Unlike studies of activism, research on cultural production does not focus on Political aspects of labor.16 But the two are alike in rarely grappling directly with creativity as a central conceptual category.17 One exception is Vicki Mayer’s study of workers in a television set factory in Manaus, Brazil, where the author endeavors to “deconstruct our received notions of creativity and to reconstruct a notion of creative action that is both social and individual in the practices of assembling.”18 Following an argument made by Joas and others that social context is key to understanding creativity, Mayer develops notions of creativity that “conjoin the interiority of mental labor with the exteriority of a world that enables its articulation.”19 In addition to emphasizing creativity’s social dimension, Mayer shows that as a discourse creativity is deployed with discrimination for purposes of social distinction and control. But it is Mayer’s discussion of creativity as a process of making do under structural constraints that is most relevant for my purposes, because it leads to two questions that are central to this chapter. What differences can we discern between deployments of “creativity” in media industries research and the trope of “creative resistance” used to describe some forms of dissent in the Arab uprisings? And how do these differences enable my elaboration of revolutionary creative labor?
“Creativity” is a strategic and discriminatory trope. It is strategic because its selective deployment reflects and perpetuates relations of politico-economic power. It is discriminatory because it is applied according to rules of exclusion and inclusion that serve criteria of social distinction. Considerations of power and distinction in creative labor differ between scholarship on media industries and research on Political forms of labor, such as activism and propaganda. In the television set factory Mayer studied, the discourse of creativity is reserved to operators in higher ranks of the industry, who exclude workers on the assembly line from creativity’s definitional scope. As Miller has shown, proponents of “creativity” have stretched the term to encompass most ways in which any activity that could remotely be described as cultural is monetized.20 In contrast, the creative resistance trope operates primarily according to political and ideological imperatives. Creative resistance refers to propaganda by people we like—in this sense creative resistance is a more glamorous, bottom-up cousin of the great euphemism public diplomacy. During the war between Israel and Lebanon in 2006, Hezbollah launched a range of stylistically bold, visually compelling propaganda videos, some aimed at mobilizing supporters, others psyops clips, many in Hebrew, aimed at demoralizing Israeli soldiers. Though the notion of resistance is central to Hezbollah’s raison d’être, and though many of the videos were rhetorically sophisticated and aesthetically slick, to my knowledge no one called these “creative resistance.” Most mainstream media coverage in the West referred to them as “propaganda,” though in some aspects they resemble revolutionary videos of the Arab uprisings, and some of them even resemble U.S. Army recruitment commercials. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/17%3A_Revolutionary_Creative_Labor/17.01%3A_Introduction.txt |
As a mercurial term that is applied at once broadly (connoting a vast and varied semantic field) and selectively (according to considerations of political power and social distinction), creativity requires definitional work to be analytically useful. In this chapter I am not interested in developing a full-scale analytical parsing of creativity’s various possible definitions and applications. I am, however, keen on discerning differences between the kind of creativity that one sees in, say, a television studio or factory floor—industrial creative labor—and the kind of creativity manifest in revolutionary creative labor. What might some of these differences be?
One must begin with the rather obvious observation that the creative labor of Egyptian, Syrian, and Tunisian revolutionaries is more confrontational than the invisible, sanctioned, unsanctioned, and even subversive types of creativity that Mayer identifies on the Manaus factory floor. Manifestations of creative labor in the Arab uprisings are not flexible, reformist, or merely subversive: spawned under life-threatening conditions, they are radical rejectionist expressions of human affects and aspirations. Rather than trying to find ways to survive or thrive in the factory, revolutionaries seek to burn the factory down, clean the debris, and build a new and utterly different edifice. This is the first and most crucial difference between industrial and revolutionary creative labor.
The centrality of the human body is a second difference between industrial and revolutionary creative labor. Though concern with the body is not vital to most research on media industries, Mayer does grapple with corporeality as an important aspect of workers’ experience, what she calls “the corporeal achievement of assembly,” and she argues that “conditioning the body to do the physical work signified an important rite of passage in the social world of the factory.”21 Assembly workers regiment their bodies in new and uncomfortable ways with the purpose of increasing productivity. Nonetheless, “the corporeality of the act of assembling the television set could not communicate a creative act in itself simply because of its exclusion from the discourse of creativity.”22 In contrast, revolutionary creative labor, I would argue, is more deeply and more intimately entangled with the human body. This is primarily a matter of resources: factory workers are provided with the tools needed to satisfy the demands of capitalist production. Revolutionaries, in contrast, are often bereft of tools and resort to very basic media. The Syrian Masasit Mati collective, which created the famous Top Goon video series lampooning Bashar al-Assad, used paper, wood, and fabric to create finger puppets and human energy to operate the puppets. Using basic materials, they miniaturized the dictator by reducing him to a finger puppet and infantilized him through satire.23 Of course, they also had a basic video camera and eventually set up a YouTube channel, but rather than being provided by “the system,” these resources (most from the seventeenth century, some from the twentieth and twenty-first) were snatched “behind the back” of the dictator to express derision of his person and rejection of his rule.
This brings us to the third divergence. In the television set factory in Manaus, assembly-line workers are subjected to a range of managerial constraints that Mayer groups under Taylorism, “parsing complex jobs into tasks,”24 and Japanization, which consists of a gamut of “social surveillance techniques.”25 Working in tandem and sometimes in contradiction, these two top-down forces constrain workers as they create opportunities to overcome constraints. In Mayer’s words, “Assemblers looked creatively for solutions to stressful limits because they had no other choice. . . . Yet workers’ creativity could also overstep expectations, leading to disciplinary actions, dismissal, or even blacklisting.”26 In contrast, revolutionary creative labor is situated farther down the sanctioned–unsanctioned creativity that Mayer evokes in her analysis. Assembly workers’ creativity is what I would call “making-do” creativity, whereas creative insurgency involves “breaking-bad” creativity.27 The first is conjured up to cope with the system; the second is deployed to topple the system. The first is framed by top-down industrial-managerial models; the second is a bottom-up expression of pent-up repressed subjectivity. The former involves bodily discipline—“The adaptation of her fingers to the fine manipulations of wires was an acquired skill”28—on the factory floor, while the second entails bodily insurrection on a literal and symbolic battlefield. In the first, Mayer points out, “unsanctioned creative actions generally stimulated more rules.”29 Whereas factory workers bent their fingers to the demands of capital, members of Masasit Mati moved puppets’ fingers to utterly reject the Syrian dictatorship. The first is adaptation; the second, rebellion.
Whereas assembly workers face managerial (and social) constraints, Arab creative activists confront often brutal and sometimes murderous repression, which grows increasingly violent as uprisings endure. If Brazilian assembly workers focus their creativity on “eking out a living,”30 Arab revolutionaries deploy creativity for the purpose of eking out a dignity, a political agency. Prerevolutionary creative dissent in countries like Egypt, Syria, and Tunisia—double-entendre parodies, strategically ambivalent artwork, and allegorical theater—can be described as subversive. In contrast, revolutionary creativity is a confrontational, no-holds-barred, high-stakes, high-risk, and potentially high-rewards gambit.
Industrial creative labor and revolutionary creative labor differ in a fourth way. Whereas the former occurs openly, the latter operates surreptitiously. In both cases, the visibility of creative labor is determined by the structural constraints already discussed. Though factory floor workers may engage in micropractices of subversion to improve their lives in the factory, they are subjected to a strong surveillance regime, and the lion’s share of their labor is exceedingly visible to their managers. But if in the factory “absences were treated as the worst infractions,”31 absence from the revolutionary public sphere constitutes an ideal situation for incumbent dictators—presence and visibility invite immediate repression. As a result, though security apparatuses attempt to spy on and capture activists, revolutionary creative labor must occur underground and be physically peripatetic to avoid arrest. In addition to resources, then, revolutionary creative labor’s “trajectories of creative migration,” as Michael Curtin called creative labor’s movement across national boundaries,32 are motivated primarily by the desire to physically stay alive, rather than by economic survival. Many Syrian revolutionary artists now live in Beirut or Berlin, and several prominent Arab uprising activists are political refugees in Europe.
A fifth and final difference between industrial and revolutionary creative labor is that the former is remunerated, however unfairly, while the latter is unwaged labor.33 I list this difference in fifth place rather than earlier in the list because this contrast is not as extreme as it may appear. Though the creative labor of most activists in the Arab uprisings remained unrecognized and unwaged, there have been several exceptions reflecting the commercial and political co-optation of revolutionary creative labor. The Egyptian surgeon turned late-night comedian, Bassem Youssef, the so-called Egyptian Jon Stewart, started his show on YouTube during the Egyptian revolution. In time, one television channel picked up the show, then a bigger channel acquired it, to considerable commercial success and global critical praise. Subsequently, the show was streamed by the Arabic-language channel of the German broadcaster Deutsche Welle, before being shut down after the military coup of Abdelfattah El-Sisi in June 2013.34 Youssef, already an affluent medical doctor, was one of a few revolutionary creative laborers who moved from unpaid to highly waged labor. The finger puppeteers of Masasit Mati, in contrast, tried crowdfunding their second season via Kickstarter, and when that effort failed, they received a grant from the Prince Claus Fund in the Netherlands. In effect, they leveraged their fame into financial support and official recognition from prestigious Western institutions, even if technically that does not constitute waged labor. But disagreements within the group led to its dissolution. Despite momentary success, then, revolutionary creative labor’s mainstream prospects are as precarious as revolutionaries’ ambitions for political rule.35 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/17%3A_Revolutionary_Creative_Labor/17.03%3A_Industrial_and_Revolutionary__Two_Types_of_Creative_Labor.txt |
This chapter has been grappling with the extent to which different contextual environments and constraints generate different types of creative labor with different levels of precarity. From the preceding critical comparison of what I called industrial and revolutionary creative labor, we can conclude that the extreme strictures of revolutionary contexts lead to a specific relation between the individual and the social. In The Creativity of Action, Joas singles out three metaphors, which emerged between 1750 and 1850, that are central to creative action: expression, from the work of Johann Gottfried Herder; and production and revolution, both elaborated by Karl Marx. Each of these metaphors, Joas argues, “represents an attempt to anchor human creativity in at least one of the three ways of relating to the world. The idea of expression circumscribes creativity primarily in relation to the subjective world of the actor.” In contrast, “the idea of production relates creativity to the objective world, the world of material objects that are the conditions and means of actions.” “And finally,” Joas concludes, “the idea of revolution assumes that there is a potential of human creativity relative to the social world, namely that we can fundamentally reorganize the social institutions that govern human coexistence.”36
Revolutionary creative labor, I conclude, entails the convergence of expression, production, and revolution. Revolutionary contexts are characterized by total upheaval—social and political but also economic and cultural—in which everything is up for grabs. These contexts of tremendous flux and peril require a total expenditure of resources, calling on people to mobilize to enact subjective and objective changes to the world they live in.
The definitional field delineated by expression, production, and revolution encompasses familiar axes of tension: the individual versus the social, the ideational against the material, the reformist in contrast to the radical. Such a field is a particularly apt space to grapple with the revolutionary creative labor emerging in the Arab uprisings. If, as Joas and Mayer argue, creativity entails coordinating a variety of means, responding to incentives, and working within constraints, and if, as I have already argued, revolutionaries respond to specific motivations and work within strictures distinct from the constraints of the factory floor (or, for that matter, the production studio), then revolutionary creative labor is indeed a distinct kind of creative labor.37
Revolutionary creative labor contributes to the creation of a subjectivity that is radically different from that of industrial labor. Jasper noted that artists can “generate and regenerate the very subjectivity they pretend only to display.”38 This echoes Lazzarato’s argument about immaterial labor, which “presupposes and results in an enlargement of productive cooperation that even includes the production and reproduction of communication and hence its most important content: subjectivity.”39 Whereas Lazzarato argues that immaterial labor changes the relationship between producer and consumer, it is productive to think of revolutionary creative labor as changing the relationship between ruler and ruled. One important aspect of Lazzarato’s thesis is that the shift from manual to immaterial labor transforms the three elements of what he calls the aesthetic model of labor—author, reproduction, and reception—by emphasizing their social rather than individual aspects. Creativity, Lazzarato concludes by way of brief mentions of Simmel’s work on intellectual labor and Bakhtin’s focus on social creativity, is social rather than individual, a point also made by Joas and Mayer.
Ordinary people from among the hitherto ruled, having become revolutionary activists, enact revolutionary creative labor to get rid of the ruler. Revolutionary creative labor, then, occasions a shift in subjectivity from the atomized docility of subjects under dictatorship to the collective rebellion of politicized agents in revolution. In Foucauldian terms, we can describe revolutionary creative labor as a technology of revolutionary selfhood. It mobilizes expressive and affective resources alongside the material resources of “noncreative” revolutionary labor— demonstrating in the street, staffing barricades, confronting security personnel, wielding sticks, shooting guns, tending to the wounded—to effect fundamental and political change.
The body is crucial to the project of revolutionary selfhood. As I have argued elsewhere 40 (though without grappling with the conceptual minutiae of creativity and labor), the body—as instrument, metaphor, symbol, medium—is central to revolutionary creative labor. Mayer explains how creativity pertains to Joas’s concept of a “situation,” by which he means “the ability of the body to move and communicate in an innovative way. . . . [C]reativity must be enacted through both the body and the social system of meanings that recognizes the action as different from the norm. . . . Creative action unifies the mind and body in doing something perceived as different. . . . This means that thought must be materialized, but also that the material is cause for later reflection.”41
But in revolutionary contexts of the twenty-first century, the body must be understood as a central and agentive node among a panoply of other media—from cardboard to digital video—that are harnessed by revolutionaries in an all-out campaign to change their lives. The body, then, must be understood as the animator of what I elsewhere called “hypermedia space,” a space of signification with multiple points of access created by interconnections among various media platforms.42 In the case of the Arab uprisings, these include media that can be characterized as mainstream (television, newspapers), new (mobile devices, social media), and old (puppetry, graffiti), alongside the oldest of them all, the human body, which operates all other media.
Revolutionary creative labor, then, is an embodied, extremely precarious practice unfolding in a life-or-death situation, one among several kinds of labor (from physical struggle to mainstream media production) that challenge authoritarian leaders. Whereas, as Mayer argues, assembly-line work is a kind of creative labor that should to be situated within the broader context of media creativity, a different kind of creativity is at work in what I defined and explicated in this chapter as revolutionary creative labor. Indeed, a final distinction can be made between forms of creative labor that are embedded in localized contexts (the factory) which are otherwise not creative (the assembly line), what in this chapter I called industrial creative labor, and revolutionary creative labor, which consists of explicit and self-conscious forms of revolutionary creativity that are intended to be launched into broader trajectories of circulation. By enacting contextually new forms of political subjectivity and directing them at radical change, revolutionary creative labor seeks to find, congeal, and mobilize publics. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/17%3A_Revolutionary_Creative_Labor/17.04%3A_Subjectivity_and_Revolutionary_Creative_Labor.txt |
Vicki Mayer—who alerted me to the work of Hans Joas—Michael Curtin, and Toby Miller have been key interlocutors on issues related to this chapter. I also thank Katerina Girginova for research assistance on creativity, Michael Curtin and Kevin Sanson for their useful feedback on the first draft of this chapter, and Marina Krikorian for editorial help.
1 Media stories and academic publications celebrating Arab revolutionary rap and graffiti have become so commonplace that we could talk of a Revolutionary Graffiti Index or an Arab Rap Index, following what Miller, referring to Richard Florida’s work, calls the Technological and Gay Indexes (Toby Miller, “A View from a Fossil: The New Economy, Creativity and Consumption—Two or Three Things I Don’t Believe In,” International Journal of Cultural Studies 7.1 [2004]: 60).
2 Clearly, this is only a small part of revolutionary labor at large, which includes demonstrating, confronting policy and security personnel, building barricades, feeding revolutionaries, tending to the wounded, and so on.
3 Marwan M. Kraidy, The Naked Blogger of Cairo: Creative Insurgency in the Arab World (Cambridge, MA: Harvard University Press, 2016).
4 James M. Jasper, The Art of Moral Protest: Culture, Biography and Creativity in Social Movements (Chicago: University of Chicago Press, 1997).
5 Vicki Mayer, Below the Line: Producers and Production Studies in the New Television Economy (Durham, NC: Duke University Press, 2011).
6 Hans Joas, The Creativity of Action (Chicago: University of Chicago Press, 1996).
7 Maurizio Lazzarato, “Immaterial Labour,” in Radical Thought in Italy: A Potential Politics, ed. Paolo Virno and Michael Hardt (Minneapolis: University of Minnesota Press, 1996), 133–147.
8 For example, Benjamin Shephard, Play, Creativity and Social Movement (New York: Routledge, 2011), pivots around the notion of play; while Glenda Ballantyne, Creativity and Critique: Subjectivity and Agency in Touraine and Ricoeur (Leiden: Brill, 2007), focuses on subjectivity and agency. The “culture-jamming” literature does not apply in this context because it concerns relatively low-risk subversion of consumer culture in relatively stable, relatively democratic, industrialized countries.
9 Jasper, Art of Moral Protest, 378–379. Jasper identifies four basic dimensions that artful protesters use: resources like technology and money; strategies, individual and group tactics; culture, shared aspects of mental worlds and their physical representations; and biography, individuals’ mental worlds, conscious and subconscious.
10 Ibid., 375.
11 Jasper uses creativity and innovation somewhat interchangeably, though definitional differences emerge in his discussion. The sociologist Doug McAdam has done extensive work on tactical innovation. For a summary introduction, see Doug McAdam, “Tactical Interaction and Innovation,” The Wiley-Blackwell Encyclopedia of Social and Political Movements, ed. David A. Snow et al. (Hoboken: Wiley-Blackwell, 2013).
12 Jasper, Art of Moral Protest, 65.
13 Ibid., 159.
14 Ibid., 219.
15 Ibid., 99. For an argument about upbringing and social support as key biographical enablers of creativity in famously “creative” people, see Howard Gardner, Creating Minds: An Anatomy of Creativity Seen through the Lives of Freud, Einstein, Picasso, Stravinsky, Eliot, Graham, and Gandhi (New York: Basic Books, 2011).
16 Political with a capital P connotes issues of state power and resistance to it, as opposed to cultural politics.
17 Media, cultural, and music production scholars have addressed labor issues, though mostly focusing on the exploitation of labor by those industries. See Toby Miller et al., Global Hollywood: Issue 2 (London: British Film Institute, 2004); Mark Andrejevic, Reality TV: The Work of Being Watched (Lanham, MD: Rowman and Littlefield, 2004); Matt Stahl, Unfree Masters: Recording Artists and the Politics of Work (Durham, NC: Duke University Press, 2013). The emerging literature on digital labor also focuses on the increasingly exploitative nature of capitalism; see Christian Fuchs, Digital Labour and Karl Marx (London: Routledge, 2014). This literature’s socio-economic focus is helpful, but only indirectly, for studying a revolutionary setting.
18 Mayer, Below the Line, 33.
19 Ibid., 32.
20 Miller, “A View from a Fossil.”
21 Mayer, Below the Line, 43.
22 Ibid.
23 See their YouTube channel at www.youtube.com/user/MasasitMati.
24 Mayer, Below the Line, 44.
25 Ibid.
26 Ibid., 47–51.
27 Jasper’s definition of creativity as an “extreme form of flexibility” (The Art of Moral Protest, 94) has a matter-of-fact resonance in a revolutionary setting.
28 Mayer, Below the Line, 58.
29 Ibid., 56–57.
30 Ibid., 58.
31 Ibid., 59.
32 Michael Curtin, Playing to the World’s Biggest Audience: The Globalization of Chinese Film and TV (Berkeley: University of California Press, 2007).
33 One aspect of the compatibility of revolutionary work with digital labor is that they are both unpaid. Though Lazarrato’s elaboration of “immaterial labor” focused on the remunerated kind, there has been an active discussion of unwaged labor in the digital era at least since Terranova’s 2000 article: Tiziana Terranova, “Free Labor: Producing Culture for the Digital Economy,” Social Text 18.2 (2000): 33–58. Andrejevic’s critique of reality television highlighted the unpaid “work of being watched” (Reality TV, 2004). Most recently, see the special issue of TripleC, “Philosophers of the World Unite! Theorising Digital Labour and Virtual Work—Definitions, Dimensions and Form,” ed. Christian Fuchs et al.; especially Brian A. Brown, “Will Work for Free: The Biopolitics of Unwaged Digital Labour,” TripleC: Communication, Capitalism and Critique 12.2 (2014): 694–712, www.triple-c.at/index.php/tripleC/article/view/538.
34 Marwan M. Kraidy, “No Country for Funny Men,” Al-Jazeera America, February 26, 2014, http://america.aljazeera.com/opinion...-funnymen.html.
35 For more details, see Kraidy, The Naked Blogger of Cairo.
36 Joas, The Creativity of Action, 71 (emphasis in original).
37 I suspect that a systematic, theoretical, and comparative examination of “action” and “labor” would unearth fascinating overlaps and differences, but this falls outside the purview of this chapter.
38 Jasper, Art of Moral Protest, 154.
39 Lazzarato, “Immaterial Labour,” 139.
40 Marwan M. Kraidy, “The Revolutionary Body Politic: Preliminary Thoughts on a Neglected Medium in the Arab Uprisings,” Middle East Journal of Culture and Communication 5.1 (2012): 68–76; “The Body as Medium in the Digital Age: Challenges and Opportunities,” Communication and Critical-Cultural Studies 10.2–3 (2013): 285–290; “The Politics of Revolutionary Celebrity in the Contemporary Arab World,” Public Culture 27.1 (2014).
41 Mayer, Below the Line, 41–42.
42 I initially elaborated it in Marwan M. Kraidy, “Governance and Hypermedia in Saudi Arabia,” First Monday 11.9 (2006), http://firstmonday.org/issues/specia...idy/index.html. For an application of the concept in the context of political activism before the Arab uprisings, see Marwan M. Kraidy, Reality Television and Arab Politics: Contention in Public Life (Cambridge: Cambridge University Press, 2010). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/17%3A_Revolutionary_Creative_Labor/17.05%3A_Notes.txt |
Conceiving of social inequality as a salient object of research for media industry studies is a tricky business. As a research matter, approaching inequality is mired in if not now displaced by a cluster of terms likediversity, multiculturalism, difference, lifestyle, andniche. Media’s role in the production of inequalities based on class, race, gender, and sexual identification is displaced onto questions of access and representation, multiculturalism and diversity, branding and audience appeal.
• 18.1: Introduction
Introduction to the goals of this chapter: identifying the conditions that created the discursive alliance between media representation and demography in the U.S, and exploring the limitations of this alliance in modern media.
• 18.2: Problem Space 1 – Employment, Content, and Demography
Examining the assumptions underlying the commonly held equivalence between increased diversity in media and demographic parity between races and genders. Historical origins of this equivalence.
• 18.3: The Current Conjecture
The current conjecture with respect to media and diversity, and how it differs from the historical status of networks, studios, and the state as targets of political protest for which the alliance of representation and demography forms redress.
• 18.4: Problem Space 2 – Difference and Power
The need for media studies to analyze race as a practice of knowledge/power, rather than only through the lens of representation.
• 18.5: Diversity as Quotidian Production Practices
Suggestions for an approach to race and media studies that analyze the organizational structures, creative processes, and social relations underlying the production of inequality in media.
• 18.6: Conclusion – From Inventory to Attachment
The importance of transitioning the focus of media studies on race from inventorying representation to analyzing the affective work of media in galvanizing and organizing the public on issues of race and inequality.
• 18.7: Notes
18: Precarious Diversity Representation and Demography
Conceiving of social inequality as a salient object of research for media industry studies is a tricky business. As a research matter, approaching inequality is mired in if not now displaced by a cluster of terms like diversity, multiculturalism, difference, lifestyle, and niche. Media’s role in the production of inequalities based on class, race, gender, and sexual identification is displaced onto questions of access and representation, multiculturalism and diversity, branding and audience appeal. As the subject of media industry studies research, approaches to the study of diversity often direct researchers to see diversity as a discrete outcome and empirically track rates of diversity in the production and expression of media content.
Thinking with the possibilities opened up by renewed energies and critical foci in media industry studies, I ask what assumptions underwrite how diversity is thought in media studies of race and difference. What evidence locates, measures, and assesses its effectiveness as a social accomplishment? Is the study of diversity a salient means of getting at the role of media in the production of inequalities? As the editors of this collection suggest, following such a research agenda means starting with the methodological assumption that diversity, like studies of creative labor, operates at multiple levels and in multiple registers, including textual representation, reception, and production, as well as in the micro transactions that circulate among different sites.1 Such transactions include critical discourses and industrial practices that organize and nominate media objects as significant and worth studying, as well as the legal and aesthetic disputes that make social differences based on race, gender, and sexual identification objects of legal oversight, political dispute, financial (dis)investment, and administrative management by studios, the FCC, global entertainment corporations, and guilds. As with studies of work objects, deep texts, and implicit ritualized relations, at each of these levels media researchers might aim to identify the quotidian practices of diversity and ask how is it framed, how it works, and to what ends.
Research on questions of diversity (as a gloss for inequality) seems especially suited to neoliberal approaches to studies of media industries as a robust site to generate new evidence about the actual practice of diversity in media organizations and institutions, its expression and production as a practical outcome of the doings that happen in particular and specific production and creative sites.2 This includes researchers asking with respect to diversity, what do creative personnel understand themselves to be doing and what notions of diversity matter, how, and where do such understandings express themselves in their actual quotidian practice?
Diversity is also the object of contentious political, legal, and academic disputes. In the United States, diversity is a practical outcome, the momentary stabilization of a discursive logic and signifying system that produces material effects in the social world. As a proxy for addressing race and a disavowal of racism and inequalities based on gender, racial, and class difference, diversity operates in a shifting nexus of legal rulings, social claims, cultural practices, and media narratives about its practical life and effects. As a key location where diversity is practiced materially and symbolically, the media too is constantly undergoing economic, institutional, and technological change, marked by the appearance and disappearance of new platforms, synergies, financial entities, and international networks of finance and production. The cultural idea of diversity circulates in a media environment where, at least on the issue of race and ethnicity, social difference is a cultural signifier of a (purportedly) postracial America. Cultural signs of diversity, such as language, sexual identifications, school textbooks, and university admissions policies, are not only contested but extremely “hot” discursive objects. As the subject of news stories, reality television, and salacious entertainment, these signs of social difference veer between “postracial” racial insignificance and disputes about the primacy of racial and ethnic difference in access to economic resources and differential exposure to personal vulnerability and social insecurity, such as environmental toxins, police abuse, youth violence, and substandard housing. In this sense, diversity is also a technology of power, a means of managing the very difference it expresses, which prompts me to focus in this chapter on the social life of diversity as a working practice, social commitment, and policy goal in the media as well as media studies scholarship.
Industry, scholarly, and market inventories of the distribution of race and gender difference in media content look to representational parity as the most salient benchmark of diversity in the entertainment business. Of course, representational parity is essentially meaningless without demography as a reference point. Hence the path to diversity in entertainment media must always pass through the “assumed link between representation and demography,” a link that has defined media studies of race and diversity in the United States over several generations now.
What are the conditions of possibility that produced the discursive alliance between representation and demography? Moreover, why (and how) did the discursive alliance between representation and demography come to settle on production as the site of correction and regulation that still organizes scholarly research, industry responses, and state intervention as means of addressing racial and ethnic disparities in U.S. media? In the remainder of the chapter, I detail the technological, discursive, social, and cultural conditions of possibility that gave rise to this initial alliance, then identify the subsequent shifts in media discourses of race and racism that give rise to a different problem space and set of research questions. Drawing on examples from recent media studies scholarship, I then consider some possible ways that researchers and scholars might approach media studies of difference, diversity, and representation that conceive of a different problem space for thinking about media and diversity. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/18%3A_Precarious_Diversity__Representation_and_Demography/18.01%3A_Introduction.txt |
Why diversity, not (in)equality? Or perhaps the assumption is that diversity is the expression of equality? Articulated most explicitly in the 1968 Report of the National Advisory Commission on Civil Disorders, or Kerner Commission Report, the discursive alliance between representation and demography turned on the conception of racial difference and the role that this conception played in contributing to the conception of blacks and the disadvantages and frustrations blacks experienced. Media was a crucial site for addressing grievances of disenfranchised blacks over lack of access to significant positions of employment and the exclusion of black images.3 Addressing racial, class, and gender inequality was never the explicit aim of this alliance; access and inclusion was. Prodded by the Civil Rights Commission and left to their own devices, television networks, newsrooms, showrunners, and advertisers entered a generation-long cycle of lurching in fits and starts toward granting access and including people of color and women in mainstream media content and employment.
Media scholars and historians 4 suggest that there was at least shared agreement among members of the civil rights establishment, black cultural nationalists, and the policy establishment about the importance of aligning television news and entertainment content, media industry employment, and the demography of minority populations. Nowhere is this consensus on the alignment of content, demography, and employment more evident than in the Kerner Commission Report. As media scholar Vicki Mayer reads it, “The Kerner Commission Report of 1968, which concluded that media representation helped fuel national racial unrest, linked problematic discourse (stereotypes) not to mass communication per se but to employment within its related industries.” She continues, “An explosion of publicly and privately financed quantitative studies of television content, employment practices, ownership patterns and cultivated audience effects buttressed social movement claims that distortions on the screen should be mediated through production practices and broadcast regulations.”5 For Mayer, issues of identity in studies of television production seemed tied to the labor that could be held responsible for the representations of race and gender on television. This historic conjuncture of the golden age of broadcast networks, the moral and political pressure of the civil rights movement, and news and entertainment content as the site of cultural affirmation, social recognition, and redress still remains the dominant framework for academic research on race and gender representation in media industries research, especially for cable and broadcast media and television. This reasoning and framework also continues to organize policy approaches to achieving media diversity.
Since 1965, for example, media and communication scholars, activists and pressure groups, journalists and critics, craft guilds and industry observers have provided periodic reports on the state of diversity in North American media and entertainment industries. These reports inventory the number of women, black, gay and lesbian, Asian American and Latino/Latina personnel employed in different production sectors of the U.S. entertainment media, from showrunners and writers in television to directors and producers in cinema. These reports also monitor the state of diversity in front of the screen (according types of characters by genre, role, setting, action, and so on).
Consider a few recent examples that illustrate the continuing influence of the discursive alignment of representation and demography as a measure of racial and gender equality. In March 2014 a respected television critic and columnist, Mo Ryan of the Huffington Post, reported on the dismal state of affairs for diversity in entertainment television: “At the outlets responsible for many top programs, women and people of color are enormously under-represented as creators. If one focuses only on the last dozen years at AMC, FX, Showtime, Netflix and HBO, around 12 percent of the creators and narrative architects in the dramatic realm were women. . . . According to the most recent stats from the Writers Guild of America, about 30.5 percent of TV staff writers are women, and about 15.6 percent of TV writers are people of color; both numbers represent modest gains from the past. San Diego State University’s Center for the Study of Women in Television and Film, which uses a different calculation method, puts the percentage of female TV writers for the 2012–13 season at 34 percent. . . . Yet according to SDSU’s most recent study, 27 percent of women bear the title executive producer, and 24 percent are a ‘creator’—numbers that have remained stagnant for a long time.”6
Sociologist and media scholar Darnell Hunt authored one of the reports cited by journalists, industry observers, and pressure groups concerned about the state of racial and gender diversity in Hollywood. Hunt’s 2014 Hollywood Diversity Report tracks longitudinal data on the distribution of actors, writers, directors, agencies, and audience in film, cable, and broadcast outlets. According to Hunt, minorities fare better as leads in cable comedies and drama compared to broadcast at 14.7 percent, while women fare worse as leads in cable comedies and dramas than in broadcast at 37.2 percent. Minorities, in contrast, are more likely to be leads on reality and other shows than on comedies and dramas in broadcast.7 Citing a Writers Guild of America West 2013 report, Hunt emphasizes that, according to the report, “diverse writers were underrepresented by a factor of about 4 to 1 among writer-producers with the most decision-making authority, both in the development of original network show concepts and in the day-to-day management of the storytelling process . . . despite the fact the minorities collectively accounted for 36.3 percent of the nation’s population in 2010.”8
In another highly respected state-of-the-industry report, Stacy Smith and her colleagues at USC’s Annenberg School of Communication provide a highly detailed annual report on the dismal state of gender and racial diversity in the Hollywood film industry.9 As with Hunt’s 2014 Hollywood Diversity Report, the San Diego State University study, and the Writers Guild of America West and Director’s Guild of America findings, Smith’s study is based on longitudinal data. Smith found a similar absence of gender, racial, and ethnic diversity in Hollywood, drawing conclusions similar to those of Hunt and other researchers. On the index of gender, participation seems somewhat more hopeful than on race and ethnicity, though the general trends suggest that despite industrial transformations in production, financing, and service delivery in television and film production overall, movement in racial and gender inclusion and participation has not kept up with these transformations.
What accounts for the persistent patterns of racial and gender exclusion reported in these empirical studies? Surely after years of reporting on such practices of exclusion, media executives, advertisers, content producers, and program purchasers are aware of the dismal state of affairs with respect to diversity in media industries. In the face of so much documented evidence about the lack of racial and gender diversity in television and cinema, what else might be going on? What else might account for the failure of these reports and the evidence they present to gain any lasting traction? What would it take at the level of policy prescriptions, industry practice, and guiding assumptions for this evidence to matter in ways that would change the practices of exclusion they report? Periodically, advocacy groups like the National Association for the Advancement of Colored People, the Mexican American Legal Defense and Education Fund, and GLAAD use these reports to leverage studio and network executives to hire more women and people of color, develop more content aimed at diverse audiences, and earmark job training programs to develop talent in different sectors of the industry.10
What if we shifted the angle of vision, treating inequality and the absence of diversity as a process? What if we see the absence of diversity, or more properly inequality, in media as a crucial component of the production of creative objects, labor relations, financing, distribution, and marketing, and not just discrete outcomes within the associated fields of production aimed at representational and demographic parity.11 Why not expand the analysis to include the very way we frame and interrogate issues of diversity? As the handful of reports cited already show, research scholars, craft guilds, industry leaders, regulators, and advocacy groups understand diversity in media industries as a matter of whether or not television, cinema, and now different sectors of new media like gaming employ a diverse workforce, which by extension is presumed to result in more diverse content.
This continues to be an important goal, to be sure, but it conceives of diversity as a fixed outcome, measurable in the number and distribution of discrete indicators like the number of minority showrunners or the number of women in lead roles. While this approach addresses questions of representational parity, it raises other questions, especially the relation between media industries and inequality, including whether correctives to inequality can be addressed by the exchange of bodies and experiences responsible for making content, rather than by exposing the assumptions, micropractices, social relations, and power dynamics that define our collective cultural common sense about the nature of social difference and the practices of inequality.
As a research agenda and public policy goal, by far the dominant approach to media diversity is framed from the vantage point of a problem whose specific roots go back over fifty years to black urban unrest, the golden age of network television, and a liberal consensus on the Great Society. In this discursive alliance, legal, cultural, social, and political assumptions located in the Great Society and the civil rights movement consensus set the terms of a framework that aligns demographic representation, the politics of representation, the conception of media and television as cultural sites of redress for racial injury, and the assumption that a corrective of the image will equal social justice and political parity. That is, the legal terms of state recognition engendered by the civil rights movement, the idea that demographic parity and media parity should be equivalent, that merely having diverse content would achieve demographic parity, and that minority access to the dominant image culture would equal social redress. So by dwelling on the conditions of possibility and the assumptions that frame media studies approaches to diversity and the empirical evidence by which media scholars measure its distribution and assess its efficacy, perhaps we can begin to account for why the discursive alignment that defines much of the research on diversity and the media has proved to matter so little in reordering the racial order of things in the media.
By conceiving of diversity as a social accomplishment and emphasizing the shifting, contested, and precarious nature of the social context and power relations that diversity elicits, organizes, and charges relative to its changing conditions of possibility, we might begin to ask different questions about media practices of diversity as a proxy for inequality. Diversity’s precarity invites probing the shift in academic and popular discourses of diversity, especially legal disputes over the very meaning and conception of difference generated by a host of new legal and cultural claims and grievances. A good place to begin might be with the disarticulation between the problems to which studies of racial and ethnic distribution in content and production were generated to provide answers and our own conjuncture, on which these studies are called upon to comment. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/18%3A_Precarious_Diversity__Representation_and_Demography/18.02%3A_Problem_Space_1__Employment_Content_and_Demography.txt |
Among the most significant elements that define the current conjuncture with respect to media and diversity, we would certainly have to account for the impact of the new international division of labor and the new international division of cultural labor, including especially the rise of new media capitals and production centers, such as Bangalore, Lagos, and Hong Kong, along with highly skilled and unskilled labor forces.12 These production centers generate specific content organized around nation, ethnicity, language, and history, all of which provides a sharp contrast to—while adding to and complicating—U.S. conceptions of diversity. So too the changing functions of culture within the global circulation of information, entertainment, and cultural products that scholars like George Yudice and Arlene Davila explain as the expediency of cultures. These expediencies add cultural diversity and celebrations of diversity and cultural differences to the working of culture and media, especially in terms of the circulation of content ranging from gaming culture to cinema.13
The post-network-television environment and the realignments of platforms for the delivery and distribution of content, as well as reception experiences and consumption practices, require a different research approach with diversity as a complex object of research.14 In this context, multicultural programming content is very much an element of branding and marketing deployed by content producers to reach precise sectors of their desired markets (for instance, reality television shows about rural white working-class families, or programs about black church women or the ordeals of Silicon Valley high-tech employees). Seen from this perspective, it would seem that in the current media ecology, diverse characters, story lines, and content producers are no longer in short supply and that the alignment between representation and demography, now understood as identifiers for market segments and lifestyle choices, has, in the gloss of the postracial society, been fully realized. This is no longer a condition of content scarcity but one of saturation and hypervisibility.
Added to corporate transformations of the U.S. media ecology is the impact of juridical and legislative state institutions and practices in legally inscribing and authorizing color-blindness in voting rights laws and college admissions. Underwritten by the principle of color-blind legal and social practices, diversity, supplemented by its proxy multiculturalism, is the expressive form for the institutionalization of what Rod Ferguson calls minoritarian discourse. For Ferguson, the discursive life of minoritarian discourse is expressed explicitly as a value commitment to diversity in the culture of university classrooms as well as in scholarly research. Ferguson cautions that this value commitment has an increasingly normative function, and thus in his view is an operation of power/knowledge that works to exclude the social actors and political struggles that made race, ethnicity, and gender discursive sites of struggle around inequality and emergent sites of non-normative imaginaries.15 This reordering provides the scholarly rationale and authority for the circulation of knowledge about lifestyle differences that find their way into production suites, executive offices, and media content.
The point is simply that unlike the conditions that produced the alignment of networks, studios, and the state as objects of political protest by groups and the alliance between demography and representation as the accepted means of redress, the conditions that define the present conjuncture are predicated on the cultural recognition of difference, the deployment of diversity as a social practice, and their normative operation as a discourse of management and regulation.16 In short, the current conjuncture destabilizes the alliance between demography and representation as a response to exclusion, invisibility, and stereotypes and reorders it around diversity and multiculturalism as markers of consumer brands, lifestyle choices, and postracial cultural appreciation.
18.04: Problem Space 2 Difference and Power
In their critical assessment of studies of race and ethnicity in media and communication studies, David Hesmondhalgh and Anamik Saha observe that production studies has given woefully little attention to questions of race and ethnicity.17 The lacuna they signal is as much the result of the analytic confinement and discursive linkages of race to people of color (and not the operation of whiteness) as it is to not appreciating the logic of creative practices, especially media, as a site of making race and practices of inequality. It follows too that inattention to race-making rather than racial representation in media studies assumes that the source of inequality and racism rests with individual preferences and dispositions of showrunners and directors, network executives, and advertising executives. Concerns with diversity and race as a practice of knowledge/power (what John Caldwell calls the deep texts of production cultures) are not endemic to the organization of media industries or research approaches to their study.18 Designing studies of media and race in the current conjuncture at the least suggests foregrounding an analytic of governmentality, televisuality, neoliberalism, and the role of diversity in making race.19
To these I would add the need for systematic attention to media production and the operation of racial knowledge as a repetition of inequality (and knowledge about differences in race, gender, and sexuality) embedded in the routine habits, assumptions, practices, rituals, and organization of cultural work. Media industry and television studies might productively address some of the concerns identified by Hesmondhalgh and Saha by engaging with creative industry and production studies research agendas to identify sites, discourses, and practices of producing difference and to study race-making practices as power/knowledge that operates as a logic of production. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/18%3A_Precarious_Diversity__Representation_and_Demography/18.03%3A_The_Current_Conjecture.txt |
In her study of the shift in the nature of the work object, its impact on social relations among television writers in Los Angeles, and the precarity of their work as writers, television scholar and director Felicia Henderson examines the transformation of the traditional work product, the television season, as a key unit of analysis.20 Specifically, she focuses on the shortening of the television season from twenty-two programs to thirteen, which is made possible by changes in delivery systems, viewing platforms, viewing practices, and contract negotiations.
The virtue of Henderson’s insight is that her analysis of the writer’s room as a site of creative production (like others in this genre) dwells on the structure of creative relationships, industrial settings, and the organizing logic that defines the production and creative processes rather than on individual personalities and attitudes of creative personnel. In terms of the applicability of her approach to a concern with race and diversity, Henderson’s research commends attention to the organizational sites, creative processes, and social relations where the practices of diversity (or impediments to diversity) operate. Such an approach appreciates the fluidity and flexibility of evidence and analysis across time and space so that foundational categories, policy mandates, political stakes, and analytic conceptions can shift with the historical, technological, and political conditions in which they embedded and which they help organize and narrate. In other words, analytically it is useful to look at the conditions that structure and organize some of the foundational assumptions and questions about diversity and television (aims, means of realizing, forms of monitoring and assessing their effectiveness).
Henderson’s approach suggests that the specific conceptions, conditions, and assumptions that produce diversity (or the twenty-two-episode season as a staple unit of network television) as a desirable goal in television are not static, nor is the nexus of institutions, interests, and stakes that support or oppose its actualization. Her approach encourages an analysis of diversity (or impediments to diversity) as a dynamic and flexible set of industrial, legal, cultural, and economic practices that the study of the precarity of creativity and diversity can bring to bear on the question. In other words, the specific conceptions, conditions, and assumptions that produce diversity as a goal and practice within creative media industries like television are not fixed, and neither is the nexus of institutions and logics that organize and express them.
The challenge is moving the research focus from the founding scene of the problem of racial access and image exclusion within television to the shifting conditions that shaped television and discourses of race, including the rise of diversity, since the Kerner Commission Report. What, in other words, are the implications for quotidian practices of inequality and making race that the shorter seasons, new delivery systems, new interactive platforms, new divisions of labor, and new relations of production crystalize? What might the impact of these developments be on the very terms within which we pose the question of inequality in television that diversity glosses? Such reframing moves the issue of diversity some way from the analytic social and political scene in which it initially appeared. It shifts the industry and analytic assumption of equating diversity and social equality with access and representational parity to one where the calculus of cultural, economic, and political difference as a basis of the production of inequality is central to media industry practices.
This approach to research on television and race scrambles foundational binaries that continue to inform industrial practice, academic approaches, and media activism: inside/outside, accuracy/stereotype, author/imitation. With respect to race and difference, the terrain is considerably more complex and urges different questions that creative industry studies might help clarify: 1) How is diversity and difference framed as a labor issue and as a matter of work process and contractual management? 2) In what respect does the international division of creative labor pressure local and national formulations of diversity as matters of representation, reparation, and labor? 3) In what respects do the new international division of cultural labor, the rise of new platforms and delivery systems, and the creative arrangements that drive new projects, genres, talent bear on the question of diversity in new and unforeseen ways, especially within different national formations defined by distinct racial projects and ethnic formations? 4) These conditions could just as well open the way for the media production of diversity as cultural normativity or a technology of power/knowledge deployed to reach lifestyle niches. A critical media industry approach to inequality (rather than merely diversity) would urge that media and ethnic/racial arrangements be located and analyzed within the context of racism, racial projects, and race making nationally and globally.
On this count, John Caldwell’s insights about reflexivity, industrial knowledge, and practices and rituals among cultural producers are exemplary;21 so too are Vicki Mayer’s considerations of the inscription of knowledge in practices at all levels of the production process, as is Sarah Banet Weiser’s work on the role of brands and branding as a mode of crafting and caring for the self in the construction of diversity.22 Finally, Timothy Havens’s explorations of the circulation of television content about blackness in the United States and the role of industry lore about race and diversity in the creative process are especially rich. So too are Darnell Hunt’s studies of African Americans who use new media technologies to write, produce, and perform alternate and nonhegemonic conceptions of complex and intersecting minoritarian identities.23 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/18%3A_Precarious_Diversity__Representation_and_Demography/18.05%3A_Diversity_as_Quotidian_Production_Practices.txt |
Jennifer Petersen’s Murder, the Media, and the Politics of Public Feeling might be taken as the kind of study that moves away from the dominance of concerns with parity and representation as routes to social and racial justice toward concern with the affective work of media in galvanizing feelings, organizing publics, and materializing grievance.24 In her concern with emotional conflicts, legal adjudications, and social negotiations over the depiction, circulation, investment, and use of the coverage of the murders (and their aftermath) of Matthew Shepard and James Byrd, Petersen traces the circulation, disputes, and impact of the emotional economy of these events. She shows how they came to matter on questions of sexuality, race, gender, and nation. Petersen’s study prompts media studies of race and diversity to consider not just what things mean but also how they matter, where, for whom, and with what effects.
To Peterson’s emphasis on the relationship of media to public feeling and how things matter, I urge attention to the concerns mobilized by media content and the resonances it generates for users as well as producers of content. Engaging research this way may at the very least complement if not reimagine insights that the nexus between representation and demography now yields. Practically, this means moving away from the assumption that a bid on image accuracy and authenticity anchored by demography will provide some assurance of social parity. It suggests moving toward the possibility that a focus on resonance and attachment might critically address the complexities of race making and the production of diversity as a technology of power in the current conjuncture. Signaling matters of concern registers a different assumption, one that considers the inscription of racial meaning as endemic media work.
Complementing media studies of production, industrial organization, and routine media practices with critical research on the intensity, duration, and locus of emotional concerns engendered by media could direct critical analytic attention to forms of attachment and identification that do a bit more than document annual diversity effects in media. The alliance of discursive and social conditions of possibility that has defined much of media studies research on race and media for several generations now suggests that we have reached a critical limit of the capacity of the alliance of demography and representation to tell us enough about the practice, production, and normalization of diversity to matter. The annual research reports on representational parity have themselves become normative, organizing and fueling policy prescriptions, research agendas, guild training programs, marketing research, and branding campaigns. Perhaps it is time to ask that our research tell us a different story about the operations of power/knowledge and the role of media in the making of racial inequality (and its potential for the making of racial justice).
18.07: Notes
1 Miranda Banks, “How to Study Media and Makers,” in The Sage Handbook of Television Studies, ed. Manuel Alvarado, Milly Buonanno, Herman Gray, and Toby Miller (London: Sage, 2014).
2 John Caldwell, Production Culture: Industrial Reflexivity and Critical Practice in Film and Television (Durham, NC: Duke University Press, 2008).
3 “The News Media and the Disorders,” in Channeling Blackness: Studies on Television and Race in America, ed. Darnell Hunt (London: Oxford, 2005); Kerner Commission, Report of the National Advisory Commission on Civil Disorders (Washington, DC: U.S. Government Printing Office, 1968); Arthur S. Fleming, Stephen Horn, Frankie M. Freeman, Manuel Ruiz Jr., Murray Saltzman, and John A. Biggs, Window Dressing on the Set: Women and Minorities in Television (Washington, DC: U.S. Commission on Civil Rights, 1977).
4 See, for instance, Aniko Bodroghkozy, Equal Time: Television and the Civil Rights Movement (Urbana: University of Illinois Press, 2012); Lynn Spigel, Welcome to the Dreamhouse: Popular Media and Postwar Suburbs (Durham, NC: Duke University Press, 2001); Laurie Ouellette, Viewers Like You? How Public TV Failed the People (New York: Columbia University Press, 2002); Sasha Torres, Black, White, and in Color: Television and Black Civil Rights (Princeton, NJ: Princeton University Press, 2003); Chon Noriega, Shot in America: Television, the State, and the Rise of Chicano Cinema (Minneapolis: University of Minnesota Press, 2000); Devorah Heitner, Black Power TV (Durham, NC: Duke University Press, 2013); Steven Classen, Watching Jim Crow: The Struggle over Mississippi TV, 1955–1969 (Durham, NC: Duke University Press, 2004); Anna McCarthy, The Citizen Machine (New York: New Press, 2010); and Vicki Mayer, Below the Line: Producers and Production Studies in the New Television Economy (Durham, NC: Duke University Press, 2011).
5 Mayer, Below the Line, 13 (my emphasis); see also Kerner Commission, Report of the National Advisory Commission on Civil Disorders.
6 Mo Ryan, “Who Creates Drama at HBO? Very Few Women or People of Color,” Huffington Post, March 7, 2014.
7 Darnell Hunt, “Hollywood Story: Diversity, Writing and the End of Television as We Know It,” in The Sage Handbook of Television Studies, ed. Alvarado et al; Ralph Bunche Center, 2014 Hollywood Diversity Report: Making Sense of the Disconnect (Los Angeles: UCLA, 2014).
8 Hunt, “Hollywood Story,” 166.
9 Stacy Smith et al., “Gender Inequalities in 500 Popular Films: Examining On Screen Portrayals and Behind-the-Scenes Employment Patters in Motion Pictures Released between 2007–2012,” Annenberg Report (Los Angeles: University of Southern California, 2013); Stacy Smith et al., “Race/Ethnicity in 500 Popular Films: Is the Key to Diversifying Cinematic Content held in the Hand of the Black Director?” Media Diversity and Social Change Initiative (Los Angeles: University of Southern California, 2013).
10 Directors Guild of America, “‘A Fair Shot’: Women Directors on Television,” www.dga/Craft/DGAQ/All-Articles/1301-Winter (accessed September 19, 2014); Directors Guild of America, “DGA Report: Employers Make No Improvement in Diversity Hiring in Episodic Television,” www.dga.org/ News/PressReleases/2014/140917.
11 Pierre Bourdieu, The Field of Cultural Production: Essays on Art and Literature (New York: Columbia University Press, 1993).
12 Michael Curtin, “Media Capitals: Toward the Study of Spatial Flows,” International Journal of Cultural Studies 6.2 (2003): 202–228; Toby Miller, Nitin Govil, John McMurria, and Richard Maxwell, Global Hollywood (London: British Film Institute, 2001); David Hesmondhalgh, The Cultural Industries (Los Angeles: Sage, 2006).
13 George Yudice, The Expediency of Culture: The Uses of Culture in the Global Era (Durham, NC: Duke University Press, 2003); Arlene Davila, Culture Works: Space, Value, and Mobility across the Neoliberal Americas (New York: New York University Press, 2012).
14 Amanda Lotz, The Television Will be Revolutionized (New York: New York University Press, 2007).
15 Roderick Ferguson, The Reorder of Things: The University and Its Pedagogies of Minority Difference (Minneopolis: University of Minnesota Press, 2012).
16 Herman Gray, “Subject(ed) to Recognition,” American Quarterly 65.4 (2013): 771–799.
17 David Hesmondhalgh and Anamik Saha, “Race, Ethnicity and Cultural Production,” Popular Communication 11.3 (2013): 179–195.
18 Caldwell, Production Culture.
19 Laurie Ouellette and James Hay, Better Living through Reality TV (London: Wiley, 2008).
20 Felicia Henderson, “Options and Exclusivity: Economic Pressures on TV Writers’ Compensation and the Effects on TV Writer’s Room Culture,” in The Sage Handbook of Television Studies, ed. Alvarado et al.
21 Caldwell, Production Culture.
22 Mayer, Below the Line; Sarah Banet-Weiser, Authentic (New York: New York University Press, 2012).
23 Timothy Havens, Black Television Travels: African American Media around the Globe (New York: New York University Press, 2013); Hunt, “Hollywood Story.”
24 Jennifer Peterson, Murder, the Media, and the Politics of Public Feelings (Bloomington: Indiana University Press, 2011). | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/18%3A_Precarious_Diversity__Representation_and_Demography/18.06%3A_Conclusion__From_Inventory_to_Attachment.txt |
When Alex Nogales, president and CEO of the National Hispanic Media Coalition (NHMC), narrates the history of his organization, he tells a story of continuity and change. The core mission of the group—to integrate Latinas/os into more jobs behind and in front of the camera, ameliorate derogatory images of Latinas/os in the media, and advocate for telecommunications policies that serve the needs of Latina/o publics—has remained consistent since the NHMC was founded in 1986. What has changed, according to Nogales, is the organization’s strategies, which have evolved with the group’s experiences in media activism and advocacy.
• 19.1: Introduction
Introduction to the goals of this chapter: to examine the actions of media advocacy groups as both long-term work and media work, and the intersection of media advocacy with media production, with a focus on the National Hispanic Media Coalition.
• 19.2: Media Advocacy as Work
Examining the NHMC's media advocacy over the last several decades as an ongoing, cumulative work process.
• 19.3: Media Advocacy as Media Work
Examining the NHMC's media work, such as operating a para-industry for training talent and honoring media worker accomplishments, and how it is partly funded by media corporations. Contrasting the NHMC's record with other corporate-supported minority advocacy groups that have taken pro-corporate stances on issues such as media ownership and net neutrality.
• 19.4: Conclusion – The Precarity and Politics of Media Advocacy Work
Summary of the reasons for precarity in media advocacy work, including challenges posed by the regulatory community, industry interests, the political culture, and other advocacy groups.
• 19.5: Notes
19: The Precarity and Politics of Media Advocacy Work
When Alex Nogales, president and CEO of the National Hispanic Media Coalition (NHMC), narrates the history of his organization, he tells a story of continuity and change. The core mission of the group—to integrate Latinas/os into more jobs behind and in front of the camera, ameliorate derogatory images of Latinas/os in the media, and advocate for telecommunications policies that serve the needs of Latina/o publics—has remained consistent since the NHMC was founded in 1986. What has changed, according to Nogales, is the organization’s strategies, which have evolved with the group’s experiences in media activism and advocacy. In his telling, the NHMC went from being a comparatively naïve organization, committed to addressing the exigent concerns of local communities, to a sophisticated group capable of exerting meaningful pressure on a national scale, especially via participation in the policymaking sphere.1
The NHMC’s emphasis on media labor has been in keeping with the priorities of other identity-based media advocacy groups who have worked to bring people of color into media industry workforces at all levels. For the NHMC, to ensure that Latinas/os have access to these jobs is, like other equal employment advocacy, to enable them to participate in a sector that had historically discriminated against them; in addition, it is to transform the kinds of stories told and perspectives voiced in media texts, from news reports to entertainment programming. While securing Latina/o jobs has been a consistent goal of the NHMC, it has had to navigate a legal environment increasingly hostile to race-conscious policies to promote diversity and a regulatory system increasingly committed to media deregulation. In response, the NHMC, like other advocacy groups, has had to rethink how to promote diversity in the absence of what had been essential regulatory tools and in a climate unreceptive to such interventions.
Media advocacy, the kind of actions undertaken by groups like the NHMC, thus not only has been centrally concerned with media labor, but has constituted its own form of work. The work of media advocacy often is a labor-intensive enterprise, one that relies on myriad forms of capital—financial, cultural, institutional—to function. While media advocacy has often depended on uncompensated labor, from the work of volunteers whose contributions create the scaffolding upon which media advocacy efforts are built to the citizens who respond to calls to action by filing letters with or calling the Federal Communications Commission (FCC) or members of Congress, it also has been guided by media advocacy professionals. These are professionals in two senses of the term: they have expertise and they are compensated for their labor.
To examine media advocacy as work is to alter the kinds of questions we ask and the kind of narratives we construct. While there are meaningful differences in how scholars have understood the political stakes, moments of opportunity, and mobilizing structures and strategies of media advocacy efforts, what they share is an understanding of media advocacy as a social movement or as a form of civic participation that has sought to transform the media to meet the communication needs of citizens in a democracy. Media advocacy campaigns are often narrated as David-and-Goliath stories, in which public interest groups try to reform the media only to be defeated by better-resourced media corporations that more successfully manipulate public opinion and gain sway over public officials.
The emphasis of media advocacy scholarship, furthermore, often is the media advocacy campaign, a temporally bounded effort undertaken at a particularly propitious moment when political changes or new technologies introduce fissures that make reform seem possible.2 Media advocacy has also often been analyzed along a success/failure binary, an assessment of how and why media advocacy has or has not attained its desired goals. As the first section of this article discusses, to see media advocacy as work is to shift our focus off outcomes and onto process and to rethink the success/failure binary that has structured much of media advocacy scholarship. Media advocacy for groups like the NHMC is a long-term, multifaceted commitment that shifts with technological, political, and regulatory changes, as well as with the increasing savvy of the media advocates themselves. Their work is continuing, not contingent on singular campaigns or issues. Viewed through this lens, media advocacy can be seen less as a rhythmic exercise in hope and failure and more as a continuous hum of activity that sometimes yields actionable policy changes, in which communities outside the official regulatory sphere make themselves legible as stakeholders in the policymaking process. To consider media advocacy as work is to see it as ongoing, cumulative, and flexible.
In addition, as the second section demonstrates, many contemporary media advocacy groups in the United States are engaged in media work, labor that contributes to, rather than interferes with, media production and the interests of media companies. Media advocacy, however, has been invisible to scholars of media labor, who mostly have been interested in how the production process under which media are made, as well as the occupational cultures and power relations structuring the mode of production, affects the narratives, values, and images that media audiences consume. Deploying ethnographic and historical methods, and focusing on a range of media, this subfield traditionally has focused on above-the-line workers (directors, writers, producers, and executives); labor within these texts is imagined as both the creative labor of artists and the managerial labor of executives, the friction between them understood as alternately stifling and generative for the production of media texts.3
More recent scholarship, however, has expanded the methods and subjects of media labor scholarship. John Caldwell, for example, has blended ethnographic research with sophisticated discourse analysis to investigate not only the diverse range of labor practices—both above and below the line—that constitute film and television production, but the discursive labor involved in shaping and sustaining the occupational cultures within the entertainment industry.4 In a similar vein, Vicki Mayer, in her Below the Line, has broadened the definition of production to include the “invisible labor” that is constitutive of television production but frequently absented in both industry and academic discourse.5 Conceptions of media labor thus have been extended to the myriad forms of work that contribute to media production and to the discursive formations that sustain its division of labor.
While media advocacy has often existed outside media production, it has also intersected with, and contributed to, both the workflow of media production and the underlying assumptions about audience and narrative that structure it. For decades, media advocacy groups’ work with media producers has been a constitutive part of their reform efforts. Increasingly, however, this collaboration has extended to advocacy groups using their position as representatives of the public to promote the policy agenda of media corporations. As the second section discusses, for some organizations, media advocacy work thus has given way to media work, their adversarial role transformed into a collaborative—or, to some critics, collusive—one with media and telecommunications corporations. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/19%3A_The_Precarity_and_Politics_of_Media_Advocacy_Work/19.01%3A_Introduction.txt |
In the United States, public participation in media policymaking is technically part of the process. The FCC is required to solicit public input on new policies or changes to existing regulations. By design, members of public are to have their say in the shaping of regulations; in actuality, the role of the public has been far more constrained. Not only have industry lobbyists and attorneys had far more purchase with policymakers than members of the public, but administrative law requires federal agencies to consult the public but does not require them to pay heed to what the public says.6 As a result, a range of social movement and civil society organizations have included media advocacy in their broader fights for social justice and political reform, and a number of dedicated media advocacy groups have emerged with the mission to reform the media. Many of these groups have been engaged in media advocacy for decades and have adapted to changes in media technologies, regulatory decisions, and broader political and social conditions.
As Becky Lentz and I have argued elsewhere, media advocacy hinges on the acquisition of media policy literacy, a set of competencies to understand not only the processes by which media policies and laws are formed, debated, and enacted, but also how to participate in a milieu of action to effect meaningful change. This literacy forms out of experience; that is, it is through sustained participation in advocacy that individuals and organizations gain the capacity to critique the sociopolitical impact of media structures, media practices, and media representations, and to strategize how best to tackle them.7 Part of this literacy involves recognizing the myriad functions of a media advocacy campaign. While campaigns have identifiable goals, they also make an advocacy group legible as a stakeholder in the policymaking process and can establish the group’s credibility with fellow advocacy practitioners. The work of the NHMC, which has been committed to media reform for nearly thirty years, exemplifies the long-term, multifaceted, and flexible nature of media advocacy work.
The history of the NHMC shows the organization expanding its understanding of how media and communications matter to the Latina/o community and accordingly increasing the scale of its activities. When the NHMC first began its media advocacy work, it focused primarily on the practices of local broadcast stations. The NHMC utilized the petition to deny license renewal to broadcast stations as its primary means of redressing discriminatory employment practices and derogatory programming. The threat of a petition often would incline local stations to negotiate with the group rather than face the legal fees and irritations of a license challenge. In its early years, the NHMC reached agreements with local Los Angeles stations and soon extended its reach to television and radio stations in heavily Latina/o areas across the United States. The NHMC, in the process, also built ties with Latina/o groups in communities across the nation and began to establish its visibility as a Latina/o rights organization centrally committed to reforming media practices.8
Throughout the 1990s, the NHMC enlarged its focus to include not only local stations but also broadcast and cable networks, along with the media conglomerates that owned them. In addition to an extensive economic boycott of the entertainment holdings of Disney-ABC in 1997, the NHMC targeted media consolidation, specifically the merger of ABC and Disney and the sale of the Spanish-language network Univision to non-Latina/o interests, in its advocacy campaigns. The NHMC was especially concerned over the potential transformation of Univision, the largest Spanish-language television network in the United States at the time, into an adjunct to Mexican and Venezuelan media empires.9 For the NHMC, media consolidation in the English-language sphere and foreign control of the Spanish-language sector would portend fewer jobs for Latinas/os, diminished opportunities for Latinas/os to gain control of their own stations, and the continued invisibility of Latina/o concerns and perspectives in the national media.
Throughout, the NHMC confronted a regulatory apparatus that was seemingly disinterested in enforcing existing policies, especially around media ownership restrictions. These experiences signaled to the NHMC a divide between policy and enforcement and exposed a persistent willingness on the part of the FCC and the federal courts to facilitate media consolidation even in the face of the commission’s own rules against it. In addition, though the NHMC was not able to prevent the sale of Univision in the 1990s, its tenacity in fighting it established the organization as a formidable Latina/o advocacy group. Univision sent representatives to meet with the NHMC in the mid-1990s, and in exchange for ceasing their legal actions, the NHMC gained programming commitments in areas like children’s educational television, which it viewed as critical to the needs of the Latina/o community.10
These experiences in the 1990s were highly instructive for the NHMC in its approach to media advocacy. It more fully committed to affecting policy at the national level—as Nogales states, the NHMC realized that the “big game” was being played in DC—and in the early 2000s hired two attorneys specifically to do policy advocacy work. In addition, its scope continued to increase as telecommunications issues of particular concern to the Latina/o community arose—for example, the expansion of broadband connectivity, the preservation of network neutrality, the maintenance of the Universal Service Fund. And as nativism accelerated in the United States in the mid-2000s over undocumented immigrants, the NHMC has made hate speech one of its top priorities, combating what NHMC executive vice president and general counsel Jessica Gonzalez refers to as “low-hanging fruit,” the programs that circulate what strikes the NHMC as particularly dangerous invective against the Latina/o community.11
Media consolidation has continued to be a top policy issue for the NHMC. Since 2003, it has worked continually to prevent the FCC from diminishing its ownership restrictions. And while it has fought some media mergers—most notably the 2011 proposed merger between T-Mobile and AT&T—it also has sanctioned mergers in exchange for concessions for communities of color. Perhaps most controversially, the NHMC encouraged the FCC to approve the merger of Comcast and NBC-Universal in 2010. When asked to serve on a Hispanic advisory board, the NHMC and other Latina/o groups negotiated a memorandum of understanding (MOU) with Comcast and NBC-U for diversity measures such as the creation of a Hispanic Advisory Council, increased Latina/o representation in the companies’ workforce, enhanced procurement diversity, and the expansion of Spanish-language broadcasting. Members of the NHMC subsequently held ex parte meetings with FCC commissioners in which they described the conditions of the MOU and asked, should the merger be approved, that enforcement of the MOU be written into its conditions.12
A galvanizing moment for the NHMC took place in 1999 and 2000, when it banded together with other identity-based advocacy groups to secure memoranda of understanding with each of the Big Four (ABC, CBS, NBC, and Fox) broadcast networks. Greg Braxton in the Los Angeles Times had reported that of the twenty-six new prime-time shows premiering across the major networks, not one had a person of color in a recurring role.13 Working in a “grand coalition” with the NAACP, the Asian Pacific America Media Coalition, and Indians in Film and Television, among others, the NHMC secured MOUs that included hiring commitments, mentorship and training programs, commitments to work with minority-controlled vendors and production companies, and designations of in-house executives to promote diversity.14 These MOUs were struck at a low point for minority advocacy work, as the federal courts and Congress by 1999 had eliminated or ruled unconstitutional all the rules adopted in the 1960s and 1970s to promote minority employment and ownership in broadcasting. Direct negotiations with the networks were, at this moment, the most immediate and advantageous way to bring more people of color into the television industry. It was this experience with the networks, according to Nogales, that shaped how the NHMC approached the NBC-Comcast merger.15
The NHMC was certainly not the only civil rights or advocacy group to support the merger. The NAACP, National Urban League, and National Action Network similarly secured an MOU with the two companies for programming and hiring commitments, as did a consortium of Asian American civil rights groups.16 The stance of these organizations put them at odds with public interest and consumer advocacy groups who had been allies, especially over media consolidation issues, including Free Press, whose then president and CEO Josh Silver labeled the merger a “comcastrophe,” fearing that with it would come an onslaught of greater levels of consolidation that would diminish diversity, raise prices, and gut network neutrality.17 While the NHMC feels ambivalent about its role—Nogales referred to the NHMC’s action as something of a “cop-out”—its actions speak to a tension within its advocacy agenda. While philosophically the NHMC sees public interest harms in media concentration, it also, as part of its mission, has prioritized the inclusion of Latina/o perspectives and narratives in the media and Latina/o access to jobs within media industries. Its decision to support the merger thus speaks to the experience of the NHMC in unsuccessfully fighting mergers of the past, its assessment of the FCC’s inclination to approve, and its estimation that this was the best way to secure some services to its community. And to be sure, identity-based media advocacy groups historically have butted heads with public interest advocacy groups over the issue of media consolidation. While the latter have imagined substantial public interest harms in enabling fewer companies to own more media properties, the former at moments have been willing to sanction media mergers in exchange for concessions, especially hiring and programming commitments.18 When the NHMC supported the NBC-Comcast merger, it followed in a longer history of civil rights organizations choosing to secure benefits for their communities at a moment when it seemed like the regulatory sphere was inhospitable to considerations of minority media rights.
As the shifting strategies of the NHMC illustrate, examining media policy advocacy as work illustrates that it is an ongoing process in which advocates continually learn and revise the optimal way to intervene in the policymaking process. Their campaigns hinge on and are informed by previous experiences with advocacy. Accordingly, media policy advocacy is a cumulative process in which advocacy groups both acquire the skill sets and resources necessary to intervene in policymaking while at the same time adjusting their expectations of what can be accomplished at particular historical junctures. Sometimes, as in the case of the NBC-Comcast merger, this experience leads advocacy groups to work with media companies and to use their standing as public interest representatives to sanction their interests. In other words, as the next section addresses, media advocacy work can constitute media work. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/19%3A_The_Precarity_and_Politics_of_Media_Advocacy_Work/19.02%3A_Media_Advocacy_as_Work.txt |
Media advocacy has long been concerned with shaping the parameters of what media production can be and how it can be profitable. Battles over, for example, media ownership limits, equal employment rules, children’s television requirements, and indecency regulations are efforts to influence the labor conditions of media companies, the composition of their workforce, and the cultural products they make. While not engaged directly in the creative labor of media production, media advocacy groups frequently have intervened in the economic and cultural logics of production. In addition, media advocacy groups have contributed their labor to media producers. Frequently this work has been advisory—the reading of scripts, for example, to ensure that the politics of representation within them are not demeaning or harmful—and accordingly, it has been part of the mission especially of identity-based advocacy groups.19 Work on behalf of media companies has more recently extended for some advocacy groups to their policy work, as they have supported positions that, to their critics, do the bidding of media companies at the expense of the communities they ostensibly represent. Critics of the NHMC’s support for the NBC-Comcast merger have read its actions in this light.
This recent synergy of interests in the policy sphere between advocacy groups and media companies is inseparable from the increased financial support advocacy groups receive from media corporations. Fund-raising, as Gonzalez has put it, is the “dirty skeleton in the closet” of advocacy work.20 While many media advocacy groups at first rely on volunteer labor, over time they require a sustained staff who can pursue both long-term and short-term objectives. Thus sustained media advocacy requires sustained access to financial support. Early media advocacy groups were funded by a combination of donations and grants from philanthropic foundations. Action for Children’s Television (ACT), for example, founded in 1969 to combat commercialism in and raise the quality of children’s programming, was supported by individual membership fees, higher donations from “benefactors,” and grants from the Ford Foundation and the Markle Foundation.21 Ford additionally was the primary funder of educational telecasters in the fifteen years leading up to the passage of the 1967 Public Broadcasting Act, and as Jefferson Pooley has demonstrated, Ford from 1998 onward has been one of the biggest benefactors of the media reform movement.22
While grants from philanthropic organizations and individual donations continue to provide substantial support for media advocacy work, they are either inaccessible or inadequate for many organizations. The NHMC, when it formed, relied on the volunteer labor of its members. In the 1990s, it formalized as an organization, secured its 501(c)(3) status as a nonprofit organization, and expanded the scope of its activities. While it initially had been difficult to attract foundation support, the NHMC in the 2000s secured a Ford Foundation grant to support its policy advocacy. Both Ford and the Media and Democracy Fund continue to support the NHMC, the latter also operating as an important advocate for the NHMC’s work with other potential funders.23 Professional and personal networks can be pivotal for media advocacy groups, often making the difference between being visible or invisible to potential funders, regardless of the significance of the organization’s advocacy commitments or its credibility with the community it represents.
While foundation support has been crucial, it also can be insufficient. Thus a number of advocacy groups rely on corporate donations and sponsorship. The NHMC itself receives financial support from media companies like Univision, Entravision, Disney/ABC, and Comcast/NBC-Universal. This funding enables the NHMC’s writers’ program, a screenwriting workshop that prepares Latina/o writers for writing careers in the television industry, and its pitch program, which trains writers to package their ideas as “pitches” and connects them to executives at broadcast and cable networks. The NHMC’s goals with these programs—to bring more Latinas/os into above-the-line creative positions in television—lines up well with the interests of media companies seeking not only potential new series but strategic hires that can underline their dedication to diversity.24
The NHMC also raises money through annual events that fuse the organization’s fund-raising with its mission to promote Latina/o talent and to honor allies and advocates for Latina/o rights. These include an annual gala held in Beverly Hills to honor Latina/o performers; an annual conference that brings together industry personnel, artists, and activists in substantive conversation about contemporary media practices and Latina/o creators and publics; a local impact awards luncheon that honors local talent in the Los Angeles area; and an impact awards reception in Washington, DC, to recognize individuals in the policymaking and legislative sphere who have championed issues central to the NHMC mission. To organize these events, the NHMC has two staff members who spend half their time on fund-raising, along with one dedicated intern to support fund-raising, out of a total staff of six full-time and two part-time employees.25
With these activities, the NHMC operates a sort of para-industry, which trains creative talent and honors the accomplishments of media workers. In return, they strengthen the NHMC’s identity as a Latina/o media advocacy organization and its personal ties with media professionals. Yet they also link the NHMC to companies whose policy objectives often contrast with its own. As both Gonzalez and Nogales insist, NHMC’s record should quell concerns that it is a shill to the companies that help fund its work, as the NHMC has routinely taken positions contrary to their interests. The organization has been a consistent advocate of network neutrality, has filed comments or signed onto comments filed by other public interest groups in support of retaining current media ownership restrictions, and has aggressively opposed some proposed media mergers that it has seen as harmful to its community.
In addition, the NHMC has sought to distance itself from other civil rights organizations that have similarly accepted corporate monies but whose integrity allegedly has been compromised for it. As Juan González and Joseph Torres have argued, civil rights stalwarts like the National Association for the Advancement of Colored People (NAACP) and the League of Latin American Citizens (LULAC), which “used to rail against the injustices of the white media,” now often advocate for policies that support media and telecommunications companies at the expense of the communities they represent.26 For González and Torres, this turn constitutes a “startling and tragic” setback for minority media rights and is directly tied to the financial support provided to these organizations by media corporations.27
Most notably, in June 2013, David Honig and his advocacy group, the Minority Media Telecommunications Council (MMTC), came under attack as being under the sway of their corporate donors.28 Honig is a long-standing media advocacy professional, who, prior to forming the MMTC, had worked for the NAACP on a range of minority media rights campaigns. Honig’s longtime experience as an advocate for minority media rights put him in strong standing to advise civil rights groups on media policy issues. And so when the MMTC—along with the NAACP, LULAC, and others—supported diminished media ownership restrictions, opposed network neutrality, and backed media mergers, other media advocates cast suspicion on the integrity of the MMTC’s position and the influence of corporate donations in its decision making.
The MMTC’s about-face on media ownership issues is of especial concern. When the FCC voted in 2003 and 2004 to diminish its existing ownership restrictions, it faced an enormous public backlash and had its rules remanded by the Third Circuit Court of Appeals for procedural violations and failure to consider how the changes would affect female and minority ownership of broadcast stations.29 When, in 2010, the FCC voted to repeal its newspaper-broadcast cross-ownership rule, the Third Circuit once again remanded the rule to the FCC and admonished it for not considering the change’s impact on female and minority ownership of broadcast stations.30 In each review, as the FCC has asked for comments on its ownership rules, the NHMC, often in collaboration with other advocacy groups, has drawn on the concern over levels of minority ownership to persuade the commission not to diminish or repeal existing regulations. Thus for one of the leading civil rights–based media advocacy groups to argue that media consolidation poses no harm to communities of color, and that the loosening of ownership restrictions could benefit them, is a tremendous opportunity for advocates of deregulation and the media companies who would benefit from it, and a substantial obstacle to public interest advocates who fear the impact of consolidation on the diversity and quality of the media.
The MMTC’s opposition to network neutrality has similarly raised the ire of advocacy groups and elicited accusations that the MMTC and the civil rights organizations with which it works have forsaken a public interest agenda for a corporate agenda. Opposition by the MMTC, NAACP, LULAC, and National Urban League to network neutrality rules indeed echoes the claims of media companies that open Internet provisions would harm communities of color by reducing jobs and inhibiting the expansion of broadband into underserved communities. James Rucker, cofounder of ColorofChange.org, has characterized this advocacy as “the deployment of our civil rights organizations in support of a corporate agenda,” one facilitated by the heavy financial support provided by telecommunications companies to these groups.31 Honig has responded to these charges by reasserting that his organization and other civil rights groups are centrally committed to protecting communities of color, accusing his “netroots” critics of paternalism toward communities of color that in fact misunderstands their interests.32
In 2013, Nogales publicly admonished Honig and the actions of the MMTC, accusing Honig of having become “too chummy with the industry.” Nogales also resigned his position on the MMTC’s board because of concerns over the organization’s ties to media corporations.33 In this, Nogales joined a chorus of media advocacy group leaders who sought to delegitimize the MMTC as an advocate of the public interest broadly, and of the civil rights community specifically, on media regulation issues. In the process, Nogales was able to distance the NHMC from damning accusations that advocacy groups who accept corporate monies become corporate mouthpieces rather than watchdogs or opponents. Such a move was necessary for the NHMC to retain its credibility with its own community and with fellow advocacy practitioners.
Thus part of the current practice of media advocacy groups is to police what counts as an acceptable relationship with a media company and what constitutes advocacy capture, the process by which public interest groups adopt the priorities of their funders over those of their communities; it is to distinguish the kinds of media work that are acceptable forms of media advocacy work. Significantly, it is the ongoing, cumulative nature of media advocacy work that has rendered the recent actions of the MMTC, NAACP, and LULAC so threatening to other advocacy groups and their allies. The power of these groups’ positions on media ownership and network neutrality hails from their clout as long-standing media advocates for communities of color and their past record of reform campaigns to ensure that the media meet the needs of a multiracial public. This work is what makes them credible advocates to policymakers, desirable allies for media and telecommunications companies, and heartbreaking adversaries to other media advocacy groups. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/19%3A_The_Precarity_and_Politics_of_Media_Advocacy_Work/19.03%3A_Media_Advocacy_as_Media_Work.txt |
Precarity—the central theme of this collection—defines media advocacy work in many ways. The NHMC has been motivated by what it has seen as the precarious status of its community. Its work has been premised on the belief that Latinas/os’ security—as well as their political, economic, and social rights—would be affected by their visibility within the media and their ability to access communication technologies. The capacity to enact reform is also precarious, as the outcome of advocacy campaigns rarely hinges only on the solidity of the arguments presented or the extent of popular support for an issue, but also depends on the ideological commitments of the regulatory community, the sway of industry interests, and the political culture at a historical juncture. The ability to do advocacy work is precarious, as groups not only have to continually raise money to support their organization, but consistently have to shore up their informational and reputational capital in order to be legible and credible stakeholders to regulators, other advocacy groups, and their own community. Indeed, the very precarity of media advocacy only underlines how critical it is to honor the ongoing labors of media advocacy groups who continually work amid uncertainty as to outcome as well as to their own survival.
When civil rights organizations become, in the words of Nogales, “too chummy” with the media corporations, when they use their standing as representatives of communities of color to promote the agenda of media companies, they only intensify the precarity of media advocacy work. Not only do they lend support to policies that most likely will diminish the diversity of voices in the public sphere, but they discredit the notion that communities of color have not been, and will not be, served well by deregulation. In this, they mask their media work as media advocacy work and upend the very purpose of media advocacy on behalf of the public interest.
19.05: Notes
1 Alex Nogales, interviewed by Allison Perlman, July 10, 2013, offices of National Hispanic Media Coalition, Pasadena, CA.
2 Philip Napoli makes this very point in his wide-ranging overview of media advocacy scholarship. See Philip Napoli, “Public Interest Media Advocacy and Activism as a Social Movement,” Communication Yearbook 33 (2009): 394–401.
3 Hortense Powdermaker, Hollywood: The Dream Factory (New York: Grosset & Dunlap, 1950); Leo Rosten, Hollywood: The Movie Colony, the Movie Makers (New York: Harcourt Brace, 1941); Thomas Schatz, The Genius of the System: Hollywood Filmmaking in the Studio Era (New York: Pantheon Books, 1988); Todd Gitlin, Inside Prime Time (New York: Pantheon Books, 1983); Horace Newcomb and Robert S. Alley, The Producer’s Medium: Conversations with Creators of American TV (New York: Oxford University Press, 1983).
4 John Caldwell, Production Culture: Industrial Reflexivity and Critical Practice in Film and Television (Durham, NC: Duke University Press, 2008).
5 Vicki Mayer, Below the Line: Producers and Production Studies in the New Television Economy (Durham, NC: Duke University Press, 2011).
6 Seeta Peña Gangadharan, “Public Participation and Agency Discretion in Rulemaking at the Federal Communications Commission,” Journal of Communication Inquiry 33 (2009): 337–353.
7 Becky Lentz and Allison Perlman, “Media Advocacy Practice Produces Media Policy Literacy,” presentation at the International Communication Association Annual Meeting, Seattle, Washington, May 25, 2014.
8 The NHMC, for example, in 1988 filed four petitions against stations in Los Angeles, and in 1989 went after WNET in New York for low levels of Latino employment. See Victor Valle, “Latino Group Challenges TV Licenses,” Los Angeles Times, November 2, 1988, 1; Victor Valle, “Latino Coalition’s Bid For KTTV: Full Assault, Long Odds,” Los Angeles Times, November 10, 1988, H1, H13; Victor Valle, “Latinas/os Claim Job Bias at KCBS,” Los Angeles Times, December 31, 1986, J1, J2; “NHMC Signs Landmark Agreement with KABC-TV,” press release, November 11, 1993, Box 7, Folder 1, Papers of the National Hispanic Media Coalition, University of California, Los Angeles (hereafter NHMC Papers).
9 For documents relating to the Disney boycott, see Box 33, Folders 1–2, NHMC; for a discussion of the struggles of Spanish language broadcasting, see Allison Perlman, Public Interests: Media Advocacy and the Struggles over U.S. Television (New Brunswick, NJ: Rutgers University Press, forthcoming), chapter 6.
10 Notice of Appeal, NHMC v. FCC, filed October 20, 1992; Notion of Intention to Intervene, NHMC v. FCC, filed November 5, 1992, Box 32, Folder 2; Settlement Agreement, Entered into Between National Hispanic Media Coalition and the Univision Television Group, Inc., May 20, 1994; Letter from Enrique Baray to Armando Durón, October 18, 1995, Box 31, Folder 3, NHMC Papers.
11 Jessica Gonzalez, interviewed by Allison Perlman, April 17, 2014, offices of National Hispanic Media Coalition, Pasadena, CA.
12 Interview with Nogales; Electronic Filing, Correction to Ex Parte Presentation Letters, filed by Jessica Gonzalez, September 21, 2010, http://apps.fcc.gov/ecfs/comment/view?id=6016055968. See the letter to Genachowski outlining the terms of the MOU, http://apps.fcc.gov/ecfs/comment/view?id=6015694189.
13 Greg Braxton, “A White, White World on TV’s Fall Schedule,” Los Angeles Times, May 28, 1999, 1.
14 Interview with Nogales; Greg Braxton, “Groups Join to Protest Exclusion, Television: Coalition Forms in Response to the Absence of Minorities on New Shows in Prime Time This Fall,” Los Angeles Times, June 25, 1999, 1; Paul Bernstein and Michael Schneider, “NAACP, NBC Reach Pact,” Daily Variety, January 6, 2000, 1; Lisa de Moraes, “TV Networks Adding Some Color For Fall: New Minority Roles Receive Little Applause,” Washington Post, May 21, 2000, A01.
15 Interview with Nogales.
16 Letter to Chairman Genachowski from Benjamin Todd Jealous, Marc H. Morial, and Rev. Al Sharpton, December 16, 2010, http://apps.fcc.gov/ecfs/comment/view?id=6016064629; Letter to Chairman Genachowski from Karen K. Narasaki, December 15, 2010, http://apps.fcc.gov/ecfs/comment/view?id=6016064415.
17 Josh Silver, “Comcastrophe: Comcast/NBC Merger Approved,” Huffington Post, January 18, 2011, www.huffingtonpost.com/josh-silver/comcastrophy-comcastnbc-m_b_810380.html.
18 Erwin G. Krasnow, Lawrence D. Longley, and Herbert A. Terry, The Politics of Broadcast Regulation (New York: St. Martin’s Press, 1982), 57–58.
19 See Kathryn C. Montgomery, “Special Interest Citizen Groups and the Networks: A Case Study of Pressure and Access,” in Telecommunications Policy Handbook, ed. Jorge Reina Schement, Felix Gutierrez, and Marvin A. Sirbu, Jr. (New York: Praeger, 1982), 241–254.
20 Interview with Gonzalez.
21 Donald Guimary, Citizens’ Groups and Broadcasting (New York: Praeger, 1975), 125–126.
22 Jefferson Pooley, “From Psychological Warfare to Social Justice: Shifts in Foundation Support for Communication Research,” in Media and Social Justice, ed. Sue Curry Jansen, Jefferson Pooley, and Lora Taub-Pervizpour (New York: Palgrave-MacMillan, 2011), 211–240.
23 Interview with Gonzalez.
24 Ibid.
25 Ibid.
26 Juan González and Joseph Torres, News for All the People: The Epic Story of Race and the American Media (London: Verso, 2012), 371.
27 Ibid., 372–376.
28 Jason McLure, “Civil Rights Group’s FCC Positions Reflect Industry Funding, Critics Say,” Center for Public Integrity, June 6, 2013, www.publicintegrity.org/2013/06/06/12769/civil-rights-groups-fcc-positions-reflect-industry-funding-critics-say.
29 Prometheus Radio Project v. FCC, 373 F.3d 372 (2004).
30 Prometheus Radio Project v. FCC, 652 F.3d 431 (2011).
31 James Rucker, “Net Neutrality, Civil Rights, and Big Telecom Dollars,” Huffington Post, September 10, 2014, www.huffingtonpost.com/james-rucker/net-neutrality-civil-rights-orgs_b_5796944.html.
32 “David Honig Pushes Back against the ‘Digital Activists,” Field Negro, July 30, 2014, http://field-negro.blogspot.com/2014...l#.VBD1i1ZnylK.
33 McLure, “Civil Rights Group’s FCC Positions Reflect Industry Funding.” | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/19%3A_The_Precarity_and_Politics_of_Media_Advocacy_Work/19.04%3A_Conclusion__The_Precarity_and_Politics_of_Media_Advocacy_Work.txt |
Across the world, trade unions have played a major role in efforts by workers to improve their conditions, defend their rights, and promote social justice in people’s working lives. Yet in the recent “turn to labor” in media and cultural studies, there has been little sustained consideration of unions.¹ The collective action and bargaining offered by unions are crucial in providing a means of limiting the problematic working conditions that, as a number of researchers have shown, are apparent in much media work, in spite of easy and flawed assumptions that the media industries provide high-quality or “easy” jobs.²
• 20.1: Introduction
Introduction to the goal of this chapter: to explore the efforts of professional/trade organizations to improve labor rights and conditions for writers in media, both within a country and internationally, using a case study of the Writers Guild of America.
• 20.2: Problems Facing Organized Labor in the Media Industries
Some of the problems facing organized labor in the media industries, including differences in precarity between union and non-union members of media industries and tension between the concept of craft unions/guilds versus general unions.
• 20.3: The Writers Guild of America in the National Context
Challenges faced by the WGA, including the tension between its status as a guild and the fact that it organizes union-like activities such as strikes and protests; the difficulty in collecting a group of individuals who normally work alone; and its need to jockey for concessions with other guilds and unions in the media industries.
• 20.4: Going Global – Guilds in an Era of Internationalization
Examining international connections between writers’ guilds, in the contexts of both professional solidarity and limiting potential outsourcing.
• 20.5: Conclusion
Summary of the tensions and contradictions involving writers’ guilds, including internationalization and competition between established, guild-member writers and aspirants.
• 20.6: Notes
20: Internationalizing Labor Activism Building Solidarity among Writers Guilds
Across the world, trade unions have played a major role in efforts by workers to improve their conditions, defend their rights, and promote social justice in people’s working lives. Yet in the recent “turn to labor” in media and cultural studies, there has been little sustained consideration of unions.1 The collective action and bargaining offered by unions are crucial in providing a means of limiting the problematic working conditions that, as a number of researchers have shown, are apparent in much media work, in spite of easy and flawed assumptions that the media industries provide high-quality or “easy” jobs.2 The labor precariousness that is the subject of this collection would be much less likely to prevail in a situation where strong unions were able to negotiate collectively on behalf of workers. In addition, the best trade unions strive to counter inequalities and exclusions based on gender, class, ethnicity, and other dimensions of social power, and these too are real problems in the media industries. Yet many media workers feel uncertain about the value of trade unions, or anxious that affiliation or identification with them will lead to the loss of work. This chapter concerns efforts by professional and trade organizations to defend and improve the rights and conditions of writers as a community of workers in the media industries, both within particular nations and internationally. It explores these issues via a case study of the Writers Guild of America (WGA).
However, our concerns are not confined to the borders of the United States. We begin by discussing various obstacles and tensions facing organized labor in the media industries. Although here we focus on the United States and the United Kingdom, many of these issues can be found internationally. We then discuss some of the ways these issues have played out historically in the specific example of the WGA, before turning to a recent significant development that raises crucial questions about media labor in an era of internationalization or, as some would have it, “globalization”: increasing efforts by the WGA to work with other writers’ labor organizations abroad, not only to prevent outsourcing of work to cheaper locations (of course a problem in many industries, media and otherwise, in the global era), but also to build solidarity. Yet some of the same problems regarding tensions between solidarity and exclusion, fairness and privilege, can be found in the context of international media labor organization, though with intriguing new dynamics that we explore below. Those new dynamics can be properly understood only when explained in the context of problems facing organized labor in the media industries, and we begin this chapter with a historical perspective on these issues. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/20%3A_Internationalizing_Labor_Activism__Building_Solidarity_among_Writers_Guilds/20.01%3A_Introduction.txt |
In many countries, media industries have been fairly highly unionized for many years. In The Cultural Front, Michael Denning tells the story of how culture came to be a major ground for leftist activism in the United States during the 1930s and 1940s,3 and he shows how this led to the American working class making its mark on dominant cultural institutions for the first time, but also how it led to the formation of organized labor institutions in the sphere of culture. For Andrew Ross,4 Denning’s perspective is a useful reminder that the industrialization of culture in the twentieth century was an opportunity for creative labor more than a threat. Industrialization made culture an object of mass production, and unlike workers in other industries, media workers could exert an influence on the shape and nature of the product. By contrast, Ross points out, “the non-commercial arts have long been a domain of insecurity, underpayment, and disposability.”5 In other countries too, the rise of media industries was accompanied by significant levels of unionization. For example, the networks that traditionally dominated British broadcasting (the BBC and ITV) were unionized from their formation in the 1920s and 1950s, respectively,6 and so was U.K. journalism (the National Union of Journalists [NUJ] was founded in 1907). The U.K. Musicians Union was formed in 1921 and by the end of the 1990s had over 31,000 members.7
Across the world in the early twenty-first century, however, media trade unions of all kinds are facing significant challenges. Attacks on trade unions in general, launched with renewed vigor starting in the 1970s and 1980s, have continued to the present day across the globe, and in many countries union membership is in steep decline.8 This, combined with the marketization of media industries enabled by government deregulation programs, has led to a real reduction in the influence of media labor unions. The power of trade unions in the media industries has almost uniformly diminished, professionally, economically, culturally, and politically. Examples can be seen in television, journalism, and music.9 Rates of unionization are extremely low in the independent television production companies that have come to occupy a key place in the European television market. Journalists’ unions have been significantly reduced in number and power, not only because of the technological “advances” of digitalization, but also because of changing employment laws and journalists’ embrace of notions of “professionalism,” which has drawn entrants to the occupation away from unions.10
Musicians’ unions illustrate some of the problems facing collective worker organization in the new media landscape in a way that suggests the dangers of precariousness for screen workers. Few workers are employed permanently as musicians, and musical labor more often than not is carried out on a freelance basis, and therefore difficult to unionize. Musicians’ unions play an important role in campaigning around various issues—for example, the regulation of live performance. But the collective bargaining over pay and conditions that is at the heart of modern trade unionism is elusive in the case of musicians outside live entertainment and orchestral work. What’s more, some of the issues that musicians’ unions take up on behalf of their members can have detrimental effects on musicians outside the union. For example, those who have already attained the status of authorship, and who are therefore more likely to gain fuller compensation through rights, are more likely to be members of a union (among other reasons, because they are more likely to feel that it is worthwhile to pay their dues). Income from “rights” of various kinds provides an important supplement to other income for many musicians and other precarious creative workers—though few workers can actually make a living from rights alone. It is perfectly understandable that unions and other associations of workers work to increase such income for their members by campaigning for stricter enforcement of intellectual property. Yet this can have the effect of stifling public culture and making content creation more expensive for workers who do not have the protection of a big company. This illustrates the potential tensions between goals that unions pursue on behalf of their members (payment via rights) and other potentially legitimate goals that might favor nonmember media workers (more open access to culture). Such tensions between solidarity and exclusion recur constantly and internationally.
The fight for improved conditions for media workers faces other challenges even within the organized labor movement. The coexistence of the terms union and guild indicates some of the tricky issues regarding different kinds of workers, and different approaches to how they might best be protected by worker organizations. There are tensions in the media and communication industries between “craft unions,” on the one hand, and those oriented toward general worker solidarity, on the other. There are also tensions between those organizations that represent above-the-line or “creative talent” workers, such as writers, actors, and directors, and those representing below-the-line “craftspeople,” technical or support workers.
Worker organization in the media industries is divided between, on the one hand, craft unions and guilds, who often aim primarily to protect the pay and conditions of existing members who have gained entry to a limited field; and on the other, general unions that adhere to inclusive goals of solidarity and equality, and see themselves much more as defending workers as a whole. This in turn relates to a fundamental problem underlying all modern trade unionism: the tension between the pressure to act as a “businesslike service organization” or as an “expression and vehicle of the historical movement of the submerged laboring masses.”11 As Alan Paul and Archie Kleingarter have shown in the most important study of the topic, the unions or guilds representing “creative” above-the-line talent in the U.S. film and television sectors managed to expand membership and bargain powerfully for their members in the late twentieth century, in spite of regulatory and technological changes that might have harmed their effectiveness.12 Some analysts have responded to the unfortunate connotations of above-the-line and below-the-line, terms derived from Hollywood accounting practices and seeming to suggest a hierarchy of labor, by treating above-the-line workers as somehow inherently privileged or more “creative” compared with technical and other workers. But in the media industries some technical workers enjoy very good pay and conditions, and many above-the-line workers suffer hardship.
Craft unions have some ambivalent features, as Vincent Mosco and Catherine McKercher have shown in a valuable account of labor organization in media and communication industries. Craft solidarity, they write, has “at times worked against the push toward mass unions, and at other times has encouraged it.”13 The International Typographical Union (ITU), which represented printers in the U.S. newspaper industry until 1986, for example, encouraged workers to identify with their union and to see it “as the institution that would provide them with a good living.”14 But Mosco and McKercher also recognize that craft solidarity can be destructive, and that the ITU, for example, tried the patience of workers as it grew into a more bureaucratic and professional bargaining institution concerned with “jurisdiction over the tools of the trade” to the exclusion of protection and promotion of the craft itself.
What is needed is strong union representation ensuring good working conditions and rights across all types of media work, nationally and internationally. Yet social and cultural changes have negatively affected trade unions in general, including media unions. One way of understanding this is via the concept of individualization, whereby workers tend to see organizations, and jobs, as opportunities for self-development rather than sources of commitment. For the most widely cited advocate of this concept, Ulrich Beck, individualization offers some new freedoms in that people become independent of restrictive traditional ties, but it also leads to competitiveness and isolation.15 In the eyes of some commentators, this leads to “an individualistic and self-centered culture of contentment that sees no virtue in forms of collective association and solidarity.”16 Such developments perhaps help to explain how, in the contemporary media industries, in Susan Christopherson’s words, “personal networks are recognized as the central mechanism both for individual career advancement and risk reduction.”17
Organizations representing creative workers face all these challenges. They also face a challenge concerning how they are perceived more widely. In a fine analysis of changes in the U.S. film and television industries, Christopherson shows how middling budget productions are being eroded both by the huge demand for cheap programming in the era of multichannel television and by the blockbuster syndrome in movies, and how this has led to a strengthening of “defensive exclusionary networks”18 that dominate access to the best jobs. Are guilds of creative workers examples of such exclusionary networks, reinforcing the privilege of the well educated and successful? This question of privilege cannot be separated from dynamics of inequality related to class, race, ethnicity, and gender. In the remainder of this chapter, we explore these issues by examining efforts by writers of film, television, and streaming media to defend—or better procure—their rights as employees within the major media industries, first by looking at some of the obstacles faced by U.S. writers in their own national context and then turning to their efforts to establish strong global connections among writers’ organizations. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/20%3A_Internationalizing_Labor_Activism__Building_Solidarity_among_Writers_Guilds/20.02%3A_Problems_Facing_Organized_Labor_in_the_Media_I.txt |
In early November 2007, certain quarters of Los Angeles transformed overnight into walking districts. For the next five months, five days a week, dozens of writers, often spectacled, wearing jeans and T-shirts and always with picket signs, walked for hours in front of various gates of the major Hollywood studios. Across the country, dozens more in New York bundled up and braved the cold to protest their rights of labor and rates of compensation. These professional film and television writers walked en masse to protest stalled negotiations with the American trade organization the Alliance of Motion Picture and Television Producers (AMPTP). For the first time in nineteen years, the Writers Guild of America (WGA) was on strike. Nationally, a poll conducted two weeks after negotiations broke off showed that 63 percent of Americans sided with the striking workers (with 4 percent favoring the studios, 33 percent unsure).19
It is rare in the United States to see striking workers marching in a number of areas across the two largest cities in the country. Even more notable was the fact that these employees were neither blue-collar laborers nor white-collar workers. They were no-collar workers.20 Unlike earlier strikes, this time writer-producers and showrunners also walked the picket lines, arguing that they could not separate their work as producers from their role as writers. The guild leadership specifically targeted showrunners early in the negotiations to get their support, not just for labor action but to read the letter of the law in such a way that their role as producers could not be separated from their role as writers. While as producers they were part of management, as writers they were employees of the studio. While some faces were familiar—Tina Fey, Rob Reiner—others had names that were familiar to audiences: Norman Lear and James L. Brooks. Still others were attached to beloved products that suddenly disappeared from homes across the globe. Writers were now positioned—in their role marching around the outside of studio buildings—as industry workers fighting for their rights.
The Writers Guild of America was first established as the Screen Writers Guild in 1933, though it was not granted a contract until 1942. The WGA, which comprises East and West branches, is the bargaining agent for professional writers who craft film, television, news, animation, streaming media, and video game scripts for American signatory companies. The Writers Guild has gone on strike six times, in 1959–1960, 1973, 1981, 1985, 1988, and 2007–2008. Three of these industry-wide walkouts were protracted, lasting many months. As they had in every previous strike, in 2007–2008, these American writers marched in circles and demanded their rights, not as artisans but as workers in a media industry. This time, though, because of the globalization of film and television distribution, as well as the rise of YouTube—where many striking writers went to speak directly to audiences—more people than ever before were aware of a strike among working writers. Not just in the United States, but globally. And not just audiences, but other writers as well.
For the writers under its protection, the WGA as a guild provides union-oriented services: it convenes and mobilizes members, addresses their concerns, negotiates and enforces contracts, lobbies on behalf of its members, and represents the face of screenwriters to the outside world. But it is its final directive—preserving the art and craft of writing—that most clearly illuminates the subtle difference between a union and a guild. The WGA sees its protection, teaching, and preservation of the work of writing as the additive dimension that distinguishes it from a traditional trade union.
Yet during moments of economic crisis or labor negotiations, writers often feel compelled to define themselves as a union first and foremost. Bob Barbash, a writer on Zane Grey Theater, explained how this perception played out during a strike in 1960: “A tremendous amount of people in the Guild . . . resent the word ‘union.’ . . . [Every] morning I had to be carrying a picket sign in front of MGM. Now that is not a Guild. That’s a union, man. When you are walking there and you are trying to stop people from crossing the line. We are an unusual group because we like to think of ourselves as [part of a] super, upper [tier of] intelligence. That we don’t work on a loading dock . . . but if you are going to have a union, you are a union.”21 In contrast, the term guild implies a focus less on working conditions and more on championing the artistry of the profession. The difference is not merely one of terminology: it has resulted in a recurring tug-of-war across the entertainment industries between different groups of writers and sometimes even within an individual writer’s conception of what they do and how their interests ought to be represented.22 The internal friction is captured in shifting definitional terms such as artist, worker, creative, laborer.
Writers must join the guild if they have surpassed a certain quantity of work with a company that has signed as a contractual partner on the guild’s collective bargaining agreement. A signatory company can be as vast as a multinational corporation or as limited as a small pro-union production company. An associate writer amasses units to gain full membership, and today writers must belong to either the WGA East (which uses the acronym WGAE) or the Writers Guild West (which prefers WGAw), depending on geography. The guild’s stated objectives are voluminous. It contracts minimum rates for specific types of work, determines writers’ screen credits, ensures payment of residuals, provides pensions and health benefits for members, engages in national policy debates concerning writers’ interests, and provides continuing education for members and the community. Some writers have seen their induction into the guild as a sign of having “made it” in the industry. Others have felt membership to be a weighty burden foisted upon them. And still others have paid little attention to what membership meant. Then there are those who view membership as a life raft. Barbara Corday, creator of Cagney & Lacey, expressed deep gratitude for the benefits afforded to veteran writers: “First of all, having residuals. Lifetime medical insurance as a backup to Medicare, as a secondary insurance. How many people outside of Congress have things like that? It’s just phenomenal.”23
Corralling this disparate group of workers, however, is an arduous task. The guild brings together thousands of individuals who predominantly perform solitary work. As Hal Kanter, creator of the series Julia, noted in the 1970s, “We writers are, collectively, a strange group of creatures and it’s a frequent source of amazement to me that the Guild is such a well-run zoo!”24 John Furia Jr., writer for The Singing Nun and president of the WGAw from 1973 to 1975, laughed as he pointed out, “We are the most individualistic group to band together.”25 Phyllis White, who worked on writing teams for various television series from the 1950s through the 1980s, noted the paradox of singular writers with unique voices aligning for a collective cause: “It’s a Guild of individuals as no other union is. You’ve got the Teamsters and there are a certain number of Teamsters who do the same job. . . . They do the same hours. They do the same thing. We don’t. . . . Trying to amalgamate this group . . . [of] nearly 5,000 into one union now is horrendous. It’s amazing that it works at all.”26 White’s sweeping claims around the specialness of writers’ work are problematic: many trade unions cover diverse members with distinct job descriptions, and the work of writers is not as rarefied as she proclaims.27 And yet the notion of collecting a community of workers who usually work alone does pose distinct difficulties.
Another major challenge for the Writers Guild is that it coexists with a number of other guilds and unions in the media industries. The other groups that negotiate with signatory companies include the Directors Guild of America (DGA), which represents directors, assistant directors, unit production managers, and production associates; the Screen Actors Guild–American Federation of Television and Radio Artists (SAG-AFTRA), which represents actors, extras, broadcast journalists, and puppeteers, among others; and the International Alliance of Theatrical Stage Employees (IATSE), which represents a diverse set of industry workers, from electricians to set carpenters, makeup artists, prop masters, cinematographers, editors, and art directors. The other three organizations service vastly larger constituencies than the WGA, and have needs so diverse that a united front proves tricky—especially when it comes time to negotiate with the monolithic Alliance of Motion Picture and Television Producers (AMPTP). The AMPTP is an enormous bargaining unit that digests the concerns of hundreds of production companies, networks, and studios and then delivers a proposal—representing the united group’s interests—to the negotiating table. Whereas in standard bargaining a union tries to garner advantage by playing off one company against another, the AMPTP positions itself so that the three creative guilds must jostle with each other, grabbing for scraps at the table. This tactic, called reverse pattern bargaining, forces each guild into what one member called “a kind of a chess game between the three unions.”28 | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/20%3A_Internationalizing_Labor_Activism__Building_Solidarity_among_Writers_Guilds/20.03%3A_The_Writers_Guild_of_America_in_the_National_C.txt |
As indicated earlier, an important way a guild might define its work differently than a union is by emphasizing promotion of the profession or craft. This has spatial dimensions that have changed in recent decades. Where once a union would look only for local, regional, or national solidarity, in an era of globalization of the media industries, solidarity for the WGA must be threefold: within their own union, member to member and between East and West; among the WGA, the DGA, SAG-AFTRA, and IATSE; and as we explore in this section, among different countries and communities of professional writers that work for the media industries.
This international dimension is not entirely new. For most of its eighty-year existence as a trade union, the Writers Guild has offered professional support to developing guilds and associations in other countries, guiding media and cultural workers in other countries on how to respond to changes in the industry The Writers Guild of America has often called for solidarity not only among its members, but also from aspirants and fellow professional screenwriters across the globe. But in this increasingly globalized era of media production, this aspect has intensified. This was particularly noticeable during the 2007–2008 strike, when the guild made it clear that it would hold accountable any writer who broke the strike. WGA members spoke with film students, instructing them not to take writing jobs with studios as screenwriters. At stake for any writer, locally or globally, was any chance of joining the union. But the guild did not stop at U.S. borders. The WGA asked screenwriters in countries affiliated with the American guild through the International Affiliation of Writers Guilds not to work for American studios during the strike as an act of global solidarity. Having this kind of control of the market on scripts was critical to a successful strike. By including prospective writers and defining them as allies, they increased the chances of unity during the strike.
There is a contradiction in this behavior, however: this unity only confirmed that pathways for international workers into the industry—especially the American industry—are barely open. In this case, solidarity can reaffirm exclusion. And this type of international cooperation is often about leveraging power more than benevolent mutual support. Kevin Sanson argues that global cities offer opportunities for advanced capitalist countries—most notably American but also British and Australian companies—to use their diverse locales, functional technical resources, and skilled practitioners at budget prices.29 The price of labor is significantly cheaper in part because international production labor is rarely unionized. The easiest way to keep costs low is to film overseas, outsourcing production and postproduction as much as possible to avoid the high costs of unionized labor. The economic and geographic structures of multiplatform global entertainment conglomerates have made transnational production the norm in what are still considered by most national and international audiences to be “Hollywood” productions.30
While much of so-called Hollywood production labor is now regularly outsourced across international borders, writing has generally stayed in the United States. There are a few jobs, including screenwriting, that tend to be culturally specific: not all jobs cross borders easily or comfortably. The specificities of language and idiom, trends in narrative structure, and cultural references and social issues make writing for a global audience particularly daunting. Companies might be eager to outsource writing to other Anglophone countries, but the reality is that this still rarely occurs. And yet the WGA seems aware that it is only a matter of time before global competition becomes more fierce. Like many other industries, major media corporations are increasingly prone to outsource work to lower-cost regional media capitals. American visual effects and digital postproduction workers’ recent organizing campaigns serve as a legitimate example of U.S. labor’s anxiety about jobs going overseas. Arguably, these developments can provide opportunities for labor in Prague or Budapest or India to earn pay, build skills, build infrastructure, and achieve professional renown. And those jobs could include writing jobs.
The WGA regularly ventures overseas for conversations with other national writers’ guilds and related organizations. While part of the mission is solidarity, they also have hopes of professionalizing their international counterparts in the hope of limiting outsourcing. This represents a model of modified inclusion, something WGA West vice president Howard Rodman explained as “we can’t give you what we have, but we will help you navigate the waters to get there—in the meantime by helping you secure better wages, we will ensure that our native industry does not see your labor as enticing.”31
Other writers’ guilds exist around the world, primarily in economically developed countries. South Africa, Israel, and Australia have strong screenwriters’ guilds. In the United Kingdom, the WGGB is part social club and part professional organization. Greece and Italy are establishing their guilds as social clubs first (with the hope that professionalization will follow).
The WGA has built connections with screenwriters’ guilds from around the world and continues to build more, in part through professional organizations like the International Association of Writers Guilds.32 Granted, the tie with each union, association, or professional organization shifts based on the changing nature of labor relations for each individual country. One example of this is in the case of New Zealand. Though writers in New Zealand have been unionized for over forty years, the Employment Contract Act of 1989 was a terrible blow to creative labor in the country. The act transformed the nature of labor in New Zealand, terminating any chance that media workers would hold rights to residuals. Norelle Scott, a member of the New Zealand Writers Guild, explained how the act decimated the power of creative labor—and it was only writers’ affiliation with the International Association of Writers Guilds that kept its membership focused on whatever rights they still controlled.33 It was through the strength of international partnerships that the New Zealand Writers Guild began to rebuild after this devastating blow. With their ties to the International Association of Writers Guilds, the New Zealand Writers Guild made steps forward, setting agendas and structures for international coproductions and discussing strategies for developing free trade agreements.
Writers in Greece, Italy, and France have over the years developed clear agendas as well—whether or not they are specifically stated. As U.S. formats and sensibilities are exported and transferred around the world, writers who work elsewhere are eager to import professional rights. Many hope in time not only that increased coproductions and transnational industry shifts will lure production dollars but that preproduction will also come to their countries. And with this importation, there is hope that the rights of professional writers will be redefined. American screenwriters see part of that process as making sure local writers protect themselves from their own native industry, no matter what form that native industry assumes.
The WGA has passed on to professional screenwriters across the world their frustration with media production and with the fact that directors, producers, and actors are nearly always paid better. In addition, writers rarely have much control over the way their scripts are used. Spanish screenwriter Agustín Díaz Yanes said, “The worst comment you can ever hear when you go and see a producer is when they say to you: ‘The screenplay is essential.’ That’s when you know they pay peanuts, if they pay at all!”34 While it is not the sanctity of the screenplay that matters, Yanes’s comment about the place of the writer on the lower end of the creative hierarchy speaks to a frustration widely shared among writers working in the global media industries.
In a global media production landscape, the unique dynamics of individual careers can obscure the trends of the media industries. It is not only the power of the major conglomerates at work but also the needs of trade organizations that guide debate and discussion, as well as actions that define patterns of inclusion and exclusion and hierarchies of power. As Bridget Conor observes in her study of labor problems surrounding the Lord of the Rings trilogy (filmed in New Zealand), extraordinary displays of “empire in action” demand our attention as we study precarious labor in a global economy.35 With the expanding frontiers of media production—even within the economy of a single film or film series—there is both a fear of what could happen if unionization is quashed on a global level and hope for what could happen if an alliance across countries were solidified among writers’ guilds. | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/20%3A_Internationalizing_Labor_Activism__Building_Solidarity_among_Writers_Guilds/20.04%3A_Going_Global__Guilds_in_an_Era_of_Internationa.txt |
The challenges of internationalization are substantial for a national union. The WGA offers one example of how a union has struggled toward regional, national, and global solidarity. But what about those who are yet to be included among the paid workers? Across the globe, professional screenwriters are negotiating the tricky waters of this international production flow. When considering media workers, it is critical to think about the role of national trade organizations and the role these labor groups play as media cross borders. Guilds believe they can ease the processes of production. Many now operate alongside city and regional governments in efforts to attract investment. But access to labor organizations is possible only for people who have established themselves within the industry. And access to the most powerful of these organizations—those in the United States—is limited to people who have already succeeded in selling a script. The aspirants—including international screenwriters trying to make it in their own countries—realize that they are both potential allies and potential competition for those already in coveted A-list writer roles. This further illustrates the tensions and contradictions at work among craft unions and guilds and how their efforts to protect workers can also serve as exclusionary devices. Nevertheless, the WGA offers an example of relatively successful collective worker organization in the media industries. That success now needs to be extended internationally, across different media jobs and social classes. But only by addressing the kinds of tensions and contradictions outlined above can organized labor fulfill its historical mission of protecting media workers.
20.06: Notes
1 Much of the recent critical research on media work has actually been addressed to “cultural work” or “creative labor,” and one reason for this choice of terms is the way work has been understood (or neglected) by policymakers and academics interested in the creative industries and the “creative economy.” See Mark Banks and David Hesmondhalgh, “Looking for Work in Creative Industries Policy,” International Journal of Cultural Policy 15.4 (2009): 1–16. For simplicity’s sake, and because of the topic of this book, we use the terms media work and media workers here. For a more detailed discussion of the relations among the concepts media industries, cultural industries, and creative industries, see Hesmondhalgh, The Cultural Industries (Los Angeles: Sage, 2013), 23. On the relations among the concepts of media work, cultural work, and creative labor, see Hesmondhalgh and Sarah Baker, Creative Labour: Media Work in Three Cultural Industries (New York: Routledge, 2011), a source some of this chapter draws on.
2 Among the notable contributions to the “turn to labor” in media and cultural studies discussing or providing evidence of such problematic conditions, see Andrew Ross, Nice Work If You Can Get It: Life and Labour in Precarious Times (New York: New York University Press, 2009); Angela McRobbie, “Clubs to Companies; Notes on the Decline of Political Culture in Speeded Up Creative Worlds,” Cultural Studies 16 (2002); Mark Banks, The Politics of Cultural Work (New York: Palgrave Macmillan, 2007); Vicki Mayer, Below the Line: Producers and Production Studies in the New Television Economy (Durham, NC: Duke University Press, 2011); Matt Stahl, Unfree Masters: Recording Artists and the Politics of Work (Durham, NC: Duke University Press, 2013); and Mark Banks, Rosalind Gill, and Stephanie Taylor, eds., Theorizing Cultural Work: Labour, Continuity and Change in the Cultural and Creative Industries (New York: Routledge, 2013). Important research on media labor has by no means been confined to media and cultural studies. A ground-breaking collection from industrial relations studies is Lois Gray and Ronald L. Seeber, eds., Under the Stars: Essays on Labor Relations in Arts and Entertainment (Ithaca, NY: ILR Press, 1996); see also Alan McKinlay and Chris Smith, eds., Creative Labour: Working in the Creative Industries (Basingstoke: Palgrave Macmillan, 2009). The contributions of geographer Susan Christopherson have been valuable; see below.
3 Michael Denning, The Cultural Front (New York: Verso, 1997).
4 Ross, Nice Work If You Can Get It, 19.
5 Ibid., 21.
6 Asa Briggs, The History of Broadcasting in the United Kingdom, vol. 5 (Oxford: Oxford University Press, 1995), 381.
7 Dave Laing, “Musicians’ Unions,” Continuum Encyclopedia of Popular Music of the World, ed. John Shepherd et al. (London: Continuum, 2003).
8 For a recent analysis of the problems of the U.S. labor movement, see Stanley Aronowitz, The Death and Life of American Labor: Towards a New Workers’ Movement (New York: Verso, 2014).
9 Alan McKinlay, “Making ‘the Bit between the Adverts’: Management, Accounting, Collective Bargaining and Work in UK Commercial Television, 1979–2005,” in Creative Labour, ed. McKinlay and Smith.
10 Meryl Aldridge and Julia Evetts, “Rethinking the Concept of Professionalism: The Case of Journalism,” British Journal of Sociology 54.4 (2003).
11 Will Herberg, “Bureaucracy and Democracy in Labor Unions,” Antioch Review 3, cited in Peter Fairbrother and Edward Webster, “Social Movement Unionism: Questions and Possibilities,” Employment Responsibilities and Rights 20 (2008): 309, who also quote an alternative formulation of the tension: “sword of justice” or “vested interest.”
12 Alan Paul and Archie Kleingarter, “The Transformation of Industrial Relations in the Motion Picture and Television Industries: Talent Sector,” in Under the Stars, ed. Gray and Seeber. Gray and Seeber’s collection is the most detailed and valuable study of media labor organizations, but its cases are confined almost entirely to the United States.
13 Vincent Mosco and Catherine McKercher, The Laboring of Communication: Will Knowledge Workers of the World Unite? (Lanham, MD: Lexington Books, 2008), 82.
14 Ibid., 104.
15 Ulrich Beck, The Brave New World of Work (Cambridge: Polity Press, 2008), 94.
16 Robert Taylor, “The Future of Employment Relations” (Swindon: Economic and Social Research Council, 2001), quoted in Richard Saundry, Valerie Antcliff, and Mark Stuart, “‘It’s More Than Who You Know’—Networks and Trade Unions in the Audio-Visual Industries,” Human Resource Management Journal 16.4 (2006): 378.
17 Susan Christopherson, “Beyond the Self-Expressive Creative Worker: An Industry Perspective on Entertainment Media,” Theory, Culture & Society 25 (2008): 89.
18 Ibid.
19 James Welsh, “WGA Lauds Public Support Polls,” Digital Spy, November 15, 2007, www.digitalspy.com/tv/ustv/news/a79902/wga-lauds-public-support-polls/.
20 In the United States, different socio-economic groups are often reductively indicated by the clothing they wear to work: formal white collars for professionals, (originally denim) blue collars for manual workers. “No collar” indicates the wearing of T-shirts by those who work from home or in the self-consciously informal IT industries. While the dress code is casual, the workload and working hours are often very demanding.
21 Bob Barbash, interview by the Writers Guild Oral History Project (Los Angeles: Writers Guild Foundation, February 24, 1978), 7, Writers Guild Foundation Archive, Los Angeles, CA.
22 The battle over self-definition is a recurring theme in Miranda J. Banks, The Writers: A History of American Screenwriters and Their Guild (New Brunswick, NJ: Rutgers University Press, 2015). This chapter draws upon research for that book and material that appears in it.
23 Barbara Corday, interview with Banks, August 30, 2013.
24 M.W., “Kanter Adds Dimension to Hyphenated Career: Writer-Prod-Dir-Emcee,” WGAw Newsletter, December 1967, 7, Writers Guild Foundation Shavelson-Webb Library, Los Angeles.
25 John Furia Jr. and David Rintels, interview by the Writers Guild Oral History Project (Los Angeles: Writers Guild Foundation, May 3, 1978), 44.
26 Robert White and Phyllis White, interview by the Writers Guild Oral History Project (Los Angeles: Writers Guild Foundation, spring 1978), 22.
27 On the politics of such distinctions, see Jason Toynbee, “How Special? Cultural Work, Copyright, Politics,” in Theorizing Cultural Work: Labour, Continuity and Change in the Cultural and Creative Industries, ed. Mark Banks, Rosalind Gil, and Stephanie Taylor (Abingdon: Routledge, 2013); Mayer, Below the Line; and especially Stahl, Unfree Masters.
28 Marc Norman, interview with Banks, June 2011.
29 Kevin Sanson, “Production Service Firms and the Spatial Dynamics of Global Media Production,” paper presented at annual meeting for the Society for Cinema and Media Studies Conference, Seattle, Washington, 2014.
30 Toby Miller, Nitin Govil, John McMurria, Richard Maxwell, and Ting Wang, Global Hollywood 2 (London: British Film Institute, 2008).
31 Howard Rodman, interview with Banks, February 15, 2011. None of this is really new. In the late 1940s, American writers saw this internationalism as both a boon for business and an encroaching threat. In 1948, Robert Pirosh, who worked with René Clair on a Maurice Chevalier film, referred to American writers who had contracts as “part of a postwar invasion of Europe through international coproductions.” Yet in 1947, an editorial in the journal The Screen Writer portended that “in the years to come, it is not inconceivable that the film industries in India and China may further encroach upon areas which we once held almost exclusively.” Editorial, The Screen Writer, July 1945, 38.
32 Studies of writers around the world include Eva Novrup Redvall on television screen authorship in Denmark and Bridget Conor on British and New Zealand screenwriters. See Redvall, Writing and Producing Television Drama in Denmark: From “The Kingdom” to “The Killing” (New York: Palgrave Macmillan, 2013); Bridget Conor, “Subjects at Work: Investigating the Creative Labour of British Screenwriters,” in Behind the Screen: Inside European Production Cultures, ed. Petr Szczepanik and Patrick Vonderau (London: Palgrave Macmillan, 2013); and her “Problems in ‘Wellywood’: Rethinking the Politics of Transnational Cultural Labour,” Flow TV 13, last modified January 28, 2011, http://flowtv.org/2011/01/problems-in-wellywood/.
33 Norelle Scott, interview with Banks, March 2014.
34 Agustín Díaz Yanes in Hablan los guionistas, dir. Alfonso S. Suárez, Sindicato de Guionistas ALMA, 2013.
35 Conor, “Problems in ‘Wellywood.’” | textbooks/socialsci/Sociology/Precarious_Creativity%3A_Global_Media_Local_Labor_(Curtin_and_Sanson)/20%3A_Internationalizing_Labor_Activism__Building_Solidarity_among_Writers_Guilds/20.05%3A_Conclusion.txt |
Objectives
Upon completion of this lesson, you should be able to:
• Become familiar with the standard ANOVA basics.
• Apply the Exploratory Data Analysis (EDA) basics for ANOVA appropriate data.
In previous statistics courses analysis of variance (ANOVA) has been applied in very simple settings, mainly involving one group or factor as the explanatory variable. In this course, ANOVA models are extended to more complex situations involving several explanatory variables. The experimental design aspects are discussed as well. Even though the ANOVA methodology developed in the course is for data obtained from designed experimental settings, the same methods may be used to analyze data from observational studies as well. However, let us keep in mind that the conclusions made may not be as sound because observational studies do not satisfy the rigorous conditions that the designed experiments are subjected to.
Note!
If you aren't familiar with the difference between observational and experimental studies, you should be reviewing introductory statistical concepts which are essential for success in this course!
"Classic" analysis of variance (ANOVA) is a method to compare average (mean) responses to experimental manipulations in controlled environments. For example, if people who want to lose weight are randomly selected to participate in a weight-loss study, each person might be randomly assigned to a dieting group, an exercise group, and a "control" group (for which there is no intervention). The mean weight loss for each group is compared to every other group.
Recall that a fundamental tenet of the scientific method is that results should be reproducible. A designed experiment provides this through replication and generates data that requires the calculation of mean (average) responses.
01: Overview of ANOVA
Using the scientific method, before any statistical analysis can be conducted, a researcher must generate a guess, or hypothesis about what is going on. The process begins with a Working Hypothesis. This is a direct statement of the research idea. For example, a plant biologist may think that plant height may be affected by applying different fertilizers. So they might say: "Plants with different fertilizers will grow to different heights".
But according to the Popperian Principle of Falsification, we can't conclusively affirm a hypothesis, but we can conclusively negate a hypothesis. So we need to translate the working hypothesis into a framework wherein we state a null hypothesis that the average height (or mean height) for plants with the different fertilizers will all be the same. The alternative hypothesis (which the biologist hopes to show) is that they are not all equal, but rather some of the fertilizer treatments have produced plants with different mean heights. The strength of the data will determine whether the null hypothesis can be rejected with a specified level of confidence.
Pictured in the graph below, we can imagine testing three kinds of fertilizer and also one group of plants that are untreated (the control). The plant biologist kept all the plants under controlled conditions in the greenhouse, to focus on the effect of the fertilizer, the only thing we know to differ among the plants. At the end of the experiment, the biologist measured the height of each plant. Plant height is the dependent or response variable and is plotted on the vertical (\(y\)) axis. The biologist used a simple boxplot to plot the difference in the heights.
This boxplot is a customary way to show treatment (or factor) level differences. In this case, there was only one treatment: fertilizer. The fertilizer treatment had four levels that included the control, which received no fertilizer. Using this language convention is important because later on we will be using ANOVA to handle multi-factor studies (for example if the biologist manipulated the amount of water AND the type of fertilizer) and we will need to be able to refer to different treatments, each with their own set of levels.
Another alternative for viewing the differences in the heights is with a means plot (a scatter or interval plot):
This second method to plot the difference in the means of the treatments provides essentially the same information. However, this plot illustrates the variability in the data with 'error bars' that are the 95% confidence interval limits around the means.
In between the statement of a Working Hypothesis and the creation of the 95% confidence intervals used to create this means plot is a 7-step process of statistical hypothesis testing, presented in the following section. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/01%3A_Overview_of_ANOVA/1.01%3A_The_Working_Hypothesis.txt |
We will cover the seven steps one by one.
Step 1: State the Null Hypothesis
The null hypothesis can be thought of as the opposite of the "guess" the researchers made: in this example, the biologist thinks the plant height will be different for the fertilizers. So the null would be that there will be no difference among the groups of plants. Specifically, in more statistical language the null for an ANOVA is that the means are the same. We state the null hypothesis as: $H_{0}: \ \mu_{1} = \mu_{2} = \ldots = \mu_{T}$ for $T$ levels of an experimental treatment.
Note
Why do we do this? Why not simply test the working hypothesis directly? The answer lies in the Popperian Principle of Falsification. Karl Popper (a philosopher) discovered that we can't conclusively confirm a hypothesis, but we can conclusively negate one. So we set up a null hypothesis which is effectively the opposite of the working hypothesis. The hope is that based on the strength of the data, we will be able to negate or reject the null hypothesis and accept an alternative hypothesis. In other words, we usually see the working hypothesis in $H_{A}$.
Step 2: State the Alternative Hypothesis
$H_{A}: \ \text{treatment level means not all equal}$
The reason we state the alternative hypothesis this way is that if the null is rejected, there are many possibilities.
For example, $\mu_{1} \neq \mu_{2} = \ldots = \mu_{T}$ is one possibility, as is $\mu_{1} = \mu_{2} \neq \mu_{3} = \ldots = \mu_{T}$. Many people make the mistake of stating the alternative hypothesis as $mu_{1} \neq mu_{2} \neq \ldots \neq \mu_{T}$, which says that every mean differs from every other mean. This is a possibility, but only one of many possibilities. To cover all alternative outcomes, we resort to a verbal statement of "not all equal" and then follow up with mean comparisons to find out where differences among means exist. In our example, this means that fertilizer 1 may result in plants that are really tall, but fertilizers 2, 3, and the plants with no fertilizers don't differ from one another. A simpler way of thinking about this is that at least one mean is different from all others.
Step 3: Set $\alpha$
If we look at what can happen in a hypothesis test, we can construct the following contingency table:
Decision
In Reality
$H_{0}$ is TRUE $H_{0}$ is FALSE
Accept $H_{0}$ correct Type II Error
$\beta$ = probability of Type II Error
Reject $H_{0}$
Type I Error
$\alpha$ = probability of Type I Error
correct
You should be familiar with type I and type II errors from your introductory course. It is important to note that we want to set $\alpha$ before the experiment (a priori) because the Type I error is the more grievous error to make. The typical value of $\alpha$ is 0.05, establishing a 95% confidence level. For this course, we will assume $\alpha$=0.05, unless stated otherwise.
Step 4: Collect Data
Remember the importance of recognizing whether data is collected through an experimental design or observational study.
Step 5: Calculate a test statistic
For categorical treatment level means, we use an $F$ statistic, named after R.A. Fisher. We will explore the mechanics of computing the $F$ statistic beginning in Chapter 2. The $F$ value we get from the data is labeled $F_{\text{calculated}}$.
Step 6: Construct Acceptance / Rejection regions
As with all other test statistics, a threshold (critical) value of $F$ is established. This $F$ value can be obtained from statistical tables or software and is referred to as $F_{\text{critical}}$ or $F_{\alpha}$. As a reminder, this critical value is the minimum value for the test statistic (in this case the F test) for us to be able to reject the null.
The $F$ distribution, $F_{\alpha}$, and the location of acceptance and rejection regions are shown in the graph below:
Step 7: Based on steps 5 and 6, draw a conclusion about H0
If the $F_{\text{\calculated}}$ from the data is larger than the $F_{\alpha}$, then you are in the rejection region and you can reject the null hypothesis with $(1 - \alpha)$ level of confidence.
Note that modern statistical software condenses steps 6 and 7 by providing a $p$-value. The $p$-value here is the probability of getting an $F_{\text{calculated}}$ even greater than what you observe assuming the null hypothesis is true. If by chance, the $F_{\text{calculated}} = F_{\alpha}$, then the $p$-value would exactly equal $\alpha$. With larger $F_{\text{calculated}}$ values, we move further into the rejection region and the $p$-value becomes less than $\alpha$. So the decision rule is as follows:
If the $p$-value obtained from the ANOVA is less than $\alpha$, then reject $H_{0}$ and accept $H_{A}$.
Note
If you are not familiar with this material, we suggest that you review course materials from your basic statistics course.
1.03: Chapter 1 Summary
The emphasis of this lesson was to reinforce the basics of ANOVA, which perhaps you may have seen in other courses. Using the greenhouse example, the seven important steps of hypothesis testing in a single factor ANOVA setting were explored. Step 2 highlighted the correct way to state and also interpret the alternative hypothesis $\left(H_{A}\right)$, while Step 3 discusses the Truth Table that includes possible errors in hypothesis testing. Step 6 discusses in detail the rejection region of the null hypothesis $\left(H_{0}\right)$.
The lesson also introduced us to some basics in ANOVA-related explanatory data analysis (EDA). The graphics such as side-by-side boxplots and mean plots are useful tools in producing a visual summary of the raw data and ANOVA results. These will serve as stepping stones to more elaborate graphical techniques we will learn throughout the course.
The concepts and methodology learned in this lesson, though seem straight forward will help us navigate more complex topics addressed in future lessons. The keywords and phrases learned in this lesson are:
• null and alternative hypotheses ($H_{0}$ and $H_{A}$)
• Type 1 and Type II errors
• significance level $(\alpha)$
• rejection region
• $F$ statistic and its critical and calculated values. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/01%3A_Overview_of_ANOVA/1.02%3A_The_7-Step_Process_of_Statistical_Hypothesis_Testing.txt |
Objectives
Upon completion of this chapter, you should be able to:
• Perform basic computations for Single Factor ANOVA and interpret the results.
• Carry out the Tukey pairwise mean comparison method.
• Learn about other pairwise mean comparison methods.
• Conduct a contrast analysis that accommodates the comparison of group means.
In this chapter, we will begin to learn the notation and the formulas to compute the fundamental quantities necessary for ANOVA-related hypothesis testing as well as mean comparison procedures. The application of these statistical procedures will be illustrated using the Greenhouse example from Chapter 1.
02: ANOVA Foundations
The idea of ANOVA is to compare different sources of variability: between sample variability and within sample variability.
As a point of review, the alternative hypothesis is what we think is going on (or what we need to conclude). Typically we are looking to find differences among at least one pair of our treatment means. Because of this, the null hypothesis (the opposite of the alternative) states that there are no differences among the group means (or that they are all equal).
To test the Null hypothesis (which is traditionally written as $H_{0}: \ \mu_{1} = \mu_{2} = \ldots = \mu_{T}$, we need to compute the test $(F)$ statistic that compares the between sample variability to within sample variability.
To see how we compute this statistic it is helpful to look at the ANOVA table. The table below is an ANOVA table (here presented blank, with no entries yet):
To define the elements of the table and fill in these quantities, let’s return to our example data (Lesson 1 Data) for the hypothetical greenhouse experiment:
Control F1 F2 F3
21 32 22.5 28
19.5 30.5 26 27.5
22.5 25 28 31
21.5 27.5 27 29.5
20.5 28 26.5 30
21 28.6 25.2 29.2
Notation
Each observation in the dataset can be referenced by two indicator subscripts, $i$ and $j$, as $Y_{ij}$.
For those of you not familiar with this notation, we use $Y$ to indicate that it is a response variable. The subscript $i$ refers to the $i^{th}$ level of the treatment; our example has 4 treatments, so $i$ will take on the values $1, 2, 3,$ and $4$.) The subscript $j$ refers to the $j^{th}$ observation (again, our example has 6 observations for each treatment so $j$ takes the values $1,2,3,4,5,$ and $6$). It is important to note that the $j^{th}$ observation is occurring within the $i^{th}$ treatment level.
subscripts $i=1$ $i=2$ $i=3$ $i=4$
Control F1 F2 F3
$j=1$ 21 32 22.5 28
$j=2$ 19.5 30.5 26 27.5
$j=3$ 22.5 25 28 31
$j=4$ 21.5 27.5 27 29.5
$j=5$ 20.5 28 26.5 30
$j=6$ 21 28.6 25.2 29.2
For example, $Y_{4,2}=27.5$.
We now can define the various means explicitly using these subscripts. The overall or Grand Mean is given by $\text{Grand Mean} = \bar{Y}_{..}$where the dots indicate that the quantity has been averaged over that subscript. For the Grand Mean, we have averaged over all $j$ observations in all $i$ treatment levels. The treatment means are given by $\text{Treatment Mean} = \bar{Y}_{i}$ indicating that we have averaged over the $j$ observations in each of the $i$ treatment levels.
We can find these in the output from the summary procedure that can be generated in SAS and the coding details are discussed in Chapter 3:
Summary Output for Lesson 1 Data
Fert _Type_ _FREQ_ mean
0 24 26.1667
Control 1 6 21.0000
F1 1 6 28.6000
F2 1 6 25.8667
F3 1 6 29.2000
In the output we see the column heading _TYPE_. The summary procedure in SAS calculates all possible means when specified, and so the _TYPE_ indicates what mean is being computed. _TYPE_ = 0 is the Grand Mean, and we can see this from the number of observations (given by _FREQ_) of 24. Each of the treatment level means is listed as _TYPE_ = 1 and we confirm that 6 replications were made for each treatment level (remember that j took on values 1 through 6).
Note that SAS automatically has ordered the treatment levels alphabetically.
The grand mean and treatment means are all we need in this example to compute the quantities for the ANOVA table. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.01%3A_Building_the_ANOVA_Table_-_Notation.txt |
When working with ANOVA, we start with the total variability in the response variable and divide or "partition" it into different parts: the between sample variability (i.e. variability due to our treatment) and the within sample variability (i.e. residual variability). The variability that is due to our treatment we of course hope is significantly large and variability in the response that is leftover can be thought of as the nuisance, "error", or "residual" variability.
To help you imagine this a bit more, think about the data storage capacity of a computer. If you have 8GB of storage total, you can ask your computer to show the types of files that are occupying the storage. The ANOVA model is (in a very elementary fashion) going to compare the variability due to the treatment to the variability left over.
From elementary statistics, when we think of computing a variance of a random variable (say $X$), we use the expression: $\text{variance} = \frac{\sum \left(X_{i} - \bar{X}\right)^{2}}{N-1} = \frac{SS}{df}$
The numerator of this expression is referred to as the Sum of Squares, or Sum of Squared deviations from the mean, or simply SS. (If you don't recognize this, then we suggest you sharpen your introductory statistics skills!) The denominator is the degrees of freedom, $(N-1)$, or $df$.
ANOVA Table Rules
1. Total SS = sum of the SS of all Sources (i.e., Total SS = Treatment SS + Error SS)
2. Total df = sum of df of all Sources
3. MS = SS/df
4. $F_{\text{calculated}} = \dfrac{\text{Treatment MS}}{\text{Error MS}}$ with numerator df = number of treatments - 1 and denominator df = error df
The ANOVA table is set up to generate quantities analogous to the simple variance calculation above. In our greenhouse experiment example:
1. We start by considering the TOTAL variability in the response variable. This is done by calculating the $SS_{\text{Total}}$ $\begin{array}{l} \text{Total SS} &= \sum_{i} \sum_{j} \left(Y_{ij} - \bar{Y}_{..}\right)^{2} \ &= \mathbf{312.47} \end{array}$ The degrees of freedom for the Total SS is $N - 1 = 24 - 1 = 23$, where $N$ is the total sample size.
2. Our next step determines how much of the variability in $Y$ is accounted for by our treatment. We now calculate $SS_{\text{Treatment}}$ or $SS_{\text{Trt}}$: $\text{Treatment SS} = \sum_{i} n_{i} \left(\bar{Y}_{i.} - \bar{Y}_{..}\right)^{2}$
Note!
The sum of squares for the treatment is the deviation of the group mean from the grand mean. So in some sense, we are "aggregating" all of the responses from that group and representing the "group effect" as the group mean.
and for our example: $\begin{array}{c} \text{Treatment SS} = 6 \times (21.0-26.1667)^{2} + 6 \times (28.6-26.1667)^{2} + \ \ldots + 6 \times (25.8667-26.1667)^{2} + 6 \times (29.2-26.1667)^{2} = 251.44 \end{array}$ Note that in this case we have equal numbers of observations (6) per treatment level, and it is, therefore, a balanced ANOVA.
3. Finally, we need to determine how much variability is "left over". This is the Error or Residual sums of squares by subtraction: $\begin{array}{l} \text{Error SS} &= \sum_{i} \sum_{j} \left(Y_{ij} - \bar{Y}_{i.}\right)^{2} = \text{Total SS} - \text{Treatment SS} \ &= 312.47 - 251.44 = \mathbf{61.033} \end{array}$ Note here that the "leftover" is really the deviation of any score from its group mean.
We can now fill in the following columns of the table:
ANOVA
Source df SS MS F
Treatment T - 1 = 3 251.44
Error 23-3=20 61.033
Total N - 1 =23 312.47
We have $T$ treatment levels and so we use $T-1$ for the $df$ for the treatment. In our example, there are 4 treatment levels (the control and the 3 fertilizers) so $T=4$ and $T-1 = 4-1 = 3$. Finally, we obtain the error $df$ by subtraction as we did with the SS.
The Mean Squares (MS) can now be calculated as: $MS_{Trt} = \frac{SS_{Trt}}{df_{Trt}} = \frac{251.44}{3} = 83.813$ and $MS_{Error} = \frac{SS_{Error}}{df_{Error}} = \frac{61.033}{20} = 3.052$
NOTE: $MS_{\text{Error}}$ will sometimes be referred as $MSE$ and we don’t need to calculate the $MS_{\text{Total}}$.
ANOVA
Source df SS MS F
Treatment 3 251.44 83.813
Error 20 61.033 3.052
Total 23 312.47
Finally, we can compute the $F$ statistic for our ANOVA. Conceptually we are comparing the ratio of the variability due to our treatment (remember we expect this to be relatively large) to the variability leftover, or due to error (and of course, since this is an error we want this to be small). Following this logic, we expect our $F$ to be a large number. If we go back and think about the computer storage space we can picture most of the storage space taken up by our treatment, and less of it taken up by error. In our example, the $F$ is calculated as: $F = \frac{MS_{Trt}}{MS_{Error}} = \frac{83.813}{3.052} = 27.46$
Source df SS MS F
Treatment 3 251.44 83.813 27.46
Error 20 61.033 3.052
Total 23 312.47
So how do we know if the $F$ is large enough to conclude we have a significant amount of variability due to our treatment? We look up the critical value of $F$ and compare it to the value we calculated. Specifically, the critical $F$ is $F_{\alpha} = F_{(0.05,3,20)} = 3.10$. The critical value can be found using tables or technology.
Finding a Critical Value of $F$
Using a Table:
Appendix Table B4
Using SAS:
data Fvalue;
q=finv(0.95, 3, 20);
put q=;
run;
proc print data=work.Fvalue;
run;
The Print Procedure
Data Set WORK.FVALUE
Obs q
1 3.09839
Most F tables actually index this value as $1 - \alpha = .95$
The $F_{calculated} > F_{\alpha}$ so we reject $H_0$ and accept the alternative $H_{A}$. The $p$-value (which we don't typically calculate by hand) is the area under the curve to the right of the $F_{calculated}$ and is the way the process is reported in statistical software. Note that in the unlikely event that the $F_{calculated}$ is exactly equal to the $F_{\alpha}$ then the $p \text{-value} = \alpha$. As the calculated $F$ statistic increases beyond the $F_{\alpha}$ and we go further into the rejection region, the area under the curve (hence the $p$-value) gets smaller and smaller. This leads us to the decisions rule: If the $p$-value is $< \alpha$ then we reject $H_{0}$. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.02%3A_Computing_Quantities_for_the_ANOVA_Table.txt |
If (and only if) we reject the null hypothesis, we then conclude at least one group is different from one other (importantly, we do not conclude that all the groups are different).
If it is the case that we reject the null, then we will want to know which group or groups are different. In our example we are not satisfied knowing at least one treatment level is different, we want to know where the difference is and the nature of the difference. To answer this question, we can follow up the ANOVA with a mean comparison procedure to find out which means differ from each other and which ones don’t.
You might think we could not bother with the ANOVA and proceed with a series of t-tests to compare the groups. While that is intuitively simple, it creates inflation of the type I error. How does this inflation of type I error happen? For a single test, $\alpha = 1 - (.95)$
The probability of committing a type I error (by random chance) for two simultaneous tests follows from the Multiplication Rule for independent events in probability. Recall that, for two independent events $\text{A}$ and $\text{B}$ the probability of $\text{A}$ and $\text{B}$ both occurring is $P(\text{A and B}) = P(\text{A}) * P(\text{B})$. So for two tests, we have $\alpha = 1 - ((.95) * (.95)) = 0.0975$ which is now larger than the α that we originally set. For our example, we have 6 comparisons, so $\alpha = 1 - (.95^{6}) = 0.2649$ which is a much larger (inflated) probability of committing a type I error than we originally set.
The multiple comparison procedures compensate for the type I error inflation (although each does so in a slightly different way).
There are several comparison procedures that can be employed, but we will start with the one most commonly used, the Tukey procedure. In the Tukey procedure, we compute a "yardstick" value based on the $MS_{Error}$ and the number of means being compared. If any two means differ by more than the Tukey $w$ value, then they are significantly different.
Step 1: Compute Tukey's $w$ value
$w = q_{\alpha (p, df_{Error})} \cdot s_{\bar{Y}}$ \begin{aligned} \text{where } & q_{\alpha} \text{ is obtained from a table of Tukey } q \text{ values} \ & p = \text{the number of treatment levels} \ & s_{\bar{Y}} = \text{standard error of a treatment mean} = \sqrt{MS_{Error}/r} \ & r = \text{number of replications} \end{aligned}
Show Tukey $q$ Values Table
df for Error Term $\alpha$ $p$ = Number of Treatments
2 3 4 5 6 7 3 9 10
5 0.05
0.01
3.64
5.70
4.6
6.98
5.22
7.80
5.67
8.42
6.03
8.91
6.33
9.32
6.58
9.67
6.80
9.97
6.99
10.24
6 0.05
0.01
3.46
5.24
4.34
6.33
4.90
7.03
5.30
7.56
5.63
7.97
5.90
8.32
6.12
8.61
6.32
8.87
6.49
9.10
7 0.05
0.01
3.34
4.95
4.16
5.92
4.68
6.54
5.06
7.01
5.36
7.37
5.61
7.68
5.82
7.94
6.00
8.17
6.16
8.37
8 0.05
0.01
3.26
4.75
4.04
5.64
4.53
6.20
4.89
6.62
5.17
6.96
5.40
7.24
5.60
7.47
5.77
7.68
5.92
7.86
9 0.05
0.01
3.20
4.60
3.95
5.43
4.41
5.96
4.76
6.35
5.02
6.66
5.24
6.91
5.43
7.13
5.59
7.33
5.74
7.49
10 0.05
0.01
3.15
4.48
3.88
5.27
4.33
5.77
4.65
6.14
4.91
6.43
5.12
6.67
5.30
6.87
5.46
7.05
5.60
7.21
11 0.05
0.01
3.11
4.39
3.82
5.15
4.26
5.62
4.57
5.97
4.82
6.25
5.03
6.48
5.20
6.67
5.35
6.84
5.49
6.99
12 0.05
0.01
3.08
4.32
3.77
5.05
4.20
5.50
4.51
5.84
4.75
6.10
4.95
6.32
5.12
6.51
5.27
6.67
5.39
6.81
13 0.05
0.01
3.06
4.26
3.73
4.96
4.15
5.40
4.45
5.73
4.69
5.98
4.88
6.19
5.05
6.37
5.19
6.53
5.32
6.67
14 0.05
0.01
3.03
4.21
3.70
4.89
4.11
5.32
4.41
5.63
4.64
5.88
4.83
6.08
4.99
6.26
5.13
6.41
5.25
6.54
15 0.05
0.01
3.01
4.17
3.67
4.84
4.08
5.25
4.37
5.56
4.59
5.80
4.78
5.99
4.94
6.16
5.08
6.31
5.20
6.44
16 0.05
0.01
3.00
4.13
3.65
4.79
4.05
5.19
4.33
5.49
4.56
5.72
4.74
5.92
4.90
6.08
5.03
6.22
5.15
6.35
17 0.05
0.01
2.98
4.10
3.63
4.74
4.02
5.14
4.30
5.43
4.52
5.66
4.70
5.85
4.86
6.01
4.99
6.15
5.11
6.27
18 0.05
0.01
2.97
4.07
3.61
4.70
4.00
5.09
4.28
5.38
4.49
5.60
4.67
5.79
4.82
5.94
4.96
6.08
5.07
6.20
19 0.05
0.01
2.96
4.05
3.59
4.67
3.98
5.05
4.25
5.33
4.47
5.55
4.65
5.73
4.79
5.89
4.92
6.02
5.04
6.14
20 0.05
0.01
2.95
4.02
3.58
4.64
3.96
5.02
4.23
5.29
4.45
5.51
4.62
5.69
4.77
5.84
4.90
5.97
5.01
6.09
24 0.05
0.01
2.92
3.96
3.53
4.55
3.90
4.91
4.17
5.17
4.37
5.37
4.54
5.54
4.68
5.69
4.81
5.81
4.92
5.92
30 0.05
0.01
2.89
3.89
3.49
4.45
3.84
4.80
4.10
5.05
4.30
5.24
4.46
5.40
4.60
5.54
4.72
5.65
4.83
5.76
40 0.05
0.01
2.86
3.82
3.44
4.37
3.79
4.70
4.04
4.93
4.23
5.11
4.39
5.27
4.52
5.39
4.63
5.50
4.74
5.60
For our greenhouse example we get: $w = q_{.05 (4, 20)} \sqrt{(3.052/6)} = 3.96(0.7132) = 2.824$
Step 2: Rank the means, calculate differences
For the greenhouse example, we rank the means as:
29.20 28.6 25.87 21.00
Start with the largest and second-largest means and calculate the difference, $29.20 - 28.60 = 0.60$, which is less than our $w$ of 2.824, so we indicate there is no significant difference between these two means by placing the letter "a" under each:
29.20 28.6 25.87 21.00
a a
Then calculate the difference between the largest and third-largest means, $29.20 - 25.87=3.33$, which exceeds the critical $w$ of 2.824, so we can label these with a "b" to show this difference is significant:
29.20 28.6 25.87 21.00
a a b
Now we have to consider whether or not the second-largest and third-largest differ significantly. This is a step that sets up a back-and-forth process. Here $28.6 - 25.87 = 2.73$, less than the critical $w$ of 2.824, so these two means do not differ significantly. We need to add a factor of "b" to show this:
29.20 28.6 25.87 21.00
a ab b
Continuing down the line, we now calculate the next difference: $28.60 - 21.00=7.60$, exceeding the critical $w$, so we now add a "c":
29.20 28.6 25.87 21.00
a ab b c
Again, we need to go back and check to see if the third-largest also differs from the smallest: $25.87 - 21.00=4.87$, which it does. So we are done.
These letters can be added to figures summarizing the results of the ANOVA.
The Tukey procedure explained above is valid only with equal sample sizes for each treatment level. In the presence of unequal sample sizes, more appropriate is the Tukey-Cramer Method, which calculates the standard deviation for each pairwise comparison separately. This method is available in SAS, R, and most other statistical softwares. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.03%3A_Tukey_Test_for_Pairwise_Mean_Comparisons.txt |
Although the Tukey procedure is the most widely used multiple comparison procedure, there are many other multiple comparison techniques.
An older approach, no longer offered in many statistical computing packages, is Fisher’s Protected Least Significant Difference (LSD). This is a method to compare all possible means, two at a time, as $t$-tests. Unlike an ordinary two-sample $t$-test, however, the method does rely on the experiment-wide error (the MSE). The LSD is calculated as: $LSD(\alpha) = t_{\alpha, df} s_{\bar{d}}$ where $t_{\alpha}$ is based on $\alpha$ and $df =$ error degrees of freedom from the ANOVA table. The standard error for the difference between two treatment means ($s_{\bar{d}}$ or SE) is calculated as: $s_{\bar{d}} = \sqrt{\frac{2s^{2}}{r}}$ where $r$ is the number of observations per treatment mean (replications) and $s^{2}$ is the MSE from the ANOVA. As in the Tukey method, any pair of means that differ by more than the LSD value differ significantly. The major drawback of this method is that it does not control $\alpha$ over for an entire set of pair-wise comparisons (the experiment-wise error rate) and hence is associated with Type 1 inflation.
The following multiple comparison procedures are much more assertive in dealing with Type 1 inflation. In theory, while we can set $\alpha$ for a single test, the fact that we have $T$ treatment levels means there are $T(T-1)/2$ tests (the number of pairs of possible comparisons), and so we need to adjust $\alpha$ to have the desired confidence level for the set of tests. The Tukey, Bonferroni, and Scheffé methods control the experiment-wise error, but in different ways. All three use a $\text{"multiplier"} * SE$ form, but differ in the form of the multiplier.
Contrasts are comparisons involving two or more factor level means (discussed more in the following section). Mean comparisons can be thought of as a subset of possible contrasts among the means. If only pairwise comparisons are made, the Tukey method will produce the narrowest confidence intervals and is the recommended method. The Bonferroni and Scheffé methods are used for general tests of possible contrasts. The Bonferroni method is better when the number of contrasts being tested is about the same as the number of factor levels. The Scheffé method covers all possible contrasts, and as a result, is the most conservative of all the methods. The drawback for such a highly conservative test, however, is that it becomes more difficult to resolve differences among means, even though the ANOVA would indicate that they exist.
When treatment levels include a control and mean comparisons are restricted to only comparing treatment levels against a control level, Dunnett’s mean comparison method is appropriate. Because there are fewer comparisons made in this case, the test provides more power compared to a test (see Section 3.7) using the full set of all pairwise comparisons.
To illustrate these methods, the following output was obtained (as we will see later on in the course) for the hypothetical greenhouse data of our example. We will be running these types of analyses later.
Fisher’s Least Significant Difference (LSD)
Since the estimated means for F1 and F3 are covered by the same colored bar, they are not significantly different using the LSD approach.
Tukey
Since the estimated means for F1 and F3 are covered by the same colored bar (red bar), they are not significantly different using Tukey's approach. Similarly, since F1 and F2 are covered by the same colored bar (blue bar) they are not significantly different using Tukey's approach.
Bonferroni
Observations from the Bonferroni approach are similar to the ones from Tukey's approach.
Scheffé
Observations from the Scheffé approach are similar to the ones from Tukey's and Bonferroni's approaches.
Dunnett
Comparisons significant at the 0.05 level are indicated by ***.
Fertilizer
Comparison
Difference
Between
Means
Simultaneous 95% Confidence Limits ***
F3 - Control 8.200 5.638 10.762 ***
F1 - Control 7.600 5.038 10.162 ***
F2 - Control 4.867 2.305 7.429 ***
We can see that the LSD method was the most liberal, that is, it indicated the largest number of significant differences between means. In this example, Tukey, Bonferroni, and Scheffé produced the same results. The Dunnett test was consistent with the other 4 methods, and this is not surprising given the small value of the control mean compared to the other treatment levels.
To get a closer look at the results of employing the different methods, we can focus on the differences between the means for each possible pair:
Comparison Difference between means
Control F1 7.6000
Control F2 4.8667
Control F3 8.2000
F1 F2 2.7333
F1 F3 0.6000
F2 F3 3.3333
and compare the 95% confidence intervals produced:
Type LSD Tukey Bonferroni Scheffé Dunnett
Comparison Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper
Control F1 5.496 9.704 4.777 10.423 4.648 10.552 4.525 10.675 5.038 10.162
Control F2 2.763 6.971 2.044 7.690 1.914 7.819 1.792 7.942 2.305 7.429
Control F3 6.096 10.304 5.377 11.023 5.248 11.152 5.125 11.275 5.638 10.762
F1 F2 0.629 4.837 -0.090 5.556 -0.2189 5.686 -0.342 5.808 X X
F1 F3 -1.504 2.704 -2.223 3.423 -2.352 3.552 -2.475 3.675 X X
F2 F3 1.229 5.437 0.510 6.156 0.3811 6.286 0.258 6.408 X X
You can see that the LSD produced the narrowest confidence intervals for the differences between means. Dunnett’s test had the next most narrow intervals, but only compares treatment levels to the control. The Tukey method produced intervals that were similar to those obtained for the LSD, and the Scheffé method produced the broadest confidence intervals.
What does this mean? When we need to be REALLY sure about our results, we should use conservative tests. If you are working in life-and-death situations, such as in most clinical trials or bridge building, you might want to be surer. If the consequences are less severe you can use a more liberal test, understanding there is more of a chance you might be incorrect (but still able to detect differences). In reality, you need to be consistent with the rigor used in your discipline. While we can't tell you which comparison to use, we can tell you the differences among the tests and the trade-offs for each one. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.04%3A_Other_Pairwise_Mean_Comparison_Methods.txt |
The paired comparisons discussed in sections 2.2 and 2.3 have the limitation that the comparisons are made only between treatment mean pairs. The contrast analysis procedure can be used to carry out comparisons of a much wider context such as comparisons of treatment level groups or even testing of trends prompting regression modeling to express the response vs. treatment relationship with treatment as a numerical predictor. In the context of a single factor ANOVA model, a linear contrast can be defined as a linear combination of the treatment means such that their numerical coefficients add to zero. Mathematically, a contrast can be represented by... $A = \sum_{i=1}^{T} a_{i} \bar{y}_{i}$ where $\bar{y}_{1}, \bar{y}_{2}, \ldots, \bar{y}_{T}$ represent the sample treatment means and $\sum_{i=1}^{T} a_{i} = 0$. The quantity $A$ is a sample statistic and serves as an estimate for the population contrast $\sum_{i=1}^{T} a_{i} \mu_{i}$. By choosing the numerical coefficients appropriately, linear contrasts can be used to make different comparisons among groups of treatment means but not limited to only mean pairs. The table below gives 4 linear contrasts defined in terms of the 3 fertilizer levels F1, F2, F3, and the Control in the greenhouse example.
Table: Greenhouse example contrasts
Ex $a_1$ $a_2$ $a_3$ $a_4$ Contrast
1 1 -1 0 0 F1-F2
2 1 1 1 -3 F1+F2+F3-3C
3 1 1 -2 0 F1+F2-2F3
4 0 1 -1 0 F2-F3
Notice that values of each list of $a_{i}$ ($i = 1,2,3,4$) add to zero. The first contrast compares the first two fertilizer types in terms of their means, and the second compares the means of the 3 fertilizer types with the Control mean. The third is a comparison between the combined effect of fertilizer types 1 and 2 with fertilizer type 3, while the last contrast compares the second and third fertilizer types.
A pair of contrasts $A = \sum_{i=1}^{T} a_{i} \bar{y}_{i}$ and $B = \sum_{i=1}^{T} b_{i} \bar{y}_{i}$ is orthogonal if the products of their numerical coefficients add to zero. This can be expressed mathematically as $\sum_{i=1}^{T} a_{i} b_{i} = 0$
A set of contrasts is said to be orthogonal if every pair of contrasts in the set is orthogonal. Two orthogonal contrasts are not correlated which means that if $A$ and $B$ are orthogonal, then $\text{Covariance} (A, B) = 0$. Furthermore, the sum of squares of the treatment usually displayed in the ANOVA table can be partitioned into a set of ($T-1)$ orthogonal contrasts each with 1 degree of freedom. Note that the maximal number of orthogonal contrasts associated with a treatment of $T$ levels is $(T-1)$ and each of them would be associated with one specific comparison independent of each other. In the table above, contrasts 1, 2, and 3 form an orthogonal set of contrasts and contrast 4 cannot be admitted into this set.
The statistical significance of a linear contrast, which can be equated to testing for the zero contrast value can be formulated using the null and alternative hypotheses: $H_{0}: \ \sum_{i=1}^{T} a_{i} \mu_{i} = 0 \text{ vs. } H_{A}: \ \sum_{i=1}^{T} a_{i} \mu_{i} \neq 0$ and can be tested using either,
$t = \dfrac{\sum_{i=1}^{T} a_{i} \bar{y}_{i}}{\sqrt{\text{MSE} \sum_{i=1}^{T} \frac{a_{i}^{2}}{n_{i}}}} \text{ with } (N-T) \text{ degrees of freedom}$ $\text{or} \nonumber$ $F = \frac{\left(\sum_{i=1}^{T} a_{i} \bar{y}_{i} \right)^{2}}{\text{MSE} \sum_{i=1}^{T} \frac{a_{i}^{2}}{n_{i}}}$
with numerator and denominator degrees of freedom equal to $1$ and $(N-T)$ respectively.
Note that MSE can be obtained from the ANOVA table. Applying the above formula, the $t$-statistic for testing contrast 2 above is... $t = \frac{\sum_{i=1}^{T} a_{i} \bar{y}_{i}}{\sqrt{\text{MSE} \sum_{i=1}^{T} \frac{a_{i}^{2}}{n_{i}}}} = \frac{28.6 + 25.867 + 29.2 + (3 * 21)}{3.052 \times \sqrt{\frac{(1+1+1+9)}{6}}} = 8.365$
with $df=20$ and has a $p$-value of .0028, indicating that the average plant height due to the combined treatment of the 3 fertilizer types differs significantly from the average plant height yielded by the control.
The above testing procedure is applicable to non-orthogonal contrasts as well. But, as non-orthogonal contrasts are not guaranteed to be uncorrelated, the conclusions arrived at may be "overlapping" and lead to redundancies. In Chapter 3, examples are provided to illustrate how software can be used to conduct contrast testing. The hypothesis testing for trends using contrasts will be discussed in Chapter 10: ANCOVA II. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.05%3A_Contrast_Analysis.txt |
Exercise $1$: Teaching Effectiveness
To compare the teaching effectiveness of 3 teaching methods, the semester average based on 4 midterm exams from five randomly selected students enrolled in each teaching method were used.
1. What is the response in this study?
2. How many replicates are there?
3. Write the appropriate null and alternative hypotheses.
4. Complete the partially filled ANOVA table given below. Round your answers to 4 decimal places.
Source df SS MS F p-value
teach_mtd 245
error
total 345.1
5. Find the critical value at $\alpha = .01$
6. Make your conclusion.
7. From the ANOVA analysis, you performed, can you detect the teaching method which yields the highest semester average? If not, suggest a technique that will.
Solution
1. Average of 4 mid-terms
2. 5
3. $H_{0}: \ \mu_{1} = \mu_{2} = \mu_{3} = \mu_{4}$, where $\mu_{1}, \mu_{2}, \mu_{3}$ are the actual semester average of a student enrolled in teaching method 1, method 2, and method 3 respectively. Ha: Not all semester averages are equal. (This means that there are at least two teaching methods that differ in their actual semester averages)
4. Source df SS MS F p-value
teach_mtd 2 245 122.5000 14.6853 0.0006
error 12 100.1 8.3417
total 14 345.1
5. 6.925
6. As the calculated $F$-statistic value = 14.6853 is more than the critical value of 6.925, $H_{0}$ should be rejected. Therefore, we can conclude that all 3 teaching methods do not have the same semester average, indicating that at least 2 teaching methods differ in their actual semester average.
7. The ANOVA conclusion indicated that not all 3 teaching methods are equally effective, but did not indicate which one yields the highest mean score. The Tukey comparison method is one procedure that shows the teaching method that yields the significantly highest average semester score.
Exercise $2$: Commuter Times
In a local commuter bus service, the number of daily passengers for 50 weeks was recorded. The purpose was to determine if the passenger volume is significantly less during weekends compared to workdays. Below are summary statistics for each day of the week. The partially filled ANOVA table, along with a Tukey plot, is shown below.
Statistics
Day N Mean SE Mean Std Dev
Sun 50 486.500 9.003 63.661
Mon 50 514.600 6.891 48.724
Tue 50 501.340 7.922 56.018
Wed 50 520.640 7.055 49.886
Thu 50 512.880 10.258 72.532
Fri 50 512.600 8.086 57.174
Sat 50 469.860 8.988 63.555
a) State the appropriate null and alternative hypotheses for this test.
Solution
$H_{0}: \ \mu_{Sun} = \mu_{Mon} = \mu_{Tues} = \mu_{Wed} = \mu_{Thurs} = \mu_{Fri} = \mu_{Sat}$
$H_{A}: \ \text{At least one } \mu_{day \ i} \neq \mu_{day \ j}, \text{ for some$ i, j = 1, 2, \ldots, 7 \text{ OR not all means are equal}\)
b) Complete the partially filled ANOVA table given below. Use two decimal places in the $F$ statistic.
Source df SS MS F p-value
Groups 100391
Error
Total 1306887
Solution
Source df SS MS F p-value
Day 6 100391 16731.8 4.76 0.0001
Error 343 1206496 3517.5
Total 349 1306887
c) Use the appropriate $F$-distribution cumulative probabilities to verify that the $p$-value for the test is approximately zero.
Solution
$p$-value $\approx 0$ (from the $F$-distribution with 6 and 343 degrees of freedom)
d) Use $\alpha=0.05$ to test if the mean passenger volume differs significantly by day of the week.
Solution
Since the $p$-value $\leq \alpha = 0.05$, we reject $H_{0}$. There is strong evidence to indicate that the mean passenger volume differs significantly by day of the week (i.e., for some days of the week, the average number of commuters is more than others, but this test does not indicate which days have a higher passenger volume).
e) Use the output to make a statement about how the mean daily passenger volume differs significantly by day of the week.
Solution
The passenger volume on Sundays is not statistically different from Saturdays and also from Tuesdays. The mean passenger volume on Saturdays is significantly lower than on workdays other than Tuesdays.
f) The management would like to know if the overall number of commuters is significantly more during workdays than during weekends. An appropriate comparison to respond to their query would be to compare the average number of commuters between workdays (Monday through Friday) and the weekend. Write the weight (coefficients) for a linear contrast to make this comparison. Test the hypothesis that the average commuter volume during the weekends is less.
Solution
The weights (coefficients) for the appropriate contrast are given below.
Day Mon Tue Wed Thu Fri Sat Sun
weight 1 1 1 1 1 -2.5 -2.5
$t = \dfrac{\sum_{i=1}^{T} a_{i} \bar{y}_{i}}{\sqrt{MSE \sum_{i=1}^{T} \frac{a_{i}^{2}}{n_{i}}}} = \dfrac{171.16}{\sqrt{3517.5 * \frac{17.5}{50}}} = 4.878$
Under the null hypothesis, this test statistic has a $t$-distribution with 343 degrees of freedom. You can obtain the $p$-value using statistical software. Recall this is a one-tailed test.
Student's t distribution with 343 DF
$x$ $P(X \leq x)$
$4.878$ $8.216815 \cdot 10^{-7} \approx 0$
This $p$-value indicates that the difference in the average number of passengers is statistically significant between workdays and weekends.
See the table below for computations:
Factor N Mean weights product weight2
Mon 50 514.6 1.0 514.6 1.00
Tue 50 501.34 1.0 501.34 1.00
Wed 50 520.64 1.0 520.64 1.00
Thu 50 512.88 1.0 512.88 1.00
Fri 50 512.6 1.0 512.6 1.00
Sat 50 469.86 -2.5 -1174.65 6.25
Sun 50 486.5 -2.5 -1216.25 6.25
Recall that the MSE (error mean squares) is 3517.5 with $df_{error}=343$.
2.07: Chapter 2 Summary
In this lesson, we became familiar with the ANOVA methodology to test for equality among treatment means. As follow-up procedures, we were also exposed to the Tukey method for paired mean comparisons which helped to identify significantly different treatment (factor) levels. The contrast analysis was also discussed as a means to compare differences among group means. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/02%3A_ANOVA_Foundations/2.06%3A_Try_It.txt |
In Chapter 2 we learned that ANOVA is based on testing the effect of the treatment relative to the amount of random error. In statistics, we call this the partitioning of variability (due to treatment and due to random variability in the measurements). This partitioning of the deviations can be written mathematically as: $\underbrace{Y_{ij} - \bar{Y}_{..}}_{(1)} = \underbrace{\bar{Y}_{i.} - \bar{Y}_{..}}_{(2)} + \underbrace{Y_{ij} - \bar{Y}_{i}}_{(3)}$ Thus, the total deviation $Y_{ij} - \bar{Y}_{..}$ in $(1)$ can be viewed as the sum of two components:
$(2)$ Deviation of estimated factor level mean around overall mean, and
$(3)$ Deviation of the $j^{th}$ response of the $i^{th}$ factor around the estimated factor level mean.
These two deviations are also called variability between groups, a reflection of differences between treatment levels and the variability within groups that serves as a proxy for the error variability among individual observations. A practitioner would however be more interested in the variability between groups as it is the indicator of treatment level differences and may have little interest in the within-group variability, expecting it to be in fact small. However, it will be seen that both these variability measures will play an important role in statistical procedures.
There are several mathematically equivalent forms of ANOVA models describing the relationship between the response and the treatment. In this chapter we will focus on the effects model, and in the next chapter three other alternative models will be introduced.
This lesson will also cover the topic of model assumptions needed to employ the ANOVA. Model diagnostics, which deal with verifying the validity of model assumptions, are also discussed, along with power analysis techniques to assess the power associated with a statistical study. How software can be used to analyze data using the statistical techniques discussed will also be presented.
03: ANOVA Models Part I
The effects model for one way ANOVA is a linear additive statistical model which relates the response to the treatment and can be expressed as $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$
where $\mu$ is the grand mean, $\tau_{i} \ (i = 1,2, \ldots,T)$ are the deviations from the grand mean due to the treatment levels and $\epsilon_{ij}$ are the error terms. The quantities $\tau_{i} \ (i = 1,2, \ldots, T)$ which add to zero, are also referred to as the treatment level effects and the errors show the amount "left over" after considering the grand mean and the effect of being in a particular treatment level.
Here’s the analogy in terms of the greenhouse experiment. Think of someone who is not aware that different fertilizers have been used walking into the greenhouse to simply inquire about plant heights in general. The overall sample mean, an estimate of the grand mean, will be a suitable response to this inquiry. On the other hand, the overall mean would not be satisfactory to the experimenter of the study, who obviously suspects that there will be height differences among different fertilizer types. Instead, what is more acceptable to the experimenter are the plant height estimates after including the effect of the treatment $\tau_{i}$.
Note
The actual plant height can never be known because there is an unknown measurement error associated with any observation. This unknown error is associated with the ith treatment level, and the jth observation is denoted $\epsilon_{ij} \ (i = 1, 2, \ldots, T, \ j = 1, 2, \ldots, n_{i})$ is a random component (noise) that reflects the unexplained variability among plants within treatment levels.
Under the null hypothesis where the treatment effect is zero, the reduced model can be written $Y_{ij} = \mu + \epsilon_{ij}$.
Under the alternative hypothesis, where the treatment effects are not zero, the full model for at least one treatment level can be written $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$.
If $SSE(R)$ denotes the error sums of squares associated with the reduced model and $SSE(F)$ denotes the error sums of squares associated with the full model, we can utilize the General Linear Test approach to test the null hypothesis by using the test statistic: $F = \frac{\left(\dfrac{SSE(R) - SSE(F)}{df_{R} - df_{F}} \right)}{\left(\dfrac{SSE(F)}{df_{F}}\right)}$
which under the null hypothesis has an $F$ distribution with the numerator and denominator degrees of freedom equal to $df_{R}-df_{F}$ and $df_{F}$ respectively, where $df_{R}$ is the degrees of freedom associated with $SSE(R)$ and $df_{F}$ is the degree of freedom associated with $SSE(F)$. It is easy to see that $df_{R}=N-1$ and $df_{F}=N-T$ where $N = \sum_{i=1}^{N} n_{i}$. Also, $SSE(R) = \sum_{i} \sum_{j} \left(Y_{ij} - \bar{Y}_{..}\right)^{2} = SS_{Total} \quad \text{See Section 2.2}$
Therefore, \begin{align} F &= \frac{\left(\dfrac{SS_{Total} - SSE}{T-1}\right)}{\left(\dfrac{SSE}{df_{Error}}\right)} \[4pt] &= \frac{\left(\dfrac{SS_{Treatment}}{df_{Treatment}}\right)}{\left(\dfrac{SSE}{df_{Error}}\right)} \[4pt] &= \frac{MS_{Trt}}{MSE} \end{align}
Note that this is the same test statistic derived in Section 2.2 for testing the treatment significance. If the null hypothesis is true, then the treatment effect is not significant. If we reject the null hypothesis, then we conclude that the treatment effect is significant, which leads to the conclusion that at least one treatment level is better than the others! | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.01%3A_The_Model.txt |
Before we draw any conclusions about the significance of the model, we need to make sure we have a "valid" model. Like any other statistical procedure, the ANOVA has assumptions that must be met. Failure to meet these assumptions means any conclusions drawn from the model are not to be trusted.
Assumptions
So what are these assumptions being made to employ the ANOVA model? The errors are assumed to be independent and identically distributed (iid) with a normal distribution having a mean of 0 and unknown equal variance.
As the model residuals serve as estimates of the unknown error, diagnostic tests to check for validity of model assumptions are based on residual plots, and thus, the implementation of diagnostic tests is also called Residual Analysis.
Diagnostic Tests
Most useful is the residual vs. predicted value plot, which identifies the violations of zero mean and equal variance. Residuals are also plotted against the treatment levels to examine if the residual behavior differs among treatments.
The normality assumption is checked by using a normal probability plot.
Residual plots can help identify potential outliers, and the pattern of residuals vs. fitted values or treatments may suggest a transformation of the response variable.
Lesson 4: SLR Model Assumptions of STAT 501 online notes discuss various diagnostic procedures in more detail.
There are various statistical tests to check the validity of these assumptions, but some may not be that useful. For example, Bartlett’s test for homogeneity is too sensitive and indicates that problems exist when they really don’t. It turns out that the ANOVA is very robust and is not badly affected by minor violations of these assumptions. In practice, a good deal of common sense and the visual inspection of the residual plots are sufficient to determine if serious problems exist.
We will employ statistical software such as SAS to conduct the residual analysis. Here are common patterns that you may encounter in the residual analysis (i.e. plotting residuals, $e$, against the predicted values, $\hat{y}$).
Figure $\PageIndex{1a}$ shows the prototype plot when the ANOVA model is appropriate for data. The residuals are scattered randomly around mean zero and variability is constant (i.e. within the horizontal bands) for all groups.
Figure $\PageIndex{1b}$ suggests that although the variance is constant, there are some trends in the response that is not explained by a linear model. Using Figure $\PageIndex{1c}$, we can depict that the linear model is appropriate as the central trend in data is a line. However, the megaphone patterns in Figure $\PageIndex{1c}$ suggest that variance is not constant.
Alert!
A common problem encountered in ANOVA is when the variance of treatment levels is not equal (heterogeneity of variance). If the variance is increasing in proportion to the mean (panel (c) in Figure $1$), a logarithmic transformation of Y can "stabilize" the variances. If the residuals vs. predicted values instead show a curvilinear trend (panel (b) in Figure $1$), then a quadratic or other transformation may help. Since finding the correct transformation can be challenging, the Box-Cox method is often used to identify the appropriate transformation, given in terms of $\lambda$ as shown below.
$y_{i}^{(\lambda)} = \begin{cases} \frac{y_{i}^{\lambda} - 1}{\lambda}, \text{ if } \lambda \neq 0, \ \ln y_{i}, \text{ if } \lambda = 0 \end{cases}$
Some $\lambda$ values result some common transformations.
transformations.
$\lambda$ $Y^{\lambda}$ Transformation
2 $Y^{2}$ Square
1 $Y^{1}$ Original (No transform)
1/2 $\sqrt{Y}$ Square Root
0 $\log (Y)$ Logarithm
-1/2 $\frac{1}{\sqrt{Y}}$ Reciprocal Square Root
-1 $\frac{1}{Y}$ Reciprocal
Using Technology
Using Minitab
To run the Box-Cox procedure in Minitab, set up the data (Simulated Data), as a stacked format (a column with treatment (or trt combination) levels, and the second column with the response variable.
Treatment Response Variable
A 12
A 23
A 34
B 45
B 56
B 67
C 14
C 25
C 36
Steps in Minitab
1. On the Minitab toolbar, choose Stat > Control Charts > Box-Cox Transformation
2. Place "Response Variable" and "Treatment" in the boxes as shown below.
3. Click OK to finish. You will get an output like this:
In the upper right-hand box, the rounded value for $\lambda$ is given from which the appropriate transformation of the response variable can be found using the chart above. Note, with a $\lambda$ of 1, no transformation is recommended.
Using SAS
The Box-Cox procedure in SAS is more complicated in a general setting. It is done through the Transreg procedure, by obtaining the ANOVA solution with regression which first requires coding the treatment levels with effect coding discussed in Chapter 4.
However, for one-way ANOVA (ANOVA with only one factor) we can use the SAS Transreg procedure without much hassle.
Steps in SAS
Suppose we have SAS data as follows.
Obs Treatment ResponseVariable
1 A 12
2 A 23
3 A 34
4 B 45
5 B 56
6 B 67
7 C 14
8 C 25
9 C 36
We can use the following SAS commands to run the Box-Cox analysis.
proc transreg data=boxcoxSimData;
model boxcox(ResponseVariable)=class(Treatment);
run;
This would generate an output as follows, which suggests a transformation using $\lambda=1$ (i.e. no transformation).
Using R
Steps in R
Load the simulated data and perform the Box-Cox transformation. Note that simulated data are in the stacked format (a column with treatment levels and a column with the response variable)
setwd("~/path-to-folder/)
simulated_data<-read.table("simulated_data.txt",header=T)
attach(simulated_data)
library(AID)#Load package AID so that we can use the Box-Cox Procedure
boxcoxfr(Response_Variable,Treatment)#Box-Cox command for One-Way ANOVA
Output
Box-Cox power transformation
data: Response_Variable and Treatment
lambda.hat: 0.93
Shapiro-Wilk normality test for transformed data (alpha = 0.05)
Level statistic p.value Normality
1 A 0.9998983 0.9807382 YES
2 B 0.9999840 0.9923681 YES
3 C 0.9999151 0.9824033 YES
Bartlett's homogeneity test for transformed data (alpha = 0.05)
Level statistic p.value Homogeneity
1 All 0.008271728 0.9958727 YES
From the output, we can see that the lambda value for the transformation is 0.93 (the same value as Minitab suggested). Since this value is very close to 1 we can use $\lambda=1$ (no transformation).
In addition, from the output, we can see that normality exists in all 3 levels (Shapiro-Wilk test) and we have the same variance (Bartlett's test).
Alternative:
We can use the command boxcox from package MASS
library(MASS)
Box_Cox_Plot<-boxcox(aov(Response_Variable~Treatment),lambda=seq(-3,3,0.01))
From the plot, we can see the 95% CL. Since $\lambda=1$ is within the interval there is no need for transformation.
#Exact lambda
lambda<-Box_Cox_Plot$x[which.max(Box_Cox_Plot$y)] #0.93
detach(simulated_data) | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.02%3A_Assumptions_and_Diagnostics.txt |
The statistical software SAS is widely used in this course, and in previous sections we came across outputs generated through SAS programs. In this section, we begin to delve further into SAS programming with a special focus on ANOVA-related statistical procedures. The STAT 480-course series is also a useful resource for additional help.
Here is the program used to generate the summary output in Section 2.1:
```data greenhouse;
input Fert \$ Height;
```
The first line begins with the word `data` and invokes the data step. Notice that the end of each SAS statement has a semicolon. This is essential. In the dataset, the data to be used and its variables are named. Note that SAS assumes variables are numeric in the input statement, so if we are going to use a variable with alpha-numeric values (e.g. F1 or Control), then we have to follow the name of the variable in the input statement with a `\$` sign.
A simple way to input small datasets is shown in this code, wherein we embed the data in the program. This is done with the word `datalines`.
```datalines;
Control 21
Control 19.5
Control 22.5
Control 21.5
Control 20.5
Control 21
F1 32
F1 30.5
F1 25
F1 27.5
F1 28
F1 28.6
F2 22.5
F2 26
F2 28
F2 27
F2 26.5
F2 25.2
F3 28
F3 27.5
F3 31
F3 29.5
F3 30
F3 29.2
;
```
The semicolon here ends the dataset.
SAS then produces an output of interest using `proc` statements, short for “procedure”. You only need to use the first four letters, so SAS code is full of `proc` statements to do various tasks. Here we just wanted to print the data to be sure it read it in OK.
```proc print data= greenhouse;
title 'Raw Data for Greenhouse Data'; run;
```
Notice that the data set to be printed is specified in the proc print command. This is an important habit to develop because if not specified, SAS will use the last created data set, out of both input data sets, and output datasets that may have been generated as a result of any SAS procedures run up to that point.
The summary procedure which was then run can be very useful in both EDA (exploratory data analysis) and obtaining descriptive statistics such as mean, variance, minimum, maximum, etc. SAS procedures including the summary procedure categorical variables are specified in the class statement. Any variable NOT listed in the class statement is treated as a continuous variable. The target variable for which the summary will be made is specified by the `var` (for variable) statement.
The `output` statement creates an output dataset and the `out=` part assigns a name of your choice to the output. Descriptive statistics also can be named. For example, in the `output` statement below, `mean=mean` and `stderr=se` have named the mean of the variable `fert` as `mean` and standard error as `se`. The output data sets of any SAS procedure will not be automatically printed. As illustrated in the code below, the print procedure would then have to be used to print the generated output. In the proc print command a title can be included as a means of identifying and describing the output contents.
```proc summary data= greenhouse;
class fert;
var height;
output out=output1 mean=mean stderr=se;
run;
proc print data=output1;
title 'Summary Output for Greenhouse Data';
run;
```
The two commands `title``; run;`right after will erase the title assignment. This prevents the same title to be used in every output generated thereafter, which is a default feature in SAS.
```title; run;
```
Summary Output for Greenhouse Data
Obs Fert TYPE FREQ mean se
1 0 24 26.1667 0.75238
2 Control 1 6 21.0000 0.40825
3 F1 1 6 28.6000 0.99499
4 F2 1 6 25.8667 0.77531
5 F3 1 6 29.2000 0.52599
3.04: Greenhouse Example in SAS
In this section we will modify our previous program for greenhouse data to run the ANOVA model. The two SAS procedures that are commonly used are: `proc glm` and `proc mixed`.
```data greenhouse;
input fert \$ Height;
datalines;
Control 21
Control 19.5
Control 22.5
Control 21.5
Control 20.5
Control 21
F1 32
F1 30.5
F1 25
F1 27.5
F1 28
F1 28.6
F2 22.5
F2 26
F2 28
F2 27
F2 26.5
F2 25.2
F3 28
F3 27.5
F3 31
F3 29.5
F3 30
F3 29.2
;
/*
Any lines enclosed between starting with "/*" & ending with "*/" will be ignored by SAS.
*/
/* Recall how to print the data and obtain summary statistics. See section 3.3*/
/*To run the ANOVA model, use proc mixed procedure*/
proc mixed data=greenhouse method=type3 plots=all;
class fert;
model height=fert;
store myresults; /*myresults is an user defined object that stores results*/
title 'ANOVA of Greenhouse Data';
run;
/*To conduct the pairwise comparisons using Tukey adjustment*/
/*lsmeans statement below outputs the estimates means,
performs the Tukey paired comparisons, plots the data. */
/*Use proc plm procedure for post estimation analysis*/
proc plm restore=myresults;
lsmeans fert / adjust=tukey plot=meanplot cl lines;
run;
/* Testing for contrasts of interest with Bonferroni adjustment*/
proc plm restore=myresults;
lsmeans fert / adjust=tukey plot=meanplot cl lines;
estimate 'Compare control + F3 with F1 and F2 ' fert 1 -1 -1 1,
'Compare control + F2 with F1' fert 1 -2 1 0/ adjust=bon;
run;
``` | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.03%3A_Anatomy_of_SAS_Programming_for_ANOVA.txt |
The first output of the ANOVA procedure as shown below, gives useful details about the model.
ANOVA of Greenhouse Data: The Mixed Procedure
Model Information
Data Set WORK.GREENHOUSE
Dependent Variable Height
Covariance Structure Diagonal
Estimation Method Type 3
Residual Variance Method Factor
Fixed Effects SE Method Model-Based
Degrees of Freedom Method Residual
The output below titled ‘Type 3 Analysis of Variance’ is similar to the ANOVA table we are already familiar with. Note that it does not include the Total SS, however it can be computed as the sum of all SS values in the table.
Type 3 Analysis of Variance
Sources DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
fert 3 251.440000 83.813333 Var(Residual)+Q(fert) MS(Residual) 20 27.46 <.0001
Residual 20 61.033333 3.051667 Var(Residual)
The output above titled “Type 3 Tests of Fixed Effects” will display the $F_{calculated}$ and p-value for the test of any variables that are specified in the model statement. Additional information can also be requested. For example, the method = type 3 option will include the Expected Mean Squares for each source, which will prove to be useful and will be seen in Chapter 6.
The Mixed Procedure also produces the following diagnostic plots:
The following display is a result of the LSmeans statement in the PLM procedure which was included in the programming code.
Differences of fert Least Squares Means
fert Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
Control 21.0000 0.7132 20 29.45 <.0001 0.05 19.5124 22.4876
F1 28.6000 0.7132 20 40.10 <.0001 0.05 27.1124 30.0876
F2 25.8667 0.7132 20 36.27 <.0001 0.05 24.3790 27.3543
F3 29.2000 0.7132 20 40.94 <.001 0.05 27.7124 30.6876
In the "Least Squares Means" table above, note that the $t$-value and $Pr >|t|$ are testing null hypotheses that each group mean= 0. (These tests usually do not provide any useful information). The Lower and Upper values are the 95% confidence limits for the group means. Note also that the least square means are the same as the original arithmetic means that were generated in the Summary procedure in Section 3.3 because all 4 groups have the same sample sizes. With unequal sample sizes or if there is a covariate present, the least square means can differ from the original sample means.
Next, the Plot= mean plot option in the LSmeans statement yields a mean plot and also a diffogram, shown below. The confidence intervals in the mean plot are commonly used to identify the significantly different treatment levels or groups. If two confidence intervals do not overlap, then the difference between the two associated means is statistically significant, which is a valid conclusion. However, if they overlap, it may be the case that the difference might still be significant. Consequently, conclusions made based on the visual inspection of the mean plot may not match with those arrived at using the table of "Difference of Least Square Means", another output of the Tukey procedure, and is displayed below.
Notice that this is different from the previous table because it displays the results of each pairwise comparison. For example, the first row shows the comparison between the control and F1. The interpretation of these results is similar to any other confidence interval for the difference in two means—if the confidence interval does not contain zero, then the difference between the two associated means is statistically significant.
Differences of fert Least Squares Means
Adjustment for Multiple Comparisons: Tukey
fert _fert Estimate Standard Error DF t Value Pr > |t| Adj P Alpha Lower Upper Adj Lower Adj Upper
Control F1 -7.6000 1.0086 20 -7.54 <.0001 <.0001 0.05 -9.7038 -5.4962 -10.4229 -4.7771
Control F2 -4.8667 1.0086 20 -4.83 0.0001 0.0006 0.05 -6.9705 -2.7628 -7.6896 -2.0438
Control F3 -8.2000 1.0086 20 -8.13 <.0001 <.0001 0.05 -10.3038 -6.0962 -11.0229 -5.3771
F1 F2 2.7333 1.0086 20 2.71 0.0135 0.0599 0.05 0.6295 4.8372 -0.08957 5.5562
F1 F3 -0.6000 1.0086 20 -0.59 0.5586 0.9324 0.05 -2.7038 1.5038 -3.4229 2.2229
F2 F3 -3.3333 1.0086 20 -3.30 0.0035 .0171 0.05 -5.4372 -1.2295 -6.1562 -0.5104
This discrepancy between the mean plot and the "Difference of Least Square Means" results occurs because the testing is done in terms of the difference of two means, using the standard error of the difference of the two-sample means, but the confidence intervals of the mean plot are computed for the individual means which are in terms of the standard error of individual sample means. Consistent results can be achieved by using the diffogram as discussed below or the confidence intervals displayed in the "difference in mean plot" available in SAS 14, but not included here.
The diffogram has two useful features. It allows one to identify the significant mean pairs and also gives estimates of the individual means. The diagonal line shown in the diffogram is used as a reference line. Each group (or factor level) is marked on the horizontal and vertical axes and has vertical and horizontal reference lines with their intersection point falling on the diagonal reference line. The $x$ or the $y$ coordinates of this intersection point which are equal is the sample mean of that group. For example, the sample mean for the Control group is about 21, which matches with the estimate provided in the "Least Squares Means" table displayed above. Furthermore, each slanted line represents a mean pair. Start with any group label from the horizontal axis and run your cursor up, along the associated vertical line until it meets a slanted line, and then go across the intersecting horizontal line to identify the other group (or factor level). For example, the lowermost solid line (colored blue) represents the Control and F2. As stated at the bottom of the chart, the solid (or blue) lines indicate significant pairs, and the broken (or red) lines correspond to the non-significant pairs. Furthermore, a line corresponding to a nonsignificant pair will cross the diagonal reference line.
The non-overlapping confidence intervals in the mean plot above indicate that the average plant height due to control is significantly different from those of the other 3 fertilizer levels and that the F2 fertilizer type yields a statistically different average plant height from F3. The diffogram also delivers the same conclusions and so, in this example, conclusions are not contradictory. In general, the diffogram always provides the same conclusions as derived from the confidence intervals of difference of least-square means shown in the "Difference of Least Square Means" table, but the conclusions based on the mean plot may differ.
There are two contrasts of interest: contrast to compare the control and F3 with F1 (i.e. $\mu_{control} - \mu_{F1} - \mu_{F2} + \mu_{F3}$) and the contrast to compare control and F2 with F1 (i.e., $\mu_{control} - 2 \mu_{F1} + \mu_{F2}$). Since we are testing for two contrasts, we should adjust for multiple comparisons. We use Bonferroni adjustment. In SAS, we can use the estimate command under proc plm to make these computations.
In general, the estimate command estimates linear combinations of model parameters and performs t-tests on them. Contrasts are linear combinations that satisfy a special condition. We will discuss the model parameters in Chapter 4.
Estimates
Adjustment for Multiplicity: Bonferroni
Label Estimate Standard Error DF t Value Pr > |t| Adj P
Compare control + F3 with F1 and F2 -4.2667 1.4263 20 -2.99 0.0072 0.0144
Compare control + F2 with F1 -10.3333 1.7469 20 -5.92 <.0001 <.0001
SAS returns both unadjusted and adjusted $p$-values. Suppose we wanted to make the comparisons at 1% level. If we ignored the multiple comparisons (i.e. using unadjusted $p$-values), the both comparisons are statistically significant. However, if we consider the adjusted $p$-values, we will fail to reject the hypothesis corresponding to the first contrast at the 1% level. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.05%3A_SAS_Output_for_ANOVA.txt |
Step 1: Import the data
The data (Lesson 1 Data) can be copied and pasted from a word processor into a worksheet in Minitab:
Step 2: Run the ANOVA
To run the ANOVA, we use the sequence of tool-bar tabs: Stat > ANOVA > One-way…
You then get the pop-up box seen below. Be sure to select from the drop-down in the upper right, "Response data are in a separate column for each factor level":
Then we double-click from the left-hand list of factor levels to the input box labeled "Responses", and then click on the box labeled Comparisons.
We check the box for Tukey and then exit by clicking on OK. To generate the Diagnostics, we then click on the box for Graphs and select the "Three in one" option:
You can now "back out" by clicking on OK in each nested panel.
Step 3: Results
Now in the Session Window, we see the ANOVA table along with the results of the Tukey Mean Comparison:
One-Way ANOVA: Control, F1, F2, F3
Method
Null Hypothesis: All means are equal
Alternative Hypothesis: Not all means are equal
Significance Level: $\alpha=0.05$
Equal variances were assumed for the analysis.
Analysis of Variance
(Extracted from the output that follows from above.)
Grouping Information Using Tukey Method
Means that do not share a letter are significantly different.
As can be seen, Minitab provides a difference in means plot, which can be conveniently used to identify the significantly different means by following the rule: if the confidence interval does not cross the vertical zero line, then the difference between the two associated means is statistically significant.
The diagnostic (residual) plots, as we asked for them, are in one figure:
Note that the Normal Probability plot is reversed (i.e, the axes are switched) compared to the SAS output. Assessing straight line adherence is the same, and the residual analysis provided is comparable to SAS output.
3.07: One-Way ANOVA Greenhouse Example in R
R Instructions: Code for the Greenhouse Data
• Load the greenhouse data.
• Calculate the overall mean, standard deviation, and standard error.
• Calculate the mean, standard deviation, and standard error for each group.
• Produce a boxplot to plot the differences in heights for each fertilizer.
• Produce a "means plot" (interval plot) to view the differences in heights for each fertilizer.
• Obtain the ANOVA table.
• Obtain Tukey’s multiple comparisons CIs and difference in means plot.
• Produce diagnostic (residuals) plots.
• Power analysis.
```setwd("~/path-to-folder/")
greenhouse_data<-read.table("greenhouse_data.txt",header=T)
```
Note that greenhouse data are in separate columns.
```attach(greenhouse_data)
my_data<-stack(greenhouse_data)
```
With this command, we put our data in a stacked format (the first column has the response variable (values) and the second column has the treatment levels (ind).
To calculate the overall mean, standard deviation, and standard error we can use the following commands:
```overall_mean<-mean(my_data\$values) #26.16667
overall_sd<-sd(my_data\$values) #3.685892
overall_standard_error<-overall_sd/sqrt(length(my_data\$values)) #0.7523795
```
To calculate the group means we can use the following command:
```group_means<-aggregate(my_data[, 1],list(my_data\$ind),mean)
# group_means
# Group.1 x
# 1 Control 21.00000
# 2 F1 28.60000
# 3 F2 25.86667
# 4 F3 29.20000
```
To calculate the group standard deviations and standard errors we can use the following commands:
```group_sd<-aggregate(my_data[, 1],list(my_data\$ind),sd)
# group_sd Group.1 x
# 1 Control 1.000000
# 2 F1 2.437212
# 3 F2 1.899123
# 4 F3 1.288410
```
```group_standard_error<-group_sd\$x/sqrt(length(my_data\$ind)/4)
# group_standard_error
# 0.4082483 0.9949874 0.7753136 0.5259911
```
To produce the Boxplot we can use the following commands:
```library("ggpubr")
boxplot(values~ind,data=my_data,
xlab="Fertilizer",ylab="Plant Height",
main="Distribution of Plant Heights by Fertilizer",
frame=TRUE)
```
To produce the means plot (interval plot) we can use the following commands:
```library("gplots")
plotmeans(values~ind,data=my_data,connect=FALSE,
xlab="Fertilizer",ylab="Plant Height",
main="Means Plot with 95% CI")
```
To obtain the ANOVA table we can use the following commands:
```anova<-aov(values~ind,my_data)
summary(anova)
```
The command summary (anova) will give you the following output:
```summary(anova)
Df Sum Sq Mean Sq F value Pr(>F)
ind 3 251.44 83.81 27.46 2.71e-07 ***
Residuals 20 61.03 3.05
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
```
We can see the degrees of freedom in the first column, the sum of squares in the second column, the mean sum of squares in the third column, the \(F\)-test statistic in the fourth column, and finally, we can see the \(p\)-value.
Note that the output doesn't give the \(SSTO\). To find it, use the identity \(SSTO=SSR+SSE\). Similarly, for the \(df\) associated with \(SSTO\), add the \(df\) of \(SSR\) and \(SSE\).
For our example, \(SSTO=251.44+61.03=312.47\)
To obtain Tukey multiple comparisons of means with a 95% family-wise confidence level we use the following command:
```library(multcomp)
library(multcompView)
tukey_multiple_comparisons<-TukeyHSD(anova,conf.level=0.95)
plot(tukey_multiple_comparisons)
tukey_multiple_comparisons
Tukey multiple comparisons of means
95% family-wise confidence level
Fit: aov(formula = values ~ ind, data = my_data)
\$ind
diff lwr upr p adj
F1-Control 7.600000 4.7770648 10.42293521 0.0000016
F2-Control 4.866667 2.0437315 7.68960188 0.0005509
F3-Control 8.200000 5.3770648 11.02293521 0.0000005
F2-F1 -2.733333 -5.5562685 0.08960188 0.0598655
F3-F1 0.600000 -2.2229352 3.42293521 0.9324380
F3-F2 3.333333 0.5103981 6.15626854 0.0171033
```
Based on this output, the Control group is significantly different from the 3 treatment groups and F3 is significantly different from F2.
To produce diagnostic (residuals) plots we use the following commands:
```#Residuals vs Fits plot
plot(anova,1)
```
```#QQ plot
plot(anova,2)
```
```#Histogram of residuals
residuals<-anova\$res #with this command we get the residuals from ANOVA
hist(residuals)
``` | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.06%3A_One-Way_ANOVA_Greenhouse_Example_in_Minitab.txt |
After completing a statistical test, conclusions are drawn about the null hypothesis. In cases where the null hypothesis is not rejected, a researcher may still feel that the treatment did have an effect. Let's say that three weight loss treatments are conducted. At the end of the study, the researcher analyzes the data and finds there are no differences among the treatments. The researcher believes that there really are differences. While you might think this is just wishful thinking on the part of the researcher, there MAY be a statistical reason for the lack of significant findings.
At this point, the researcher can run a power analysis. Recall from your introductory text or course that power is the ability to reject the null when the null is really false. The factors that impact power are sample size (larger samples lead to more power), the effect size (treatments that result in larger differences between groups will have differences that are more readily found), the variability of the experiment, and the significance of the type 1 error.
As a note, the most common type of power analysis are those that calculate needed sample sizes for experimental designs. These analyses take advantage of pilot data or previous research. When power analysis is done ahead of time, it is a PROSPECTIVE power analysis. This example is a retrospective power analysis, as it is done after the experiment is completed.
So back to our greenhouse example. Typically we want power to be at 80%. Again, power represents our ability to reject the null when it is false, so a power of 80% means that 80% of the time our test identifies a difference in at least one of the means correctly. The converse of this is that 20% of the time we risk not rejecting the null when we really should be rejecting the null.
Using our greenhouse example, we can run a retrospective power analysis (just a reminder, we typically don't do this unless we have some reason to suspect the power of our test was very low).
This is one analysis where Minitab is much easier and still just as accurate as SAS, so we will use Minitab to illustrate this simple power analysis in detail and follow up the analysis with SAS.
Power Analysis Techniques
Using SAS
Steps in SAS
Let us now consider running the power analysis in SAS. In our greenhouse example with 4 treatments (control, F1, F2, and F3), the estimated means were $21, 28.6, 25.877, 29.2$ respectively. Using ANOVA, the estimated standard deviation of errors was 1.747 (which is obtained by $\sqrt{MSE}=\sqrt{3.0517}$. There are 6 replicates for each treatment. Using these values, we could employ SAS POWER procedure to compute the power of our study retrospectively.
proc power;
onewayanova alpha=.05 test=overall
groupmeans=(21 28.6 25.87, 29.2)
npergroup=6 stddev=1.747
power=.;
run;
As with MINITAB, we see that the retrospective power analysis for our greenhouse example yields a power of 1. If we re-do the analysis ignoring the CONTROL treatment group, then we only have 3 treatment groups: F1, F2, and F3. The ANOVA with only these three treatments yields an MSE of $3.735556$. Therefore the estimated standard deviation of errors would be $1.933$. We will have a power of 0.731 in this modified scenario, as shown in the below output.
Suppose, we ask the question of how many replicates we would need to obtain at least 80% power to detect a difference in the means of our greenhouse example with the same group means but with different variability in data (i.e. standard deviations should be different). We can use SAS POWER to answer this question.
We can see that with a standard deviation of 1.747, if we have only 2 replicates in each of the four treatments we can detect the differences in greenhouse example means with more than 80% power. However, as the data get noisier (i.e. as standard deviation increases) we need more replicates to achieve 80% power in the same example.
Using Minitab
Steps in Minitab
In Minitab select STAT > Power and Sample Size > One-Way ANOVA
Since we have a one-way ANOVA we select this test (you can see there are power analyses for many different tests, and SAS will allow even more complicated options).
When you look at our filled-in dialogue box, you will notice we have not entered a value for power. This is because Minitab will calculate whichever box you leave blank (so if we needed sample size we would leave sample size blank and fill in a value for power). From our example, we know the number of levels is 4 because we have four treatments. We have six observations for each treatment so the sample size is 6. The value for the maximum difference in the means is 8.2 (we simply subtracted the smallest mean from the largest mean, and the standard deviation is 1.747. Where did this come from? The MSE, available from the ANOVA table, is about 3, and hence the standard deviation is $\sqrt{3}=1.747$).
After we click OK we get the following output:
If you follow this graph you see that power is on the y-axis and the power for the specific setting is indicated by a red dot. It is hard to find, but if you look carefully the red dot corresponds to a power of 1. In practice, this is very unusual, but can be easily explained given that the greenhouse data was constructed to show differences.
We can ask the question, what about differences among the treatment groups, not considering the control? All we need to do is modify some of the input in Minitab.
Note the differences here as in the previous screenshot. We now have 3 levels because we are only considering the three treatments. The maximum differences among the means and also the standard deviation are also different.
The output now is much easier to see:
Here we can see the power is lower than when including the control. The main reason for this decrease is that the difference between the means is smaller.
You can experiment with the power function in Minitab to provide you with sample sizes, etc. for various powers. Below is some sample output when we ask for various power curves for various sample sizes, a kind of "what if" scenario.
Just as a reminder, power analyses are most often performed BEFORE an experiment is conducted, but occasionally, a power analysis can provide some evidence as to why significant differences were not found.
Using R
Steps in R
With the following commands we will get the power analysis for the greenhouse example:
groupmeans<-c(21,28.6,25.87,29.2)
power.anova.test(groups=4,n=6,between.var=var(groupmeans),within.var=3.05,sig.level=0.05)
Balanced one-way analysis of variance power calculation
groups = 4
n = 6
between.var = 13.96823
within.var = 3.05
sig.level = 0.05
power = 1
NOTE: n is the number in each group.
If we want to produce a power plot by increasing the sample size and the variance (like the one produced by SAS) we can use the following commands:
groupmeans<-c(21,28.6,25.87,29.2)
n<-c(seq(2,8,by=1))
p<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=3.05,sig.level=0.05)
p1<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=4,sig.level=0.05)
p2<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=6.25,sig.level=0.05)
p3<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=9,sig.level=0.05)
p4<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=16.05,sig.level=0.05)
p5<power.anova.test(groups=4,n=n,between.var=var(groupmeans),within.var=25,sig.level=0.05)
plot(n,p$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA", ylim=c(0,1)) lines(n,p$power, col = "blue")
abline(h=0.80)
par(new=TRUE)
plot(n,p1$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA",ylim=c(0,1)) lines(n,p1$power, col = "red")
par(new=TRUE)
plot(n,p2$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA",ylim=c(0,1)) lines(n,p2$power, col = "green")
par(new=TRUE)
plot(n,p3$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA",ylim=c(0,1)) lines(n,p3$power, col = "brown")
par(new=TRUE)
plot(n,p4$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA",ylim=c(0,1)) lines(n,p4$power, col = "purple")
par(new=TRUE)
plot(n,p5$power,ylab="Power",xlab="Sample size per group",main="Overall F test for One-Way ANOVA",ylim=c(0,1)) lines(n,p5$power, col = "gray")
text(locator(1),"var=3.05",col="blue")
text(locator(1),"var=4",col="red")
text(locator(1),"var=6.25",col="green")
text(locator(1),"var=9",col="brown")
text(locator(1),"var=16",col="purple")
text(locator(1),"var=25",col="gray") | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.08%3A_Power_Analysis.txt |
Exercise $1$: Diet Study
The weight gain due to 4 different diets given to 24 calves is shown below.
a) Write the appropriate null and alternative hypotheses to test if the weight gain differs significantly among the 4 diets.
Solution
$H_{0}: \ \mu_{1} = \mu_{2} = \mu_{3} = \mu_{4}$ vs. $H_{a}: \ \mu_{i} \neq \mu_{j} for some \(i, j = 1,2,3,4$ OR "Not all means are equal"
Note: Here, $\mu_{i}, \ i=1,2,3,4$ are the actual mean weight gains due to diet1, diet2, diet3, and diet4, respectively.
b) Analyze the data and write your conclusion.
Solution
Using SAS...
data Lesson3_ex1;
input diet \$ wt_gain;
datalines;
diet1 12
diet1 10
diet1 13
diet1 11
diet1 12
diet1 09
diet2 18
diet2 19
diet2 18
diet2 18
diet2 19
diet2 19
diet3 10
diet3 12
diet3 13
diet3 16
diet3 14
diet3 13
diet4 19
diet4 20
diet4 18
diet4 19
diet4 18
diet4 19
;
ods graphics on;
proc mixed data= Lesson3_ex1 plots=all; class diet;
model wt_gain = diet;
contrast 'Compare diet1 with diets 2,3,4 combined ' diet 3 -1 -1 -1;
store result1;
title 'ANOVA of Weight Gain Data';
run;
ods html style=statistical sge=on;
proc plm restore=result1;
lsmeans diet/ adjust=tukey plot=meanplot cl lines;
run;
The ANOVA results shown below indicate that the diet effect is significant with an $F$-value of 51.27 ($p$-value <.0001). This means that not all diets provide the same mean weight gain. The diffogram below indicates the significant different pairs of diets identified by solid blue lines. The estimated mean weight gains from diets 1, 3, 2, and 4 are 11, 13, 18.1, and 19 units respectively. The diet pairs that have significantly different mean weight gains are (1,2), (1,4), (3,2), and (3,4).
Partial Output:
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
diet 3 20 51.27 <.0001
Exercise $2$: Commuter Times
Above is a diffogram depicting the differences in daily commuter time (in hours) among regions of a metropolitan city. Answer the following.
a) Name the regions included in the study.
Solution
SOUT, MIDW, NORT, and WEST
b) How many red or blue lines are to be expected?
Solution
4 choose 2 = 6 red or blue lines
c) Which pairs of regions have significantly different average commuter times?
Solution
(SOUT and NORT), (SOUT and WEST), (MIDW and NORT), and (MIDW and WEST) have significantly different mean commuter times.
d) Write down the estimated mean daily commuter time for each region.
Solution
Region SOUT MIDW NORT WEST
Estimated mean commuter time in hours 8.7 10.5 16 16.2
3.10: Chapter 3 Summary
The primary focus in this chapter was to establish the foundation for developing mathematical models for a one-way ANOVA setting. The effects model was then discussed along with the ANOVA model assumptions and diagnostics. The other focus was to illustrate, using the greenhouse example, how SAS and Minitab can be utilized to run an ANOVA model. Sections 3.3-3.6 were devoted to this purpose and include details on SAS and Minitab ANOVA basics, together with guidance in the interpretation of the outputs. Software-based diagnostics tests to detect the validity of model assumptions were also discussed, along with the power analysis procedure which computes any one of the four quantities of sample size, power, effect size, and the significance level, given the other three.
The next chapter will be a continuation of this lesson. Three more different versions of ANOVA model equations that represent a single factor experiment will be discussed. These are known as Overall Mean, Cell Means, and Dummy Variable Regression models. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/03%3A_ANOVA_Models_Part_I/3.09%3A_Try_It.txt |
Objectives
By the end of this chapter, students will be able to:
• Apply the overall mean, cell means, and dummy variable regression models for a one-way ANOVA and interpret the results.
• Identify the design matrix and the parameter vector for each ANOVA model studied.
• Recognize aspects of ANOVA programming computations.
This is a continuation of the previous lesson, and in this lesson, three more alternative ANOVA models are introduced. ANOVA models are derived under the assumption of linearity of model parameters and additivity of model terms so that every model will follow the general linear model (GLM): $Y=X \beta + \mathcal{E}$. In later sections of this lesson, we will see that the appropriate choice of $X$, the design matrix, will result in a different ANOVA model. This lesson will also shed insight into the similarities of how ANOVA calculations are done by most software, regardless of which model is being used. Finally, the concept of a study diagram is also discussed, demonstrating its usefulness when building a statistical model and designing an experiment.
04: ANOVA Models Part II
In the past lessons, we carried out the ANOVA computations conceptually in terms of deviations from means. For the calculation of total variance, we used the deviations of the individual observations from the overall mean, while the treatment SS was calculated using the deviations of treatment level means from the overall mean, and the residual or error SS was calculated using the deviations of individual observations from treatment level means. In practice, however, to achieve higher computational efficiency, SS for ANOVA is computed utilizing the following mathematical identity: $SS = \sum \left(Y_{i} - \bar{Y}\right)^{2} = \sum Y_{i}^{2} - \frac{\left(\sum Y_{i}\right)^{2}}{N}$
This identity is commonly called the working formula or machine formula. The second term on the right-hand side is often referred to as the correction factor (CF).
For computing the SS for the total variance of the responses, the formula above can be used as it is, but modifications need be made for others. For example, to compute the treatment SS, the above equation has to be modified as: $SS_{treatment} = \sum_{i=1}^{T} \frac{\left(\sum_{j=1}^{n_{i}} Y_{ij}\right)^{2}}{n_{i}} - \frac{\left(\sum Y_{i}\right)^{2}}{N}$
We will examine three new ANOVA models (Models 1, 2, 3), as well as the effects model (Model 4) from the previous lesson, defined as follows:
Model 1 - The Overall Mean Model
$Y_{ij} = \mu + \epsilon_{ij}$ which simply fits an overall or "grand" mean'. This model reflects the situation where $H_{0}$ is true, implying that $\mu_{1} = \mu_{2} = \ldots = \mu_{T}$.
Model 2 - The Cell Means Model
$Y_{ij} = \mu_{i} + \epsilon_{ij}$ where $\mu_{i}, \ i=1,2,...,T$ are the factor level means. Note that in this model, there is no overall mean being fitted.
Model 3 - Dummy Variable Regression
$Y_{ij} = \mu + \mu_{i} + \epsilon_{ij}, \text{ fitted as } Y_{ij} = \beta_{0} + \beta_{Level \ 1} + \beta_{Level \ 2} + \ldots + \beta_{Level \ r-1} + \epsilon_{ij}$ where $\beta_{Level \ 1}, \beta_{Level \ 2}, \ldots, \beta_{Level \ T-1}$ are regression coefficients for $T-1$ indicator-coded regression "dummy" variables that are correspond to the $T-1$ categorical factor levels. The $T^{th}$ factor level mean is given by the regression intercept $\beta_{0}$.
Model 4 - The Effects Model
$Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$ where $\tau_{i}$ are the the deviations of each factor level mean from the overall mean so that $\sum_{i=1}^{T} \tau_{i} = 0$.
Each of these four models can be written as a general linear model (GLM): $\mathbf{Y} = \mathbf{X} \beta + \boldsymbol{\mathcal{E}}$ simply by changing the design matrix $\mathbf{X}$. Thus to perform the data analysis, in terms of the computer coding instructions, the appropriate numerical values for the $\mathbf{X}$ matrix elements will need to be inputted. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.01%3A_How_is_ANOVA_Calculated.txt |
Model 1 - The Overall Mean Model
$Y_{ij} = \mu + \epsilon_{ij}$ which simply fits an overall or "grand" mean. This model reflects the situation where $H_{0}$ is true, implying that $\mu_{1} = \mu_{2} = \ldots = \mu_{T}$.
To understand how various facades of the model relate to each other, let us look at a toy example with 3 treatments (or factor levels) and 2 replicates of each treatment.
We have 6 observations, which means that $\mathbf{Y}$ is a column vector of dimension 6 and so is the error vector $\boldsymbol{\mathcal{E}}$ where its elements are the random error values associated with the 6 observations. In the GLM model of $\mathbf{Y} = \mathbf{X} \boldsymbol{\beta} + \boldsymbol{\mathcal{E}}$, the design matrix $\mathbf{X}$ for the overall mean model turns out to be a 6-dimensional column vector of ones. The parameter vector, $\boldsymbol{\beta}$, is a scalar equal to $\mu$, the overall population mean.
That is, $\mathbf{Y} = \begin{bmatrix} 2 \ 1 \ 3 \ 4 \ 5 \ 6 \end{bmatrix}, \ \mathbf{X} = \begin{bmatrix} 1 \ 1 \ 1 \ 1 \ 1 \ 1 \end{bmatrix}, \boldsymbol{\beta} = [\mu], \text{ and } \boldsymbol{\epsilon} = \begin{bmatrix} \epsilon_{1} \ \epsilon_{2} \ \epsilon_{3} \ \epsilon_{4} \ \epsilon_{5} \ \epsilon_{6} \end{bmatrix}$
Using the method of least squares, the estimates of the parameters in $\boldsymbol{\beta}$ are obtained as: $\boldsymbol{\hat{\beta}} = (\mathbf{X}' \mathbf{X})^{-1} \mathbf{X}' \mathbf{Y}$
Using the estimate $\boldsymbol{\hat{\beta}}$, the $i^{th}$ predicted response $\mathbf{\hat{y}_{i}}$ can be computed as $\mathbf{\hat{y}_{i}} = \mathbf{x_{i}}'$, where $\mathbf{x_{i}}'$ denotes the $i^{th}$ row vector of the design matrix.
In this simplest of cases, we can see how the matrix algebra works. The term $\mathbf{X}' \mathbf{X}$ would be: $[1 \ 1 \ 1 \ 1 \ 1 \ 1] * \begin{bmatrix} 1 \ 1 \ 1 \ 1 \ 1 \ 1 \end{bmatrix} = 1 + 1 + 1 + 1 + 1 + 1 = 6 = n$
The term $\mathbf{X}' \mathbf{Y}$ would be: $[1 \ 1 \ 1 \ 1 \ 1 \ 1] * \begin{bmatrix} 2 \ 1 \ 3 \ 4 \ 5 \ 6 \end{bmatrix} = 2 + 1 + 3 + 4 + 5 + 6 = 21 = \sum Y_{i}$
So in this case, the estimate $\mathbf{b}$ as expected is simply the overall mean $= \hat{\mu} = \bar{y}_{..} = 21/6 =3.5$
Note that the exponent of $\mathbf{X}' \mathbf{X}$ in the formula above indicates arithmetic division as $\mathbf{X}' \mathbf{X}$ is a scalar increase in this case. In the more general setting, the superscript of '-1 ' indicates the inverse operation in matrix algebra.
To perform these matrix operations in SAS IML, we will open a regular SAS editor window, and then copy and paste three components from the file (IML Design Matrices) as shown below.
SAS: Overall Mean Model
Steps in SAS
Step 1
Procedure initiation, and specification of the dependent variable vector, $\mathbf{Y}$.
For our example we have:
/* Initiate IML, define response variable */
proc iml;
y={
2,
1,
3,
4,
6,
5};
Step 2
We then enter a design matrix $\mathbf{X}$. For the Overall Mean model and our example data, we have:
x={
1,
1,
1,
1,
1,
1};
Step 3
We can now copy and paste a program for the matrix computations to generate results (regression coefficients and ANOVA output):
beta=inv(x*x)*x*y;
beta_label={"Beta_0","Beta_1","Beta_2","Beta_3"};
print beta [label="Regression Coefficients"
rowname=beta_label];
n=nrow(y);
p=ncol(x);
j=j(n,n,1);
ss_tot = (y*y) - (1/n)*(y*j)*y;
ss_trt = (beta*(x*y)) - (1/n)*(y*j)*y;
ss_error = ss_tot - ss_trt;
total_df=n-1;
trt_df=p-1;
error_df=n-p;
ms_trt = ss_trt/(p-1);
ms_error = ss_error / error_df;
F=ms_trt/ms_error;
empty={.};
row_label= {"Treatment", "Error", "Total"};
col_label={"df" "SS" "MS" "F"};
trt_row= trt_df || ss_trt || ms_trt || F;
error_row= error_df || ss_error || ms_error || empty;
tot_row=total_df || ss_tot || empty || empty;
aov = trt_row // error_row // tot_row;
print aov [label="ANOVA"
rowname=row_label
colname=col_label];
Here is a quick video walk-through to show you the process for how you can do this in SAS. (Right-click and select "Show All" if your browser does not display the entire screencast window.)
Video $1$: Walkthrough for ANOVA using the SAS overall mean model.
The program can then be run to produce the following output:
Regression Coefficients
Beta_0 3.5
We see the estimate of the regression coefficient for $\beta_{0}$ equals 3.5, which indeed is the overall mean of the response variable, and is also the same value we obtained above using "by-hand" calculations. In this simple case, where the treatment factor has not entered the model, the only item of interest from the ANOVA table would be the $SS_{Error}$ for later use in the General Linear $F$-test.
If you like to see the internal calculations further, you may optionally add the following few lines, to the end of the calculation code.
/* (Optional) Intermediates in the matrix computations */
xprimex=x*x; print xprimex;
xprimey=x
*y; print xprimey;
xprimexinv=inv(x*x); print xprimexinv;
check=xprimexinv*xprimex; print check;
SumY2=beta
*(x*y); print SumY2;
CF=(1/n)*(y
*j)*y; print CF;
This additional code produces the following output:
xprimex xprimey xprimeinv
6 21 0.1666667
check SumY2 CF
1 73.5 73.5
From this we can verify the computations for the $SS_{treatment} = \sum Y_{i}^{2} - \frac{\left(\sum Y_{i}\right)^{2}}{n} = \sum Y_{2} - CF = 0$.
The "check" calculation confirms that $(\mathbf{X}' \mathbf{X})^{-1} \mathbf{X}' \mathbf{X} = 1$, which in fact defines the matrix division operation. In this simple case, it amounts to simple division by $n$, but in other models that we will work with, the matrix division process is more complicated and is explained here. In general, the inverse of a matrix $\mathbf{A}$, denoted $\mathbf{A}^{-1}$, is defined by the matrix identity $\mathbf{A}^{-1} \mathbf{A} = I$, where $I$ is the identity matrix (a diagonal matrix of $1$’s). In this example, $\mathbf{A}$ is replaced by $\mathbf{X}' \mathbf{X}$, which is a scalar and equals 6.
R: Overall Mean Model
Steps in R
1. Define response variable and design matrix
y<-matrix(c(2,1,3,4,6,5), ncol=1)
x<-matrix(c(1,1,1,1,1,1), ncol=1)
2. Regression coefficients
beta<-solve(t(x)%*%x)%*%(t(x)%*%y) #3.5
3. Calculate the entries of the ANOVA Table
n<-nrow(y)
p<-ncol(x)
J<-matrix(1,n,n)
ss_tot = (t(y)%*%y) - (1/n)*(t(y)%*%J)%*%y #17.5
ss_trt = t(beta)%*%(t(x)%*%y) - (1/n)*(t(y)%*%J)%*%y #0
ss_error = ss_tot - ss_trt #17.5
total_df=n-1 #5
trt_df=p-1 #0
error_df=n-p #5
MS_trt = ss_trt/(p-1)
MS_error = ss_error / error_df #3.5
F=MS_trt/MS_error
4. Creating the ANOVA table
ANOVA <- data.frame(
c ("","Treatment","Error", "Total"),
c("DF", trt_df,error_df,total_df),
c("SS", ss_trt, ss_error, ss_tot),
c("MS", "", MS_error, ""),
c("F","","",""),
stringsAsFactors = FALSE)
names(ANOVA) <- c(" ", " ", " ","","")
5. Print the ANOVA table
print(ANOVA)
# 1 DF SS MS F
# 2 Treatment 0 0
# 3 Error 5 17.5 3.5
# 4 Total 5 17.5
6. Intermediates in the matrix computations
xprimex<-t(x)%*%x # 6
xprimey<-t(x)%*%y # 21
xprimexinv<-solve(t(x)%*%x) # 0.1666667
check<-xprimexinv*xprimex # 1
SumY2<-t(beta)%*%(t(x)%*%y) # 73.5
CF<-(1/n)*(t(y)%*%J)%*%y # 73.5` | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.02%3A_The_Overall_Mean_Model.txt |
Model 2 - The Cell Means Model
$Y_{ij} = \mu_{i} + \epsilon_{ij}$ where $\mu_{i}, \ i=1,2, \ldots, T$ are the factor level means. Note that in this model, there is no overall mean being fitted.
The cell means model does not fit an overall mean, but instead fits an individual mean for each of the treatment levels. Let us run this model for the same data assuming that each pair of observations arise from one treatment level, so that T, the number of treatment levels equals 3. We then have to replace the design matrix in the IML code with:
/* The Cell Means Model */
x={
1 0 0,
1 0 0,
0 1 0,
0 1 0,
0 0 1,
0 0 1};
Notice that each column represents a specific treatment level and is using indicator coding: $1$ for the rows corresponding to the observations receiving the specified treatment level, and $0$ for the other rows. It can be seen that $r=2$ is the number of replicates for each treatment level. Observe that column 1 generates the mean for treatment level 1, column 2 for treatment level 2, and column 3 for treatment level 3.
To write the cell means model as a GLM, let $\mathbf{X} = \begin{bmatrix} 1 & 0 & 0 \ 1 & 0 & 0 \ 0 & 1 & 0 \ 0 & 1 & 0 \ 0 & 0 & 1 \ 0 & 0 & 1 \end{bmatrix} = \begin{bmatrix} \mathbf{x_{1}}' \ \mathbf{x_{2}}' \ \mathbf{x_{3}}' \ \mathbf{x_{4}}' \ \mathbf{x_{5}}' \ \mathbf{x_{6}}' \end{bmatrix}$ where $\mathbf{x_{i}}'$ is the $i^{th}$ row vector of the design matrix.
The parameter vector $\boldsymbol{\beta}$ is a 3-dimensional column vector and is defined by $\boldsymbol{\beta} = \begin{bmatrix} \beta_{0} \ \beta_{1} \ \beta_{2} \end{bmatrix} = \begin{bmatrix} \mu_{1} \ \mu_{2} \ \mu_{3} \end{bmatrix}$
The parameter estimates $\boldsymbol{\hat{\beta}}$ can again be found using the least squares method. One can verify that $\mu_{i} = \bar{y}_{i}$, the $i^{th}$ treatment mean, for $i=1,2,3$. Using this estimate, the resulting estimated regression equation for the cell means model is, $\boldsymbol{\hat{Y}} = \mathbf{X} \boldsymbol{\hat{\beta}}$ which produces $\boldsymbol{\hat{y}_{i}} = \boldsymbol{x_{i}}' \begin{bmatrix} \hat{\mu}_{1} \ \hat{\mu}_{2} \ \hat{\mu}_{3} \end{bmatrix}$.
We then re-run the program with the new design matrix to get the following output:
Regression Coefficients
Beta_0 1.5
Beta_1 3.5
Beta_2 5.5
The regression coefficients $\beta_{0}$, $\beta_{1}$, and $\beta_{2}$ are now the means for each treatment level, and in the ANOVA table, we see that the $SS_{Error}$ is 1.5. This reduction in the $SS_{Error}$ is the $SS_{treatment}$. Notice that the error SS of the Overall Mean model is the sum of the SS values for Treatment and Error term in this model, which means that by not including the treatment effect in that model, its error SS has been unduly inflated.
Adding the optional code given in Section 4.2 to compute additional Internal computations, we can obtain:
xprimex
2 0 0
0 2 0
0 0 2
check
1 0 0
0 1 0
0 0 1
xprimey
3
7
11
SumY2
89.5
CF
73.5
xprimexinv
0.5 0 0
0 0.5 0
0 0 0.5
Here we can see that $\mathbf{X}' \mathbf{X}$ now contains diagonal elements that are the $n_{i}$ = number of observations for each treatment level mean being computed. In addition, we can verify that $CF = \sum Y^{2} - CF=16$, or the working formula equals the treatment $SS$.
We can now test for the significance of the treatment by using the General Linear $F$ test: $F = \frac{SSE_{reduced} - SSE_{full} / dfE_{reduced} - dfE_{full}}{SSE_{full} / dfE_{full}}$
The Overall Mean model is the "Reduced" model, and the Cell Means model is the "Full" model. From the ANOVA tables, we get: $F = \frac{17.5 - 1.5/5 - 3}{1.5/3} = 16$ which can be compared to $F_{.05,2,3} = 9.55$.
Using R
Steps in R - Cell Means Model
1. Define response variable and design matrix
y<-matrix(c(2,1,3,4,6,5), ncol=1)
x<matrix(c(1,0,0,1,0,0,0,1,0,0,1,0,0,0,1,0,0,1),ncol=3,nrow=6,byrow=TRUE)
2. Regression coefficients
beta<-solve(t(x)%*%x)%*%(t(x)%*%y)
# beta
# [,1]
# [1,] 1.5
# [2,] 3.5
# [3,] 5.5
3. Calculate the entries of the ANOVA Table
n<-nrow(y)
p<-ncol(x)
J<-matrix(1,n,n)
ss_tot = (t(y)%*%y) - (1/n)*(t(y)%*%J)%*%y #17.5
ss_trt = t(beta)%*%(t(x)%*%y) - (1/n)*(t(y)%*%J)%*%y #16
ss_error = ss_tot - ss_trt #1.5
total_df=n-1 #5
trt_df=p-1 #2
error_df=n-p #3
MS_trt = ss_trt/(p-1) #8
MS_error = ss_error / error_df #0.5
F=MS_trt/MS_error #16
4. Creating the ANOVA table
ANOVA <- data.frame(
c ("","Treatment","Error", "Total"),
c("DF", trt_df,error_df,total_df),
c("SS", ss_trt, ss_error, ss_tot),
c("MS", MS_trt, MS_error, ""),
c("F",F,"",""),
stringsAsFactors = FALSE)
names(ANOVA) <- c(" ", " ", " ","","")
5. Print the ANOVA table
print(ANOVA)
# 1 DF SS MS F
# 2 Treatment 2 16 8 16
# 3 Error 3 1.5 0.5
# 4 Total 5 17.5
6. Intermediates in the matrix computations
xprimex<-t(x)%*%x
# xprimex
# [,1] [,2] [,3]
# [1,] 2 0 0
# [2,] 0 2 0
# [3,] 0 0 2
xprimey<-t(x)%*%y
# xprimey
# [,1]
# [1,] 3
# [2,] 7
# [3,] 11
xprimexinv<-solve(t(x)%*%x)
# xprimexinv
# [,1] [,2] [,3]
# [1,] 0.5 0.0 0.0
# [2,] 0.0 0.5 0.0
# [3,] 0.0 0.0 0.5
check<-xprimexinv%*%xprimex
# check
# [,1] [,2] [,3]
# [1,] 1 0 0
# [2,] 0 1 0
# [3,] 0 0 1
SumY2<-t(beta)%*%(t(x)%*%y) #89.5
CF<-(1/n)*(t(y)%*%J)%*%y #73.5 | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.03%3A_Cell_Means_Model.txt |
Model 3 - Dummy Variable Regression
$Y_{ij} = \mu + \mu_{i} + \epsilon_{ij}, \text{ fitted as } Y_{ij} = \beta_{0} + \beta_{Level \ 1} + \beta_{Level \ 2} + \ldots + \beta_{Level \ T-1} + \epsilon_{ij}$
where $\beta_{Level \ 1}, \beta_{Level \ 2}, \ldots, \beta_{Level \ T-1}$ are regression coefficients for $T-1$ indicator-coded regression "dummy" variables that are correspond to the $T-1$ categorical factor levels. The $T^{th}$ factor level mean is given by the regression intercept $\beta_{0}$.
The General Linear Model (GLM) applied to data with categorical predictors can be viewed from a regression modeling perspective as an ordinary multiple linear regression (MLR) with "dummy" coding, also known as indicator coding, for the categorical treatment levels. Typically, software performing the MLR will automatically include an intercept, which corresponds to the first column of the design matrix and is a column of $1$'s. This automatic inclusion of the intercept can lead to complications when interpreting the regression coefficients.
The SAS Mixed procedure, and also the GLM procedure which we may encounter later, use the "Dummy Variable Regression" model. For the $Y$ data used in sections 4.2 and 4.3, the design matrix for this model can be entered into IML as:
/* Dummy Variable Regression Model */
x = {
1 1 0,
1 1 0,
1 0 1,
1 0 1,
1 0 0,
1 0 0};
Notice that in the above design matrix, there are only two indicator columns even though there are three treatment levels in the study. It is because, similar to the matrix below, if we were to have a design matrix with another indicator column representing the third treatment level, the resulting 4 columns would form a set of linearly dependent columns, a mathematical condition that will hinder the computation process any further as explained below. $\begin{bmatrix} 1 & 1 & 0 & 0 \ 1 & 1 & 0 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 1 \end{bmatrix}$
The above matrix containing all 4 columns has the property that the sum of columns 2-4 will equal the first column representing the intercept. As a result, a mathematical condition called singularity is created and the matrix computations will not run. So one of the treatment levels is omitted from the coding in the design matrix above for IML and the eliminated level is called the ‘reference’ level. In SAS, typically, the treatment level with the highest label is defined as the reference level and so, in this study, it is treatment level 3.
Note that the parameter vector for the dummy variable regression model is $\boldsymbol{\beta} = \begin{bmatrix} \mu_{1} \ \mu_{2} \ \mu_{3} \end{bmatrix}$.
Running IML, with the design matrix for the dummy variable regression model, we get the following output;
Regression Coefficients
Beta_0 5.5
Beta_1 -4
Beta_2 -2
The coefficient $\beta_{0}$ is the mean for treatment level 3. The mean for treatment level 1 is then calculated from $\hat{\beta}_{0} + \hat{\beta}_{1} = 1.5$. Likewise, the mean for treatment level 2 is calculated as $\hat{\beta}_{0} + \hat{\beta}_{2} = 3.5$.
Notice that the $F$ statistic calculated from this model is the same as that produced from the Cell Means model.
ANOVA
Treatment df SS MS F
2 16 8 16
Error 3 1.5 0.5
Total 5 17.5
Using Technology
Minitab Example
We can confirm our ANOVA table now by running the analysis in software such as Minitab.
Steps in Minitab
First input the data:
In Minitab, different coding options allow the choice of the design matrix which can be done as follows:
Stat > ANOVA > General Linear Model > Fit General Linear Model and place the variables in the appropriate boxes:
Then select Coding… and choose the (1,0) coding as shown below:
Select OK to exit the nested windows. This produces the regular ANOVA output:
Analysis of Variance
And also the Regression Equation:
Regression Equation
y = 5.500 - 4.000 trt_level1 - 2.000 trt_level2 + 0.0 trt_level3
SAS Example
Steps in SAS
In SAS, the default coding is indicator coding, so when you specify the option
model y=trt / solution;
you get the regression coefficients:
Solution for Fixed Effects
Effect trt Estimate Standard Error DF t Value Pr > |t|
Intercept 5.5000 0.5000 3 11.00 0.0016
trt level1 -4.0000 0.7071 3 -5.66 0.0109
trt level2 -2.0000 0.7071 3 -2.83 0.0663
trt level3 0
And the same ANOVA table:
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
trt 2 16.000000 8.000000 Var(Residual)+Q(trt) MS(Residual) 3 16.00 0.0251
Residual 3 1.500000 0.500000 Var(Residual)
The Intermediate calculations for this model are:
xprimex
6 2 2
2 2 0
2 0 2
check
1 -2.22E-16 0
3.331E-16 1 0
0 0 1
xprimey
21
3
7
SumY2
89.5
CF
73.5
xprimexinv
0.5 -0.5 -0.5
-0.5 1 0.5
-0.5 0.5 1
R Example
Steps in R
1. Define response variable and design matrix
y<-matrix(c(2,1,3,4,6,5), ncol=1)
x = matrix(c(1,1,0,1,1,0,1,0,1,1,0,1,1,0,0,1,0,0),ncol=3,nrow=6,byrow=TRUE)
2. Regression coefficients
beta<-solve(t(x)%*%x)%*%(t(x)%*%y)
# beta
# [,1]
# [1,] 5.5
# [2,] -4.0
# [3,] -2.0
3. Calculate the entries of the ANOVA Table
n<-nrow(y)
p<-ncol(x)
J<-matrix(1,n,n)
ss_tot = (t(y)%*%y) - (1/n)*(t(y)%*%J)%*%y #17.5
ss_trt = t(beta)%*%(t(x)%*%y) - (1/n)*(t(y)%*%J)%*%y #16
ss_error = ss_tot - ss_trt #1.5
total_df=n-1 #5
trt_df=p-1 #2
error_df=n-p #3
MS_trt = ss_trt/(p-1) #8
MS_error = ss_error / error_df #0.5
F=MS_trt/MS_error #16
4. Creating the ANOVA table
ANOVA <- data.frame(
c ("","Treatment","Error", "Total"),
c("DF", trt_df,error_df,total_df),
c("SS", ss_trt, ss_error, ss_tot),
c("MS", MS_trt, MS_error, ""),
c("F",F,"",""),
stringsAsFactors = FALSE)
names(ANOVA) <- c(" ", " ", " ","","")
5. Print the ANOVA table
print(ANOVA)
# 1 DF SS MS F
# 2 Treatment 2 16 8 16
# 3 Error 3 1.5 0.5
# 4 Total 5 17.5
6. Intermediates in the matrix computations
xprimex<-t(x)%*%x
# xprimex
# [,1] [,2] [,3]
# [1,] 6 2 2
# [2,] 2 2 0
# [3,] 2 0 2
xprimey<-t(x)%*%y
# xprimey
# [,1]
# [1,] 21
# [2,] 3
# [3,] 7
xprimexinv<-solve(t(x)%*%x)
# xprimexinv
# [,1] [,2] [,3]
# [1,] 0.5 -0.5 -0.5
# [2,] -0.5 1.0 0.5
# [3,] -0.5 0.5 1.0
check<-xprimexinv%*%xprimex
# check
# [,1] [,2] [,3]
# [1,] 1.000000e+00 0.000000e+00 0
# [2,] -1.110223e-16 1.000000e+00 0
# [3,] 0.000000e+00 -1.110223e-16 1
SumY2<-t(beta)%*%(t(x)%*%y) # 89.5
CF<-(1/n)*(t(y)%*%J)%*%y # 73.5
7. Regression Equation and ANOVA table
trt_level1<-x[,2]
trt_level2<-x[,3]
model<-lm(y~trt_level1+trt_level2)
8. With the command summary(model) we can get the following output:
Call:
lm(formula = y ~ trt_level1 + trt_level2)
Residuals:
1 2 3 4 5 6
0.5 -0.5 -0.5 0.5 0.5 -0.5
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 5.5000 0.5000 11.000 0.00161 **
trt_level1 -4.0000 0.7071 -5.657 0.01094 *
trt_level2 -2.0000 0.7071 -2.828 0.06628 .
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 0.7071 on 3 degrees of freedom
Multiple R-squared: 0.9143, Adjusted R-squared: 0.8571
F-statistic: 16 on 2 and 3 DF, p-value: 0.02509
From the output, we can see the estimates for the coefficients are b0=5.5, b1=-4, b2=-2 and the F-statistic is 16 with a p-value of 0.02509.
By using the estimates we can write the regression equation:
y=5.5-4 trt_level1-2 trt_level2+0 trt_level3
9. With the command anova(model) we can get the following output
Analysis of Variance Table
Response: y
Df Sum Sq Mean Sq F value Pr(>F)
trt_level1 1 12.0 12.0 24 0.01628 *
trt_level2 1 4.0 4.0 8 0.06628 .
Residuals 3 1.5 0.5 ---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Note: R is giving the sequential sum of squares in the ANOVA table. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.04%3A_Dummy_Variable_Regression.txt |
Model 4 - The Effects Model
$Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$ where $\tau_{i}$ are the the deviations of each factor level mean from the overall mean so that $\sum_{i=1}^{T} \tau_{i} = 0$.
In the effects model that we discussed in chapter 3, the treatment means were not estimated but instead, the $\tau_{i}$'s, or the deviations of treatment means from the overall mean, were estimated. The model must include the overall mean, which is estimated by the intercept, and hence the design matrix to be inputted for IML is:
/* The Effects Model */
x={
1 1 0,
1 1 0,
1 0 1,
1 0 1,
1 -1 -1,
1 -1 -1};
Here we have another omission of a treatment level, but for a different reason. In the effects model, we have the constraint $\sum \tau_{i}=0$. As a result, coding for one treatment level can be omitted.
Running the IML program with this design matrix yields:
Regression Coefficients
Beta_0 3.5
Beta_1 -2
Beta_2 0
The regression coefficient Beta_0 is the overall mean and the coefficients $\beta_{1}$ and $\beta_{2}$ are $\tau_{1}$ and $\tau_{2}$, respectively. The estimate for $\tau_{3}$ is obtained as $-(\tau_{1})-(\tau_{2})=2.0$.
In Minitab, if we change the coding now to be Effect coding (-1,0,+1), which is the default setting, we get the following:
Regression Equation
y = 3.500 - 2.000 trt_A - 0.000 trt_B + 2.000 trt_C
The ANOVA table is the same as for the dummy-variable regression model above. We can also observe that the factor level means and General Linear F Statistics values obtained for all 3 representations (cell means, dummy coded regression and effects coded regression) are identical, confirming that the 3 representations are identical.
The intermediates were:
xprimex
6 0 0
0 4 2
0 2 4
check
1 0 0
0 1 0
0 0 1
xprimey
21
-8
-4
SumY2
89.5
CF
73.5
xprimexinv
0.1666667 0 0
0 0.3333333 -0.166667
0 -0.166667 0.3333333
By coding treatment or factor levels into numerical terms, we can use regression methods to perform the ANOVA.
To state the effects model $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$ as a regression model, we need to include $\mu, \tau_{i}, \ldots, \tau_{T}$ as elements in the parameter vector $\boldsymbol{\beta}$ of the GLM model. Note that, in the case of equal replication at each factor level, the deviations satisfy the following constraint: $\sum_{i=1}^{T} \tau_{i} = 0$
This implies one of the $\tau_{i}$ parameters is not needed since it can be expressed in terms of the other $T-1$ parameters and need not be included in the $\boldsymbol{\beta}$ parameter vector. We shall drop the parameter $\tau_{T}$ from the regression equation, as it can be expressed in terms of the other $T-1$ parameters $\tau_{i}$ as follows: $\tau_{T} = - \tau_{1} - \tau_{2} - \ldots - \tau_{T-1}$
Thus, the $\boldsymbol{\beta}$ vector of the GLM is a $T \times 1$ vector containing only the parameters $\mu, \tau_{1}, \ldots, \tau_{T-1}$ for the linear model.
To illustrate how a linear model is developed with this approach, consider a single-factor study with $T=3$ factor levels when $n_{1}=n_{2}=n_{3}=2$. The $\mathbf{Y}$, $\mathbf{X}$, $\boldsymbol{\beta}$, and $\boldsymbol{\epsilon}$ matrices for this case are as follows: $\mathbf{Y} = \begin{bmatrix} Y_{11} \ Y_{12} \ Y_{21} \ Y_{22} \ Y_{31} \ Y_{32} \end{bmatrix}, \ \mathbf{X} = \begin{bmatrix} 1 & 1 & 0 \ 1 & 1 & 0 \ 1 & 0 & 1 \ 1 & 0 & 1 \ 1 & -1 & -1 \ 1 & -1 & -1 \end{bmatrix}, \ \boldsymbol{\beta} = \begin{bmatrix} \beta_{0} \ \beta_{1} \ \beta_{2} \end{bmatrix}, \ \boldsymbol{\epsilon} = \begin{bmatrix} \epsilon_{11} \ \epsilon_{12} \ \epsilon_{21} \ \epsilon_{22} \ \epsilon_{31} \ \epsilon_{32} \end{bmatrix}$
where $\beta_{0}$, $\beta_{1}$, and $\beta_{2}$ correspond to $\mu$, $\tau_{1}$, and $\tau_{2}$ respectively.
Note that the vector of expected values $\mathbf{E\{Y\}} = \mathbf{X} \boldsymbol{\beta}$ yields the following: \begin{align} \mathbf{E\{Y\}} & = \mathbf{X} \boldsymbol{\beta} \ \begin{bmatrix} E\{Y_{11}\} \ E\{Y_{12}\} \ E\{Y_{21}\} \ E\{Y_{22}\} \ E\{Y_{31}\} \ E\{Y_{32}\} \end{bmatrix} & = \begin{bmatrix} 1 & 1 & 0 \ 1 & 1 & 0 \ 1 & 0 & 1 \ 1 & 0 & 1 \ 1 & -1 & -1 \ 1 & -1 & -1 \end{bmatrix} \begin{bmatrix} \beta_{0} \ \beta_{1} \ \beta_{2} \end{bmatrix} \ &= \begin{bmatrix} \mu + \tau_{1} \ \mu + \tau_{1} \ \mu + \tau_{2} \ \mu + \tau_{2} \ \mu - \tau_{1} - \tau_{2} \ \mu - \tau_{1} - \tau_{2} \end{bmatrix} \end{align}
Since $\tau_{3} = -\tau_{1} - \tau_{2}$, as shown above, we see that $E\{Y_{31}\} = E\{Y_{32}\} = \mu + \tau_{3}$. Thus, the above $\mathbf{X}$ matrix and $\boldsymbol{\beta}$ vector representation provides the appropriate expected values for all factor levels as expressed below: $E\{Y_{ij}\} = \mu + \tau_{i}$
Using R: Effects Model
Steps in R
1. Define response variable and design matrix
y<-matrix(c(2,1,3,4,6,5), ncol=1)
x = matrix(c(1,1,0,1,1,0,1,0,1,1,0,1,1,-1,-1,1,-1,-1),ncol=3,nrow=6,byrow=TRUE)
2. Regression coefficients
beta<-solve(t(x)%*%x)%*%(t(x)%*%y)
# beta
# [,1]
# [1,] 3.5
# [2,] -2.0
# [3,] 0.0
3. Calculate the entries of the ANOVA Table
n<-nrow(y)
p<-ncol(x)
J<-matrix(1,n,n)
ss_tot = (t(y)%*%y) - (1/n)*(t(y)%*%J)%*%y #17.5
ss_trt = t(beta)%*%(t(x)%*%y) - (1/n)*(t(y)%*%J)%*%y #16
ss_error = ss_tot - ss_trt #1.5
total_df=n-1 #5
trt_df=p-1 #2
error_df=n-p #3
MS_trt = ss_trt/(p-1) #8
MS_error = ss_error / error_df #0.5
F=MS_trt/MS_error #16
4. Creating the ANOVA table
ANOVA <- data.frame(
c ("","Treatment","Error", "Total"),
c("DF", trt_df,error_df,total_df),
c("SS", ss_trt, ss_error, ss_tot),
c("MS", MS_trt, MS_error, ""),
c("F",F,"",""),
stringsAsFactors = FALSE)
names(ANOVA) <- c(" ", " ", " ","","")
5. Print the ANOVA table
print(ANOVA)
# 1 DF SS MS F
# 2 Treatment 2 16 8 16
# 3 Error 3 1.5 0.5
# 4 Total 5 17.5
6. Intermediates in the matrix computations
xprimex<-t(x)%*%x
# xprimex
# [,1] [,2] [,3]
# [1,] 6 0 0
# [2,] 0 4 2
# [3,] 0 2 4
xprimey<-t(x)%*%y
# xprimey
# [,1]
# [1,] 21
# [2,] -8
# [3,] -4
xprimexinv<-solve(t(x)%*%x)
# xprimexinv
# [,1] [,2] [,3]
# [1,] 0.1666667 0.0000000 0.0000000
# [2,] 0.0000000 0.3333333 -0.1666667
# [3,] 0.0000000 -0.1666667 0.3333333
check<-xprimexinv%*%xprimex
# check
# [,1] [,2] [,3]
# [1,] 1 0 0
# [2,] 0 1 0
# [3,] 0 0 1
SumY2<-t(beta)%*%(t(x)%*%y) #89.5
CF<-(1/n)*(t(y)%*%J)%*%y # 73.5
7. Regression Equation and ANOVA table
trt_level1<-x[,2]
trt_level2<-x[,3]
model<-lm(y~trt_level1+trt_level2)
8. With the command summary(model) we can get the following output:
Call:
lm(formula = y ~ trt_level1 + trt_level2)
Residuals:
1 2 3 4 5 6
0.5 -0.5 -0.5 0.5 0.5 -0.5
Coefficients:
Estimate Std. Error t value Pr(>|t|)
(Intercept) 3.500e+00 2.887e-01 12.124 0.00121 **
trt_level1 -2.000e+00 4.082e-01 -4.899 0.01628 *
trt_level2 -1.282e-16 4.082e-01 0.000 1.00000
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Residual standard error: 0.7071 on 3 degrees of freedom
Multiple R-squared: 0.9143, Adjusted R-squared: 0.8571
F-statistic: 16 on 2 and 3 DF, p-value: 0.02509
From the output we can see the estimates for the coefficients are b0=3.5, b1=-2, b2=0 and the F-statistic is 16 with a p-value of 0.02509.
By using the estimates we can write the regression equation:
y=3.5-2 trt_level1-0 trt_level2+2 trt_level3
The estimator $\tau_{3}$ is obtained as $-\tau_{1} - \tau_{2} = 2$
9. With the command anova(model) we can get the following output:
Analysis of Variance Table
Response: y
Df Sum Sq Mean Sq F value Pr(>F)
trt_level1 1 16.0 16.0 32 0.01094 *
trt_level2 1 0.0 0.0 0 1.00000
Residuals 3 1.5 0.5
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
Note that R is giving the sequential sum of squares in the ANOVA table. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.05%3A_Computational_Aspects_of_the_Effects_Model.txt |
In Section 1.1 we encountered a brief description of an experiment. The description of an experiment provides a context for understanding how to build an appropriate statistical model. All too often, mistakes are made in statistical analyses because of a lack of understanding of the setting and procedures in which a designed experiment is conducted. Creating a study diagram is one of the best ways to address this, in addition to being intuitive. A study diagram is a schematic diagram that captures the essential features of the experimental design. Here, as we explore the computations for a single factor ANOVA in a simple experimental setting, the study diagram may seem trivial. However, in practice and in lessons to follow in this course, the ability to create accurate study diagrams usually makes a substantial difference in getting the model right.
In our example, as described in Section 1.1, a plant biologist thinks that plant height may be affected by the fertilizer type and three types of fertilizer were chosen to investigate this claim. Next, 24 plants were randomly chosen and 4 batches, with 6 plants in each, were assigned individually to the 3 fertilizer types; the last batch was left untreated, constituting the Control group. The researchers kept all the plants under controlled conditions in the greenhouse. The individual containerized plants were randomly assigned the fertilizer treatment levels to produce 6 replications of each of the fertilizer applications.
Here is the data from the example that we were using in this lesson:
Control F1 F2 F3
21 32 22.5 28
19.5 30.5 26 27.5
22.5 25 28 31
21.5 27.5 27 29.5
20.5 28 26.5 30
21 28.6 25.2 29.2
So we have a description of the treatment levels and how they were assigned to individual experimental units (the potted plant), and we see the data organized in a table. But what are we missing? A key question is: how was the experiment conducted? This question is a practical one and is answered with a study diagram. These are usually hand-drawn depictions of a real setting, indicating the treatments, levels of treatments, and how the experiment was laid out. They are not typically works of art and no one should ever feel embarrassed by a lack of artistic ability to draw one. For this example, we need to draw a greenhouse bench, capable of holding the 4 × 6 = 24 experimental units:
The diagram identified the response variable, listed the treatment levels, and indicated the random assignment of treatment levels to these 24 experimental units on the greenhouse bench.
This randomization and the subsequent experimental layout we would identify as a Completely Randomized Design (CRD). We know from this schematic diagram that we need a statistical model that is appropriate for a one-way ANOVA in a Completely Randomized Design (CRD).
Furthermore, once the plant heights are recorded at the end of the study, the experimenter may observe that the variability in the growth may possibly be influenced by a second factor besides the fertilizer level. A careful examination of the layout of the plants in the study diagram may perhaps reveal this additional factor. For example, if the growth is higher in the plants placed on the row nearer to the windows, it is reasonable to assume that sunlight also plays a role and to redesign the experiment as a randomized completely block design (RCBD) with rows as a blocking factor. Note that design aspects of experiments are covered in Chapters 7 and 8.
Being able to draw and reproduce a study diagram is very useful in identifying the components of the ANOVA models. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.06%3A_The_Study_Diagram.txt |
Exercise $1$: Design Matrix
Below is a design matrix for a data set of a recent study.
$\begin{bmatrix} 1 & 1 & 0 & 0 \ 1 & 1 & 0 & 0 \ 1 & 1 & 0 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 1 \ 1 & -1 & -1 & -1 \ 1 & -1 & -1 & -1 \ 1 & -1 & -1 & -1 \end{bmatrix} \nonumber$
a) Identify the number of treatment levels and replicates.
Solution
4 treatment levels and 3 replicates
b) Name the model and write its equation.
Solution
This design matrix corresponds to the effects model, and the model equation is $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$, where $i=1,2,3,4$, $j=1,2,3$, and $\sum_{i=1}^{4} \tau_{i} = 0$.
c) Write the equation and the design matrix that corresponds to the cell means model.
Solution
The equation for the cell means model is: $Y_{ij} = \mu + \epsilon_{ij}$, where $i=1,2,3,4$ and $j=1,2,3$. The design matrix corresponding to the cell means model is: $\begin{bmatrix} 1 & 0 & 0 & 0 \ 1 & 0 & 0 & 0 \ 1 & 0 & 0 & 0 \ 0 & 1 & 0 & 0 \ 0 & 1 & 0 & 0 \ 0 & 1 & 0 & 0 \ 0 & 0 & 1 & 0 \ 0 & 0 & 1 & 0 \ 0 & 0 & 1 & 0 \ 0 & 0 & 0 & 1 \ 0 & 0 & 0 & 1 \ 0 & 0 & 0 & 1 \end{bmatrix} \nonumber$
d) Write the equation and the design matrix that corresponds to the dummy variable regressions model.
Solution
The equation for the 'dummy variable regression' model is: $Y_{ij} = \mu + \mu_{i} + \epsilon_{ij}$ for $i=1,2,3$ and $j=1,2,3$. $Y_{4j} = \mu + \epsilon_{4j}$
The design matrix is given below. Note that the last 3 rows correspond to the 4th treatment level which is the reference category and its effect is estimated by the model intercept.
$\begin{bmatrix} 1 & 1 & 0 & 0 \ 1 & 1 & 0 & 0 \ 1 & 1 & 0 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 1 & 0 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 1 \ 1 & 0 & 0 & 0 \ 1 & 0 & 0 & 0 \ 1 & 0 & 0 & 0 \end{bmatrix} \nonumber$
4.08: Chapter 4 Summary
This chapter, together with Chapter 3, covered four different versions of single-factor ANOVA models. They are: Overall Mean, Cell Means, Dummy Variable Regression, and Effects Coded Regression models. This lesson also provided the coding compatible with the SAS IML procedure, which facilitates the ANOVA computations using Matrix Algebra in a GLM setting. The method of least squares was used to estimate model parameters yielding a prediction equation for the response in terms of the treatment level. This prediction tool will show to be more useful in ANCOVA settings where model predictors are both categorical and numerical (more details on ANCOVA in Chapters 9 and 10). The prediction process can be utilized effectively only with a sound knowledge of the parameterization process for each ANOVA model, which we have been able to acquire as the design matrix was an input resource for running the IML code and the knowledge of the parameter vector was useful in interpreting the prediction (regression) equations.
Finally, using the greenhouse example, the concept of a study diagram was discussed. Though a simple visual tool, a study diagram may play an important role in identifying new predictors so that perhaps a pre-determined ANOVA model can be extended to include additional factors to create a multi-factor model discussed in Chapters 5 and 6. In addition to identifying the treatment design, the study diagram also helps in choosing an appropriate randomization design, a topic discussed in Chapters 7 and 8. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/04%3A_ANOVA_Models_Part_II/4.07%3A_Try_It.txt |
Learning Objectives
Upon completion of this lesson, you should be able to:
1. Identify factorial, nested, and cross-nested treatment designs.
2. Use main effects and interaction effects in factorial designs.
3. Create nested designs and identify the nesting effects.
4. Use statistical software to analyze data from different treatment designs via ANOVA and mean comparison procedures.
Researchers often identify more than one experimental factor of interest. One alternative is to set up separate, independent experiments in which a single treatment (or factor) is used in each experiment, and data from each experiment to be analyzed as we have done using a one-way ANOVA. This approach might have the advantage of a concentrated focus on the single treatment of interest and the simplicity of computations. However, there are several disadvantages as well.
• First, environmental factors or experimental material conditions may change during the process. This could distort the assessment of the relative importance of different treatments on the response variable.
• Second, it is inefficient. Setting up and running multiple separate experiments usually will involve more work and resources.
• Last, and probably the most important, this one-at-a-time approach does not allow the examination of how several treatments jointly impact the response.
ANOVA methodology can be extended to accommodate this multi-factor setting. Here are Dr. Rosenberger and Dr. Shumway talking about some of the things to look out for as you work your way through this lesson.
Video \(1\): Experimental design drives analysis.
To put it into perspective, let’s take a look at the phrase "Experimental Design" a term that you often hear. We are going to take this colloquial phrase and divide it into two formal components:
1. The Treatment Design
2. The Randomization Design
We will use the treatment design component to address the nature of the experimental factors under study and the randomization design component to address how treatments are assigned to experimental units. An experimental unit is defined to be that which receives a specific treatment level and in a multi-factor setting, a specific treatment or factor combination. In the single-factor greenhouse example, which is an experiment, the experimental unit is a single plant receiving one specific fertilizer level. Note that the ANOVA model pertaining to a given study depends on both the treatment design and the randomization process.
The following figure illustrates the conceptual division between the treatment design and the randomization design. The terms that are in boldface type will be addressed in detail in this or future lessons.
05: Multi-Factor ANOVA
In multi-factor experiments, combinations of factor levels are applied to experimental units. The single-factor greenhouse experiment discussed in previous lessons can be extended to a multi-factor study by including plant species as an additional factor along with fertilizer type. This addition of another factor may prove to be useful, as one fertilizer type may be most effective on one specific plant species! In other words, the optimal height growth is perhaps attainable by a unique combination of fertilizer type and plant species. A treatment design that provides the opportunity to determine this best combination is a factorial design, where responses are observed at each level of a given factor combined with each level of all other factors. In this setting, factors are said to be crossed.
A factorial design with $t$ factors is identified using the $l_{1} l_{2} \ldots l_{t}$ notation, where $l_{i}$ is the number of levels of factor $i$ $(i=1,2, \ldots,t)$. For example, a factorial design with 2 factors A and B, where A has 4 levels and B has 3 levels, will have the $4 \times 3$ notation.
One complete replication of a factorial design with $t$ factors requires $(l_{1} \times l_{2} \times \ldots \times l_{t})$ experimental units, and this quantity is called the replicate size. If $r$ is the number of complete replicates, then $N$, the total number of observations, equals $r \times (l_{1} \times l_{2} \times \ldots \times l_{t})$.
It is easy to see that with the addition of more and more crossed factors, the replicate size will increase rapidly and design modifications have to be made to make the experiment more manageable.
In a factorial experiment, as combinations of different factor levels play an important role, it is important to differentiate between the lone (or main) effects of a factor on the response and the combined effects of a group of factors on the response.
The main effect of factor A is the effect of A on the response ignoring the effect of all other factors. The main effect of a given factor is equivalent to the factor effect associated with the single-factor experiment using only that particular factor.
The combined effect of a specific combination of $l$ different factors is called the interaction effect (more details later). The interaction effect of most interest is the two-way interaction effect and is denoted by the product of the two letters assigned to the two factors. For example, the two-way interaction effects of a factorial design with 3 factors A, B, C are denoted AB, AC, and BC. Likewise, the three-way interaction effect of these 3 factors is denoted by ABC.
Let us now examine how the degrees of freedom ($df$) values of a single-factor ANOVA can be extended to the ANOVA of a two-factor factorial design. Note that the interaction effects are additional terms that need to be included in a multi-factor ANOVA, but the ANOVA rules studied in Chapter 2 for single-factor situations still apply for the main effect of each factor. If the two factors of the design are denoted by A and B with $a$ and $b$ as their number of levels respectively, then the $df$ values of the two main effects are $(a-1)$ and $(b-1)$. The $df$ value for the two-way interaction effect is $(a-1)(b-1)$, the product of $df$ values for A and B. The ANOVA table below gives the layout of the df values for a $2 \times 2$ factorial design with 5 complete replications. Note that in this experiment, $r$ equals 5, and $N$ is equal to 20.
Source d.f.
Factor A $(a - 1) = 1$
Factor B $(b - 1) = 1$
Factor A × Factor B $(a - 1)(b - 1) = 1$
Error $19 - 3 = 16$
Total $N-1=(nab)-1=19$
If in the single-factor model of $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$ $\tau_{i}$ is effectively replaced with $\alpha_{i} + \beta_{i} + (\alpha \beta)_{ij}$, then the resulting equation shown below will represent the model equation of a two-factor factorial design.
$Y_{ijk} = \mu + \alpha_{i} + \beta_{j} + (\alpha \beta)_{ij} + \epsilon_{ijk}$ where $\alpha_{i}$ is the main effect of factor A, $\beta_{j}$ is the main effect of factor B, and $(\alpha \beta)_{ij}$ is the interaction effect $(i=1,2,\ldots,a, \ j=1,2,\ldots,b, \ k=1,2,\ldots,r)$.
This reflects the following partitioning of treatment deviations from the grand mean: $\underbrace{ \bar{Y}_{ij.} - \bar{Y}_{...} }_{\begin{array}{c} \text{Deviation of estimated treatment mean} \ \text{around overall mean} \end{array}} = \underbrace{ \bar{Y}_{i..} - \bar{Y}_{...} }_{A \text{ main effect}} + \underbrace{ \bar{Y}_{.j.}-\bar{Y}_{...} }_{B \text{ main effect}} + \underbrace{ \bar{Y}_{ij.} - \bar{Y}_{i..} - \bar{Y}_{.j.} + \bar{Y}_{...} }_{AB \text{ interaction effect}}$
The main effects for Factor A and Factor B are straightforward to interpret, but what is an interaction? Delving more, an interaction can be defined as the failure of the response to one factor to be the same at different levels of another factor. Notice that $(\alpha \beta)_{ij}$, the interaction term in the model, is multiplicative, and as a result may have a large and important impact on the response variable. Interactions go by different names in various fields. In medicine, for example, physicians most times ask what medication you are on before prescribing a new medication. They do this out of a concern for interaction effects of either interference (a canceling effect) or synergism (a compounding effect).
Graphically, in a two-factor factorial with each factor having 2 levels, the interaction can be represented by two non-parallel lines connecting means (adapted from Zar, H. Biostatistical Analysis, 5th Ed., 1999). It is because the interaction reflects the failure of the difference in response between the two different levels of one factor to be the same, for both levels of the other factor. So, if there is no interaction, then this difference in response will be the same, which will graphically result in two parallel lines. In the interaction plots below, parallel lines are a consistent feature in all settings with no interaction. In plots depicting interaction, the lines do cross (or would cross if the lines kept going).
In graph 1 there is no effect of Factor A, a small effect of Factor B (and if there were no effect of Factor B the two lines would coincide), and no interaction between Factor A and Factor B.
Graph 2 shows a large effect of Factor A, small effect of Factor B, and no interaction.
Graph 3 shows no effect of Factor A, larger effect of Factor B, and no interaction.
In graph 4 there is a large effect of Factor A, a large effect of Factor B , and no interaction.
In graph 5 there is no effect of Factor A and no effect of Factor B, but an interaction between A and B.
In graph 6 there is a large effect of Factor A and no effect of Factor B, with a slight interaction between A and B.
In graph 7 there is no effect of Factor A and a large effect of Factor B, with a very large interaction.
In graph 8 there is a small effect of Factor A and a large effect of Factor B, with a large interaction.
In the presence of multiple factors with their interactions, multiple hypotheses can be tested and for a two-factor factorial design. They are:
Main Effect of Factor A:
$\begin{array}{l} H_{0}: \ \alpha_{1} = \alpha_{2} = \ldots = \alpha_{a} = 0 \ H_{A}: \ \text{not all } \alpha_{i} \text{ are equal to 0} \end{array}$
Main Effect of Factor B:
$\begin{array}{l} H_{0}: \ \beta_{1} = \beta_{2} = \ldots = \beta_{b} = 0 \ H_{A}: \ \text{not all } \beta_{j} \text{ are equal to 0} \end{array}$
A × B Interaction:
$\begin{array}{c} H_{0} \ \text{there is no interaction} \ H_{A}: \ \text{an interaction exists} \end{array}$
When testing these hypotheses, it is important to test for the significance of the interaction effect first. If the interaction is significant, the main effects are of no consequence; rather, the differences among different factor level combinations should be looked into. The greenhouse example, extended to include a second (crossed) factor, will illustrate the steps.
5.01: Factorial or Crossed Treatment Designs
Let's return to the greenhouse example with plant species also as a predictive factor, in addition to fertilizer type. The study then becomes a 2×4 factorial as 2 types of plant species and 4 types of fertilizers are investigated. The total number of experimental units (plants) that are needed now is 48, as r=6 and there are 8 plant species and fertilizer type combinations.
The data might look like this:
The ANOVA table would now be constructed as follows:
The data presented in the table above are in unstacked format. One needs to convert this into a stacked format when attempting to use statistical software. The SAS code is as follows.
The data presented in the table above are in unstacked format. One needs to convert this into a stacked format when attempting to use statistical software. The SAS code is as follows.
```data greenhouse_2way;
input fert \$ species \$ height;
datalines;
control SppA 21.0
control SppA 19.5
control SppA 22.5
control SppA 21.5
control SppA 20.5
control SppA 21.0
control SppB 23.7
control SppB 23.8
control SppB 23.8
control SppB 23.7
control SppB 22.8
control SppB 24.4
f1 SppA 32.0
f1 SppA 30.5
f1 SppA 25.0
f1 SppA 27.5
f1 SppA 28.0
f1 SppA 28.6
f1 SppB 30.1
f1 SppB 28.9
f1 SppB 30.9
f1 SppB 34.4
f1 SppB 32.7
f1 SppB 32.7
f2 SppA 22.5
f2 SppA 26.0
f2 SppA 28.0
f2 SppA 27.0
f2 SppA 26.5
f2 SppA 25.2
f2 SppB 30.6
f2 SppB 31.1
f2 SppB 28.1
f2 SppB 34.9
f2 SppB 30.1
f2 SppB 25.5
f3 SppA 28.0
f3 SppA 27.5
f3 SppA 31.0
f3 SppA 29.5
f3 SppA 30.0
f3 SppA 29.2
f3 SppB 36.1
f3 SppB 36.6
f3 SppB 38.7
f3 SppB 37.1
f3 SppB 36.8
f3 SppB 37.1
;
run;
/*The code to generate the boxplot
for distribution of height by species organized by fertilizer
in Figure 5.1*/
proc sort data=greenhouse_2way; by fert species;
proc boxplot data=greenhouse_2way;
plot height*species (fert);
run;
```
As a preliminary step in Exploratory Data Analysis (EDA), a side-by-side boxplot display of height vs. species organized by fertilizer type would be an ideal graphic. As the plot shows, the height differences between species are variable among fertilizer types (see for example the difference in height between SppA and SppB for Control is much less than that for F3). This indicates that fert*species could be a significant interaction prompting a factorial model with interaction.
To run the two-factor factorial model with interaction in SAS `proc mixed`, we can use:
```/*Runs the two-factor factorial model with interaction*/
proc mixed data=greenhouse_2way method=type3;
class fert species;
model height = fert species fert*species;
store out2way;
run;
```
In the `proc mixed` procedure, similar to when running the single factor ANOVA. The name of the data set is specified in the `proc mixed` statement and so is the `method=type 3` option that specifies the way the F test is calculated. The `fert` and `species` factors that are both categorical are included in the class statement. The terms (or effects) in the model statement are consistent with the source effects in the layout of the "theoretical" ANOVA table illustrated in 5.1. Finally, the `store` command stores the elements necessary for the generation of the LS-Means interval plot.
Recall the two ANOVA rules, applicable to any model: (a). the df values add up to total df and (b). the sums of squares add up to total sums of squares. As seen by the output below, the df values and also the sums of squares follow these rules. (It is easy to confirm that the total sum of squares = 1168.732500, by the 2nd ANOVA rule.)
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
fert 3 745.437500 248.479167 Var(Residual)+Q(fert,fert*species) MS(Residual) 40 73.10 <.0001
species 1 236.740833 236.740833 Var(Residual)+Q(species,fert*species) MS(Residual) 40 69.65 <.0001
fert*species 3 50.584167 16.861389 Var(Residual)+Q(fert*species) MS(Residual) 40 4.96 0.0051
Residual 40 135.970000 3.399250 Var(Residual)
Rule
In a model with the interaction effect, the interaction term should be interpreted first. If the interaction effect is significant, then do NOT interpret the main effects individually. Instead, compare the mean response differences among the different factor level combinations.
In general, a significant interaction effect indicates that the impact of the levels of Factor A on the response depends upon the level of Factor B and vice versa. In other words, in the presence of a significant interaction, a stand-alone main effect is of no consequence. In the case where an interaction is not significant, the interaction term can be dropped and a model without the interaction should be run. See Section 5.1.1a: The Additive Model (No Interaction)).
Now applying the above rule for this example, the small p-value of 0.0051 displayed in the table above indicates that the interaction effect is significant, which means that the main effects of either fert or species should not be considered individually. It is the average response differences among the fert and species combinations that matter. In order to determine the statistically significant fert and species combinations, a suitable multiple comparison procedure, such as Tukey and Kramer procedure can be performed on the LS-Means of the interaction effect (i.e.: the treatment combinations).
The necessary follow-up SAS code to perform this procedure is given below.
```ods graphics on;
proc plm restore=out2way;
lsmeans fert*species / adjust=tukey plot=(diffplot(center) meanplot(cl ascending)) cl lines;
/* Because the 2-factor interaction is significant, we work with
the means for treatment combination*/
run;
```
SAS Output for the LSmeans:
fert*species Least Squares Means
fert species Estimate Standard Error DF t Value Pr > |t| Alpha Lower Upper
control SppA 21.0000 0.7527 40 27.90 <.0001 0.05 19.4788 22.5212
control SppB 32.7000 0.7527 40 31.49 <.0001 0.05 22.1788 25.2212
f1 SppA 28.6000 0.7527 40 38.00 <.0001 0.05 27.0788 30.1212
f1 SppB 31.6167 0.7527 40 42.00 <.0001 0.05 30.0954 33.1379
f2 SppA 25.8667 0.7527 40 34.37 <.0001 0.05 24.3454 27.3879
f2 SppB 30.0500 0.7527 40 39.92 <.0001 0.05 28.5288 31.5712
f3 SppA 29.2000 0.7527 40 38.79 <.0001 0.05 27.6788 30.7212
f3 SppB 37.0667 0.7527 40 49.25 <.0001 0.05 35.5454 38.5879
Note that the \(p\)-values here (Pr > t) are testing the hypotheses that the fert and species combination means = 0. This may be of very little interest. However, a comparison of mean response values for different species and fertilizer combinations may prove to be more beneficial and can be derived from the diffogram shown in Figure \(2\). Again recall that, if the confidence interval does not contain zero, then the difference between the two associated means is statistically significant.
Notice also that we see a single value for the standard error based on the MSE from the ANOVA, rather than a separate standard error for each mean (as we would get from Proc Summary for the sample means). Again in this example, with equal sample sizes and no covariates, the lsmeans will be identical to the ordinary means displayed in the Summary Procedure.
There are total of 8 fert*species combinations resulting a total of \(\tbinom{8}{2} = 28\) pairwise comparisons. From the diffogram for differences in fert*species combinations, we see that 10 of them are not significant and 18 of them are significant at a 5% level after Tukey adjustment (more about diffograms). The information used to generate the diffogram is presented in the table for differences of fert*species least squares means in the SAS output (this table is not displayed here).
We can save the differences estimated in SAS `proc mixed` and utilize `proc sgplot` to create the plot of differences in mean response for the fert*species combinations as shown in Figure \(3\). The CIs shown are the Tukey adjusted CIs. SAS code to produce Figure \(3\) is not given in these notes. The interpretations of the plot are similar to what we observed from the diffogram in Figure \(2\).
In addition to comparing differences in mean responses for the fert*species combinations, the SAS code shared above will also produce the line plot for multiple comparisons of means for fert*species combinations (shown in Figure \(4\)) and the plot of means responses organized in the ascending order with 95% CIs for fert*species combinations (shown in Figure \(5\)).
The line plot in Figure \(4\) connects groups in which the LS-means are not statistically different and displays a summary of which groups have similar means. The plot of means with 95% CIs in Figure \(5\) illustrates the same result, although it uses unadjusted CIs. We have organized the plot in the ascending order of estimated means to make it easy to draw conclusions.
Using LSMEANS, subsequent to performing an ANOVA will help to identify the significantly different treatment level combinations. In other words, the ANOVA doesn't end with a \(p\)-value for an \(F\)-test. A small \(p\)-value signals the need for a mean comparison procedure.
5.1.01: Two-Factor Factorial - Greenhouse Example (SAS)
In a factorial design, we first look at the interactions for significance. In the case where interaction is not significant, then we can drop the interaction term from our model, and we end up with an additive model.
For a two-factor factorial, the model we initially consider (as we have discussed in Section 5.1) is: $Y_{ij} = \mu_{..} + \alpha_{i} + \beta_{j} + (\alpha \beta)_{ij} + \epsilon_{ijk}$
Note that the interaction term, $(\alpha \beta)_{ij}$, is a multiplicative term.
If the interaction is found to be non-significant, then the model reduces to: $Y_{ij} = \mu_{..} + \alpha_{i} + \beta_{j} + \epsilon_{ijk}$ Here we can see that the response variable is simply a function of adding the effects of the two factors.
Example $1$: Glucose in Blood Serum
As an example, (adapted from Kuehl, 2000), let's look at a study designed to evaluate two chemical methods used for assaying the amount of glucose in blood serum. A large volume of blood serum served as a starting point for the experiment. The blood serum was divided into three portions, each of which was 'doped' or augmented by adding an additional amount of glucose. Three doping levels were used. Samples of the doped serum were then assayed for glucose concentration by one of two chemical methods. This type of ‘doping’ experiment is commonly used to compare the sensitivity of assay methods.
The amount of glucose detected in each sample was recorded and is presented in the table below.
Solution
The model was run as a two-factor factorial and produced the following results:
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
method 1 263.733889 263.733889 Var(Residual) + Q(method, method*doping) MS(Residual) 12 98.35 <.0001
doping 2 57026 28513 Var(Residual) + Q(doping, method*doping) MS(Residual) 12 10632.5 <.0001
method*doping 2 13.821111 6.910556 Var(Residual) + Q(method*doping) MS(Residual) 12 2.58 0.1172
Residual 12 32.180000 2.681667 Var(Residual)
Here we can see that the interaction of method*doping was not significant (p-value > 0.05) at a 5% level. We drop the interaction effect from the model and run the additive model. The resulting ANOVA table is:
The Mixed Procedure
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
method 1 263.733889 263.733889 Var(Residual)+Q(method, method) MS(Residual) 14 80.26 <.0001
doping 2 57026 28513 Var(Residual) + Q(doping,doping) MS(Residual) 14 8677.63 <.0001
1Residual 14 46.001111 3.285794 Var(Residual)
The Error SS is now 46.001, which is the sum of the interaction SS and the error SS of the model with the interaction. The df values were also added the same way. This example shows that any term not included in the model gets added into the error term, which may erroneously inflate the error especially if the impact of excluded term on the response is not negligible.
The Error SS is now 46.001, which is the sum of the interaction SS and the error SS of the model with the interaction. The df values were also added the same way. This example shows that any term not included in the model gets added into the error term, which may erroneously inflate the error especially if the impact of excluded term on the response is not negligible.
method Least Squares Means
method Estimate Standard Error DF t Value Pr >|t| Alpha Lower Upper
1 123.40 0.6042 14 204.23 <.0001 0.05 122.10 124.70
2 115.74 0.6042 14 191.56 <.0001 0.05 114.45 117.04
doping Least Squares Means
Doping Estimate Standard Error DF t Value Pr >|t| Alpha Lower Upper
1 43.67 0.7400 14 59.01 <.0001 0.05 42.08 45.25
2 136.77 0.7400 14 184.81 <.0001 0.05 135.18 138.35
3 178.28 0.7400 14 240.92 <.0001 0.05 176.70 179.87
Here, we can see that the response variable, the amount of glucose detected in a sample, is the overall mean PLUS the effect of the method used PLUS the effect of the glucose amount added to the original sample. (Hence, the additive nature of this model!) | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/05%3A_Multi-Factor_ANOVA/5.01%3A_Factorial_or_Crossed_Treatment_Designs/5.1.01%3A_Two-Factor_Factorial_-_Greenhouse_Example_%28SAS%29/5.1.1a%3A_The_Additi.txt |
For Minitab, we also need to convert the data to a stacked format (Lesson 4 2 way Stacked Dataset). Once we do this, we will need to use a different set of commands to generate the ANOVA. We use...
Stat > ANOVA > General Linear Model > Fit General Linear Model
and get the following dialog box:
Click on Model…, hold down the shift key and highlight both factors. Then click on the Add box to add the interaction to the model.
These commands will produce the ANOVA results below which are similar to the output generated by SAS (shown in the previous section).
Analysis of Variance
Following the ANOVA run, you can generate the mean comparisons by
Stat > ANOVA > General Linear Model > Comparisons
Then specify the fert*species interaction term for the comparisons by checking the box.
Then choose Graphs to get the following dialog box, where "Interval plot for difference of means" should be checked.
The outputs are shown below.
Grouping Information Using the Tukey Method and 95% Confidence
Means that do not share a letter are significantly different.
5.1.3a: The Additive
• Load the greenhouse data.
• Produce a boxplot to plot the differences in heights for each species organized by fertilizer.
• Produce a “means plot” (interval plot) to view the differences in heights for each species organized by fertilizer.
• Obtain the ANOVA table with interaction.
• Obtain Tukey’s multiple comparisons CIs, grouping, and plot.
1. Load the greenhouse data by using the following commands:
```setwd("~/path-to-folder/")
greenhouse_2way_data <-read.table("greenhouse_2way_data.txt",header=T)
attach(greenhouse_2way_data)
```
2. Produce the Boxplot by using the following commands:
```library("ggpubr")
boxplot(height ~ species*fertilizer, data = greenhouse_2way_data,
xlab = "Species", ylab = "Plant Height",
main="Distribution of Plant Height by Species",
frame = TRUE)
```
3. Produce the means plot (interval plot) by using the following commands:
```library("gplots")
plotmeans(height ~ interaction(species,fertilizer), data = greenhouse_2way_data,connect=FALSE,n.label=FALSE,
xlab = "Fertilizer*species", ylab = "Plant Height",
main="Means Plot with 95% CI")
```
4. Obtain the ANOVA table with interaction by using the following commands:
```anova<-aov(height~fertilizer+species+fertilizer*species,greenhouse_2way_data)
summary(anova)
# Df Sum Sq Mean Sq F value Pr(>F)
# fertilizer 3 745.4 248.48 73.10 2.77e-16 ***
# species 1 236.7 236.74 69.64 2.71e-10 ***
# fertilizer:species 3 50.6 16.86 4.96 0.00508 **
# Residuals 40 136.0 3.40
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
```
5. Obtain Tukey multiple comparisons of means with 95% family-wise confidence level by using the following commands:
```library(multcomp)
library(multcompView)
tukey_multiple_comparisons<-TukeyHSD(anova,conf.level=0.95,ordered=TRUE)
tukey_multiple_comparisons
Tukey multiple comparisons of means
95% family-wise confidence level
factor levels have been ordered
Fit: aov(formula = height ~ fertilizer + species + fertilizer * species, data = greenhouse_2way_data)
\$fertilizer
diff lwr upr p adj
f2-control 5.608333 3.5908095 7.625857 0.0000000
f1-control 7.758333 5.7408095 9.775857 0.0000000
f3-control 10.783333 8.7658095 12.800857 0.0000000
f1-f2 2.150000 0.1324762 4.167524 0.0328745
f3-f2 5.175000 3.1574762 7.192524 0.0000002
f3-f1 3.025000 1.0074762 5.042524 0.0013828
\$species
diff lwr upr p adj
SppB-SppA 4.441667 3.365986 5.517348 0
\$`fertilizer:species`
diff lwr upr p adj
control:SppB-control:SppA 2.700000 -0.7025601 6.102560 0.2100548
f2:SppA-control:SppA 4.866667 1.4641065 8.269227 0.0010962
f1:SppA-control:SppA 7.600000 4.1974399 11.002560 0.0000003
f3:SppA-control:SppA 8.200000 4.7974399 11.602560 0.0000001
f2:SppB-control:SppA 9.050000 5.6474399 12.452560 0.0000000
f1:SppB-control:SppA 10.616667 7.2141065 14.019227 0.0000000
f3:SppB-control:SppA 16.066667 12.6641065 19.469227 0.0000000
f2:SppA-control:SppB 2.166667 -1.2358935 5.569227 0.4721837
f1:SppA-control:SppB 4.900000 1.4974399 8.302560 0.0009970
f3:SppA-control:SppB 5.500000 2.0974399 8.902560 0.0001745
f2:SppB-control:SppB 6.350000 2.9474399 9.752560 0.0000138
f1:SppB-control:SppB 7.916667 4.5141065 11.319227 0.0000001
f3:SppB-control:SppB 13.366667 9.9641065 16.769227 0.0000000
f1:SppA-f2:SppA 2.733333 -0.6692268 6.135893 0.1979193
f3:SppA-f2:SppA 3.333333 -0.0692268 6.735893 0.0584747
f2:SppB-f2:SppA 4.183333 0.7807732 7.585893 0.0072041
f1:SppB-f2:SppA 5.750000 2.3474399 9.152560 0.0000832
f3:SppB-f2:SppA 11.200000 7.7974399 14.602560 0.0000000
f3:SppA-f1:SppA 0.600000 -2.8025601 4.002560 0.9991227
f2:SppB-f1:SppA 1.450000 -1.9525601 4.852560 0.8685338
f1:SppB-f1:SppA 3.016667 -0.3858935 6.419227 0.1150225
f3:SppB-f1:SppA 8.466667 5.0641065 11.869227 0.0000000
f2:SppB-f3:SppA 0.850000 -2.5525601 4.252560 0.9922487
f1:SppB-f3:SppA 2.416667 -0.9858935 5.819227 0.3344595
f3:SppB-f3:SppA 7.866667 4.4641065 11.269227 0.0000001
f1:SppB-f2:SppB 1.566667 -1.8358935 4.969227 0.8173904
f3:SppB-f2:SppB 7.016667 3.6141065 10.419227 0.0000019
f3:SppB-f1:SppB 5.450000 2.0474399 8.852560 0.0002022
```
We can see the mean differences for fertilizer combinations, for the two species and for all fertilizer*species combinations. By using the confidence intervals or the p-values we can conclude which of these combinations are significant or not.
6. Obtain Tukey grouping by using the following commands:
```tukey_grouping<-multcompLetters4(anova,tukey_multiple_comparisons)
print(tukey_grouping)
\$fertilizer
f3 f1 f2 control
"a" "b" "c" "d"
\$species
SppB SppA
"a" "b"
\$`fertilizer:species`
f3:SppB f1:SppB f2:SppB f3:SppA f1:SppA f2:SppA control:SppB control:SppA
"a" "b" "b" "bc" "bc" "cd" "de" "e"
```
7. Obtain a plot of differences in mean response for fertilizer*species combinations by using the following commands:
```par(mar=c(4.1,13,4.1,2.1))
plot(tukey_multiple_comparisons,las=2)
detach(greenhouse_2way_data)
```
5.1.03: Two-Factor Factorial - Greenhouse Example (R)
• Load the glucose in blood serum data.
• Obtain the ANOVA table with interaction.
• Obtain the ANOVA table without interaction.
• Obtain estimators and CIs for means for each treatment level.
• Obtain Tukey’s multiple comparisons CIs and grouping.
1. Load the glucose in blood serum data by using the following commands:
```setwd("~/path-to-folder/")
glucose_data <- read.table("glucose_data.txt",header=T)
attach(glucose_data)
```
2. Obtain the ANOVA table with interaction by using the following commands:
```anova<-aov(glucose ~ factor(method) + factor(doping) + factor(method)*factor(doping),data=glucose_data)
summary(anova)
Df Sum Sq Mean Sq F value Pr(>F)
factor(method) 1 264 264 98.347 3.92e-07 ***
factor(doping) 2 57026 28513 10632.526 < 2e-16 ***
factor(method):factor(doping) 2 14 7 2.577 0.117
Residuals 12 32 3
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
```
Here we can see that the interaction term is not significant, and we can drop it from the model. Also, notice that I have defined method and doping as factors since they have numeric values.
3. Obtain the ANOVA table without interaction by using the following commands:
```anova1<-aov(glucose ~ factor(method) + factor(doping),data=glucose_data)
summary(anova1)
Df Sum Sq Mean Sq F value Pr(>F)
factor(method) 1 264 264 80.27 3.58e-07 ***
factor(doping) 2 57026 28513 8677.63 < 2e-16 ***
Residuals 14 46 3
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
```
The Error SS is now 46.001, which is the sum of the interaction SS and the error SS of the model with the interaction. The df values were also added the same way.
4. Obtain estimators and CIs for means for each treatment level by using the following commands:
```library(lsmeans)
lsmeans(anova1,"method")
method lsmean SE df lower.CL upper.CL
1 123 0.604 14 122 125
2 116 0.604 14 114 117
Results are averaged over the levels of: doping
Confidence level used: 0.95
lsmeans(anova1,"doping")
doping lsmean SE df lower.CL upper.CL
1 43.7 0.74 14 42.1 45.3
2 136.8 0.74 14 135.2 138.4
3 178.3 0.74 14 176.7 179.9
Results are averaged over the levels of: method
Confidence level used: 0.95
```
5. Obtain Tukey’s multiple comparisons CIs and grouping by using the following commands:
```tukey_multiple_comparisons<-TukeyHSD(anova1,conf.level=0.95,ordered=TRUE)
tukey_multiple_comparisons
Tukey multiple comparisons of means
95% family-wise confidence level
factor levels have been ordered
Fit: aov(formula = glucose ~ factor(method) + factor(doping), data = glucose_data)
\$`factor(method)`
diff lwr upr p adj
1-2 7.655556 5.822828 9.488283 4e-07
\$`factor(doping)`
diff lwr upr p adj
2-1 93.10000 90.36089 95.83911 0
3-1 134.61667 131.87755 137.35578 0
3-2 41.51667 38.77755 44.25578 0
tukey_grouping<-multcompLetters4(anova1,tukey_multiple_comparisons)
print(tukey_grouping)
\$`factor(method)`
1 2
"a" "b"
\$`factor(doping)`
3 2 1
"a" "b" "c"
detach(glucose_data)
``` | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/05%3A_Multi-Factor_ANOVA/5.01%3A_Factorial_or_Crossed_Treatment_Designs/5.1.02%3A_Two-Factor_Factorial_-_Greenhouse_Example_%28Minitab%29.txt |
When setting up a multi-factor study, sometimes it is not possible to cross the factor levels. In other words, because of the logistics of the situation, we may not be able to have each level of treatment be combined with each level of another treatment.
Here is an example:
A research team interested in the lifestyle of high school students conducted a study to compare the activity levels of high school students across the 3 geographic regions in the United States, Northeast (NE), Midwest (MW), and the West (W). The study also included the comparison of activity levels among cities within each region. Two school districts were chosen from two major cities from each of these 3 regions and the response variable, the average number of exercise hours per week for high school students for each school district was recorded.
A diagram to illustrate the treatment design can be set up as follows. Here, the subscript $i$ identifies the regions, and the subscript $j$ indicates the cities:
The table above shows the data obtained: the grand mean, the marginal means which are the treatment level means, and finally, the cell means. The cell means are the averages of the two school district mean activity levels for each combination of Region and City.
This example drives home the point that the levels of the second factor (City) cannot practically be crossed with the levels of the first factor (Region) as cities are specific or unique to regions. Note that the cities are identified as 1 or 2 within each region. But it is important to note that city 1 in the Northeast is not the same as city 1 in the Midwest. The concept of nesting does come in useful to describe this type of situation and the use of parentheses is appropriate to clearly indicate the nesting of factors. To indicate that the City is nested within the factor Region, the notation: City(Region) will be used. Here, City is the nested factor and Region is the nesting factor.
We can partition the deviations as before into the following components: $\underbrace{Y_{ijk} - \bar{Y}_{...}}_{\text{Total deviation}} = \underbrace{\bar{Y}_{i..} - \bar{Y}_{...}}_{\text{A main effect}} + \underbrace{\bar{Y}_{ij.} - \bar{Y}_{i..}}_{\text{Specific B effect when A at the } i^{th} \text{ level}} + \underbrace{Y_{ijk} - \bar{Y}_{ij.}}_{\text{Residual}}$
For Factor B
When stating the Null Hypothesis for Factor B, the nested effect, alternative notation has to be used.
Up to this point, we have been stating Null Hypotheses in terms of the means (e.g. $H_{0}:\ \mu_{1} = \mu_{2} = \ldots = \mu_{k}$), but we can alternatively state a Null Hypothesis in terms of the parameters for that treatment in the model. For example, for the nesting factor A, we could also state the Null Hypothesis as $H_{0}: \ \alpha_{\text{Northeast}} = \alpha_{\text{Midwest}} = \alpha_{\text{West}} = 0 \text{ or } H_{0}: \ \text{all } \alpha_{i} = 0$
For the nested factor B, the Null Hypothesis should differentiate between the nesting and the nested factors, because we are evaluating the nested factor within the levels of the nesting factor.
So for the nested factor (City, nested within Region), we have the Null Hypothesis. $H_{0}: \ \text{all } \beta_{j(i)} = 0 \text{ vs. } H_{A}: \ \text{not all } \beta_{j(i)} = 0 \text{ for } j=1, 2,$
The $F$-tests can then proceed as usual using the ANOVA results. The first two columns of the ANOVA table should be as follows on the next page.
Note
1. There is no interaction between a nested factor and its nesting factor.
2. The nested factors always have to be accompanied by their nested factor. This means that the effect B does not exist and B(A) represents the effect of B within the factor A
3. df of B(A) = df of B + df of A*B (This is simply a mathematically correct identity and may not be of much practical use, as effects B(A) and A*B cannot coexist)
4. The residual effect of any ANOVA model is a nested effect - the replicate effect nested within the factor level combinations. Recall that the replicates are considered homogeneous and so any variability among them serves to estimate the model error.
5.02: Nested Treatment Design
Here is the SAS code to run the ANOVA model for the hours of exercise for high school students example discussed in lesson 5.2:
```data Nested_Example_data;
infile datalines delimiter=',';
input Region \$ City \$ ExHours;
datalines;
NE,NY,30
NE,NY,35
NE,Pittsburgh,18
NE,Pittsburgh,20
MW,Chicago,10
MW,Chicago,9
MW,Detroit,20
MW,Detroit,22
W,LA,18
W,LA,19
W,Seattle,4
W,Seattle,6
;
/*to run the nested ANOVA model*/
proc mixed data=Nested_Example_data method=type3;
class Region City;
model ExHours = Region City(Region);
store nested1;
run;
/*to obtain the resulting multiple comparison results*/
ods graphics on;
proc plm restore=nested1;
lsmeans Region / adjust=tukey plot=meanplot cl lines;
lsmeans City(Region) / adjust=tukey plot=meanplot cl lines;
run;
```
When we run this SAS program, here is the output that we are interested in:
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
Region 2 424.666667 212.333333 Var(Residual)+Q(Region, City(Region)) MS(Residual) 6 65.33 <.0001
City(Region) 3 496.750000 165.583333 Var(Residual)+Q(City(Region)) MS(Residual) 6 50.95 0.0001
Residual 6 19.500000 3.250000 Var(Residual)
The \(p\)-values above indicate that both Region and City(Region) are statistically significant. The plots and charts below obtained from the Tukey option specify the means which are significantly different.
The exercise hours on average are statistically higher in the northeastern region compared to the midwest and the west while the average exercise hours of these two regions are not significantly different.
Also, the comparison of the means between cities indicates that the high schoolers in New York city exercise significantly more than the other cities in the study. The exercise levels are similar among Detroit, Pittsburgh, and LA, while exercise levels of high schoolers in Chicago and Seattle are similar but significantly lower than all other cities in the study.
These grouping observations are further confirmed by the lines plots below. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/05%3A_Multi-Factor_ANOVA/5.02%3A_Nested_Treatment_Design/5.2.01%3A_Nested_Model_in_SAS.txt |
In Minitab, for the following (Nested Example Data):
Stat > ANOVA > General Linear Model > Fit General Linear Model
Enter the factors 'Region' and 'City' in the Factors box, then click on Random/Nest...Here is where we specify the nested effect of City in Region.
The output is shown below.
Factor Information
Model Summary
Following the ANOVA run, you can generate the mean comparisons by
Stat > ANOVA > General Linear Model > Comparisons
Then specify "Region" and "City(Region)" for the comparisons by checking the boxes.
Then choose Graphs to get the following dialog box, where "Interval plot for difference of means" should be checked.
The outputs are as follows.
Comparison for Ex_hours
Tukey Pairwise Comparisons: Region
Grouping Information Using Tukey Method and 95% Confidence
Means that do not share a letter are significantly different.
Tukey Pairwise Comparisons: (City)Region
Grouping Information Using Tukey Method and 95% Confidence
Means that do not share a letter are significantly different.
5.2.03: Nested Model in R
• Load the Exercise Hours data.
• Obtain the ANOVA table for the nested treatment design.
• Obtain estimators and CIs for means for each region and city.
• Obtain means plot for region and city within the region.
• Obtain Tukey’s multiple comparisons CIs.
1. Load the Exercise Hours data by using the following commands:
```setwd("~/path-to-folder/")
ex_hours_data <- read.table("ex_hours_data.txt",header=T)
attach(ex_hours_data)
```
2. Obtain the ANOVA table for the nested treatment design by using the following commands:
```nested<-aov(Ex_hours ~ Region+Region/City,data=ex_hours_data)
summary(nested)
# Df Sum Sq Mean Sq F value Pr(>F)
# Region 2 424.7 212.33 65.33 8.46e-05 ***
# Region:City 3 496.8 165.58 50.95 0.000116 ***
# Residuals 6 19.5 3.25
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
```
3. Obtain estimators and CIs for means for each region and city by using the following commands:
```library(lsmeans)
lsmeans(nested,"Region")
# Region lsmean SE df lower.CL upper.CL
# MW 15.2 0.901 6 13.04 17.5
# NE 25.8 0.901 6 23.54 28.0
# W 11.8 0.901 6 9.54 14.0
#Results are averaged over the levels of: City
#Confidence level used: 0.95
lsmeans(nested,"City")
#City Region lsmean SE df lower.CL upper.CL
# Chicago MW 9.5 1.27 6 6.38 12.62
# Detroit MW 21.0 1.27 6 17.88 24.12
# NY NE 32.5 1.27 6 29.38 35.62
# Pittsburgh NE 19.0 1.27 6 15.88 22.12
# LA W 18.5 1.27 6 15.38 21.62
# Seattle W 5.0 1.27 6 1.88 8.12
#Confidence level used: 0.95
```
4. Obtain means plot for region and city within region by using the following commands:
```library(plotrix)
region_means<-as.data.frame(lsmeans(nested,"Region"))
plotCI(x = region_means\$lsmean,y = NULL ,li = region_means\$lower.CL, ui = region_means\$upper.CL, xaxt = "n", xlab="Region",ylab="ExHours")
axis(1, at=1:3, labels=region_means\$Region)
```
```city_means<-as.data.frame(lsmeans(nested,"City"))
City_Region<-paste(city_means\$City,city_means\$Region)
plotCI(x = city_means\$lsmean,y = NULL ,li = city_means\$lower.CL, ui = city_means\$upper.CL, xaxt = "n", xlab="City(Region)",ylab="ExHours")
axis(1, at=1:6, labels=City_Region)
```
5. Obtain Tukey’s multiple comparisons CIs by using the following commands:
```library(multcomp)
library(multcompView)
tukey_multiple_comparisons_region<-TukeyHSD(nested,"Region",conf.level=0.95,ordered=TRUE)
tukey_multiple_comparisons_region
Tukey multiple comparisons of means
95% family-wise confidence level
factor levels have been ordered
Fit: aov(formula = Ex_hours ~ Region + Region/City, data = ex_hours_data)
# \$Region
# diff lwr upr p adj
#MW-W 3.5 -0.4112978 7.411298 0.0747598
#NE-W 14.0 10.0887022 17.911298 0.0000836
plot(tukey_multiple_comparisons_region)
```
```tukey_multiple_comparisons_city<-TukeyHSD(nested,"Region:City",conf.level=0.95,ordered=TRUE)
cities<-as.data.frame(na.omit(tukey_multiple_comparisons_city\$"Region:City"))
cities
# diff lwr upr p adj
# MW:Chicago-W:Seattle 4.5 -4.96579743 13.965797 0.5867601138
# W:LA-W:Seattle 13.5 4.03420257 22.965797 0.0087623039
# NE:Pittsburgh-W:Seattle 14.0 4.53420257 23.465797 0.0072411812
# MW:Detroit-W:Seattle 16.0 6.53420257 25.465797 0.0035459602
# NE:NY-W:Seattle 27.5 18.03420257 36.965797 0.0001761692
# W:LA-MW:Chicago 9.0 -0.46579743 18.465797 0.0626471065
# NE:Pittsburgh-MW:Chicago 9.5 0.03420257 18.965797 0.0491884424
# MW:Detroit-MW:Chicago 11.5 2.03420257 20.965797 0.0198221594
# NE:NY-MW:Chicago 23.0 13.53420257 32.465797 0.0004610102
# NE:Pittsburgh-W:LA 0.5 -8.96579743 9.965797 1.0000000000
# MW:Detroit-W:LA 2.5 -6.96579743 11.965797 0.9752059356
# NE:NY-W:LA 14.0 4.53420257 23.465797 0.0072411812
# MW:Detroit-NE:Pittsburgh 2.0 -7.46579743 11.465797 0.9960158169
# NE:NY-NE:Pittsburgh 13.5 4.03420257 22.965797 0.0087623039
# NE:NY-MW:Detroit 11.5 2.03420257 20.965797 0.0198221594
library(plotrix)
city_diff<-as.character(c("
MW:Chicago-W:Seattle","W:LA-W:Seattle", "NE:Pittsburgh-W:Seattle","MW:Detroit-W:Seattle","NE:NY-W:Seattle ","W:LA-MW:Chicago ","NE:Pittsburgh-MW:Chicago","MW:Detroit-MW:Chicago","NE:NY-MW:Chicago ","NE:Pittsburgh-W:LA ","MW:Detroit-W:LA","NE:NY-W:LA ", "MW:Detroit-NE:Pittsburgh", "NE:NY-NE:Pittsburgh","NE:NY-MW:Detroit"))
par(mar=c(8, 4, 2, 2) + 0.1)
plotCI(x = cities\$diff,y = NULL ,li = cities\$lwr, ui = cities\$upr, xaxt = "
n",ylab="Differences of Means",xlab="")
abline(h=0)
axis(1, at=1:15, labels=city_diff,las = 2, cex.axis = 0.8)
```
5.03: Crossed-Nested Designs
Multi-factor studies can involve factor combinations in which factors are crossed and/or nested. These treatment designs are based on the extensions of the concepts discussed so far.
Consider an example (from Canavos and Koutrouvelis, 2009) where machines in an assembly process are evaluated for assembly times. There were three factors of interest: Machine ID (1, 2, or 3), Configuration (1 or 2), and Power level (1, 2, or 3).
It turns out that each machine can be operated at each power level, and so these factors can be crossed. Also, each configuration can be operated at each power level and so these factors also are crossed. But the configurations (1 or 2) are unique to each machine. As a result, the configuration is nested within the machine.
The statistical model contains both crossed and nested effects and is: $Y_{ijkl} = \mu + \alpha_{i} + \beta_{j(i)} + \gamma_{k} + (\alpha \gamma)_{ik} + (\beta \gamma)_{j(i) k} + \epsilon_{ijk}$
with the ANOVA table as follows:
Notice that the two main effects, Machine and Power, are included in the model along with their interaction effect. The nested relationship of Configuration within Machine is represented by the Configuration(Machine) term and the crossed relationship between Configuration and Power is represented by their interaction effect.
Notice that the main effect Configuration and the crossed effect Configuration × Machine are not included in the model. This is consistent with the facts that a nested effect cannot be represented as the main effect and also that a nested effect cannot interact with its nesting effect. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/05%3A_Multi-Factor_ANOVA/5.02%3A_Nested_Treatment_Design/5.2.02%3A_Nested_Model_in_Minitab.txt |
Exercise $1$: $\text{CO}_{2}$ Emissions
To study the variability in CO2 emission rate by global regions 4 countries: US, Britain, India, and Australia were chosen. From each country, 3 major cities were chosen and the emission rates for each month for the year 2019 were collected.
1. What type of model is this?
1. nested
2. cross-nested
3. factorial
2. How many factors?
1. 4
2. 3
3. 2
3. The replicates are...
1. 12 months of 2019
2. countries US, Britain, India, and Australia
3. major cities in US, Britain, India, and Australia
4. The residual effect in the ANOVA model is...
1. country*city*month
2. month(city(country))
3. month(country*city)
5. How many degrees of freedom?
1. 66
2. 132
3. 88
Show Answers and Explanations
Answers
Q1 - 1. nested
Q2 - 3. 2 factors
Q3 - 1. 12 months of 2019
Q4 - 2. month(city(country)
Q5 - 2. 132
Explanations
Residual effect (or error term) is the month(city(country), which is the nested effect of "month", the replicate, within the combinations of the two factors "country" and "city". One way to double-check this answer is to verify if the df values are the same.
\begin{aligned} \text{error term df} &= \text{total df} - (\text{sum df values of the model terms}) \ &= (144-1) - (\text{country df} + \text{city(country) df}) \ &= 143 - (3+2*4) \ &= 143-11 = 132 \end{aligned}
$\text{df for city(country)} = 11(12) = 132$
See the Section 5.2 ANOVA table for df formula.
Exercise $2$
A military installation is interested in evaluating the speed of reloading a large gun. Two methods of reloading are considered, and 3 groups of cadets were evaluated (slight, average, and heavy individuals). Three teams were set up within each group and they wanted to identify the fastest team within each group to go on to a demonstration for the military officials. Each team performed the reloading with each method two times (two replications).
1. Identify (i.e. name) the treatment design.
1. nested
2. cross-nested
3. factorial
2. They started to construct the ANOVA table which is given below. Given that there are a total of 36 observations in the dataset, there seems to be a missing source of variation in the analysis. What is this source of variation?
1. Team*group*method
2. Team (Group)*method
3. Team*Group
3. How many degrees of freedom are associated with the error term?
1. 6
2. 24
3. 18
Show Answers
Q1 - 2. cross-nested
Q2 - 2. Team (Group)*method
Q3 - 3. 18
Exercise $3$: GPA Comparisons
The GPA comparison of four popular majors—biology, business, engineering, and psychology—between males and females is of interest. For 6 semesters, the average GPA of each of these majors for male and female students was computed.
1. What type of model is this?
1. nested
2. cross-nested
3. crossed
2. How many factors?
1. 4
2. 3
3. 2
3. The replicates are...
1. semesters
2. majors
3. gender
4. The residual effect in the ANOVA model is...
1. major*gender*semester
2. semester(gender*major)
3. semester(major(gender))
5. How many degrees of freedom?
1. 48
2. 40
3. 2
Show Answers and Explanations
Answers
Q1 - 3. crossed
Q2 - 3. 2 factors
Q3 - 1. semesters
Q4 - 2. semester(gender*major)
Q5 - 2. 40
Explanations
Residual effect (or error term) is semester (gender*major). The error term is the nested effect of "semester", the replicate nested within gender*major, which is the "combined effect" of the factors. One way to double-check is to verify if df values are the same.
\begin{aligned} \text{error term df} &= \text{total df} - (\text{sum df values of the model terms}) \ &= (48-1) - (\text{major df + gender df + major*gender df}) \ &= 47 - (3 + 1 + 3*1) \ = 40 \end{aligned}
$\text{df for semester(major*gender)} = 5 * 8 = 40$
See the Section 5.2 ANOVA table for df formula.
5.05: Chapter 5 Summary
In this lesson, we discussed important elements of the "Treatment Design," one of the two components of an "Experimental Design." We are now familiar with the main effects and interaction effects of a factorial design.
In a full factorial design, the experiment is carried out at every factor level combination. Most factorial studies do not go beyond a two-way interaction, and if a two-way interaction is significant, the mean response values should be compared among different combinations of the two factors rather than among the single factor levels. In other words, the focus should be on response vs. interaction effect rather than response vs. main effects. An Interaction plot is a useful graphical tool to understand the extent of interactions among factors (or treatments) with parallel lines indicating no interaction.
In a nested design, the experiment need not be conducted at every combination of levels in all factors. Given two factors in a nested design, there is a distinction between the nested and the nesting factor. The levels of the nested factor may be unique to each level of the nesting factor. Therefore, the comparison of the nested factor levels should be made within each level of the nesting factor—a fact that should be kept in mind when stating null and alternative hypotheses for the nested factor(s), and also when writing programming code.
5.06: Treatment Design Summary (Optional Enrichment Material)
In an effort to summarize how to think about sums of squares and degrees of freedom and how this translates into a model that can be implemented in SAS, Dr. Rosenberger walks you through this process in the videos below. Pay attention to the subscripts and these are the keys to understanding this material. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/05%3A_Multi-Factor_ANOVA/5.04%3A_Try_It.txt |
Overview
So far, in our discussion of treatment designs, we have made the (unstated) assumption that the treatment levels were chosen intentionally by the researcher as dictated by his/her specific interests. The scope of inference in this situation is limited to the specific (or fixed) levels used in the study. However, this is not always the case. Sometimes, treatment levels may be a (random) sample of possible levels, and the scope of inference is to a larger population of all possible levels.
If it is clear that the researcher is interested in comparing specific, chosen levels of treatment, that treatment is called a fixed effect. On the other hand, if the levels of the treatment are a sample of a larger population of possible levels, then the treatment is called a random effect.
Learning Objectives
Upon completion of this lesson, you should be able to:
1. Extend the treatment design to include random effects.
2. Understand the basic concepts of random-effects models.
3. Calculate and interpret the intraclass correlation coefficient.
4. Combining fixed and random effects in the mixed model.
5. Work with mixed models that include both fixed and random effects.
06: Random Effects and Introduction to Mixed Models
When a treatment (or factor) is a random effect, the model specifications together with relevant null and alternative hypotheses will have to be changed. Recall the cell means model defined in Chapter 4 for the fixed effect case, which has the model equation: $Y_{ij} = \mu_{i} + \epsilon_{ij}$ where $\mu_{i}$ are parameters for the treatment means.
For the single factor random effects model we have: $Y_{ij} = \mu_{i} + \epsilon_{ij}$ where $\mu_{i}$ and $\epsilon_{ij}$ are independent random variables such that $\mu_{i} \overset{iid}{\sim} \mathcal{N} \left(\mu, \sigma_{\mu}^{2}\right)$ and $\epsilon_{ij} \overset{iid}{\sim} \mathcal{N} \left(0, \sigma_{\epsilon}^{2}\right)$. Here, $i = 1, 2, \ldots, T$ and $j = 1, 2, \ldots, n_{i}$, where $n_{i} \equiv n$ if balanced.
Notice that the random effects ANOVA model is similar in appearance to the fixed effects ANOVA model. However, the treatment mean $\mu_{i}$'s are constant in the fixed-effect ANOVA model, whereas in the random-effects ANOVA model the treatment mean $\mu_{i}$'s are random variables.
Note that the expected mean response, in the random effects model stated above, is the same at every treatment level and equals $\mu$.
$E \left(Y_{ij}\right) = E \left(\mu_{i} + \epsilon_{ij}\right) = E \left(\mu_{i}\right) + E \left(\epsilon_{ij}\right) = \mu$
The variance of the response variable (say $\sigma_{Y}^{2}$) in this case can be partitioned as: $\sigma_{Y}^{2} = V \left(Y_{ij}\right) = V \left(\mu_{i} + \epsilon_{ij}\right) = V \left(\mu_{i}\right) + V \left(\epsilon_{ij}\right) = \sigma_{\mu}^{2} + \sigma_{\epsilon}^{2}$ as $\mu_{i}$ and $\epsilon_{ij}$ are independent random variables.
Similar to fixed effects ANOVA model, we can express the random effects ANOVA model using the factor effect representation, using $\tau_{i} = \mu_{i} - \mu$. Therefore the factor effects representation of the random effects ANOVA model would be: $Y_{ij} = \mu + \tau_{i} + \epsilon_{ij}$ where $\mu$ is a constant overall mean, and $\tau_{i}$ and $\epsilon_{ij}$ are independent random variables such that $\tau_{i} \overset{iid}{\sim} \mathcal{N} \left(0, \sigma_{\mu}^{2}\right)$ and $\epsilon_{ij} \overset{iid}{\sim} \mathcal{N} \left(0, \sigma_{\epsilon}^{2}\right)$. Here, $i = 1, 2, \ldots, T$ and $j = 1, 2, \ldots, n_{i}$, where $n_{i} \equiv n$ if balanced. Here, $\tau_{i}$ is the effect of the randomly selected $i^{th}$ level.
The terms $\sigma_{\mu}^{2}$ and $\sigma_{\epsilon}^{2}$ are referred to as variance components. In general, as will be seen later in more complex models, there will be a variance component associated with each effect involving at least one random factor.
Variance components play an important role in analyzing random effects data. They can be used to verify the significant contribution of each random effect to the variability of the response. For the single factor random-effects model stated above, the appropriate null and alternative hypothesis for this purpose is: $H_{0}: \ \sigma_{\mu}^{2} = 0 \text{ vs. } H_{A}: \ \sigma_{\mu}^{2} > 0$
Similar to the fixed effects model, an ANOVA analysis can then be carried out to determine if $H_{0}$ can be rejected.
The MS and the df computations of the ANOVA table are the same for both the fixed and random-effects models. However, the computations of the F-statistics needed for hypothesis testing require some modification.
Specifically, the F statistics denominator will no longer always be the mean squared error (MSE or MSERROR) and will vary according to the effect of interest (listed in the Source column of the ANOVA table). For a random-effects model, the quantities known as Expected Means Squares (EMS), shown in the ANOVA table below, can be used to identify the appropriate F-statistic denominator for a given source in the ANOVA table. These EMS quantities will also be useful in estimating the variance components associated with a given random effect. Note that the EMS quantities are in fact the population counterparts of the mean sums of squares (MS) that we are already familiar with. In SAS the proc mixed, method=type3 option will generate the EMS column in the ANOVA table output.
Note
Variance components are NOT synonymous with mean sums of squares. Variance components are usually estimated by using the Method of Moments where algebraic equations, created by setting the mean sums of squares (MS) equal to the EMS for the relevant effects, are solved for the unknown variance components. For example, the variance component for the treatment in the single-factor random effects discussed above can be solved as:
$s_{\text{among trts}}^{2} = \frac{MS_{trt} - MS_{error}}{n}$
This is by using the two equations:
$MS_{error} = \sigma_{\epsilon}^{2}$
$MS_{trt} = \sigma_{\epsilon}^{2} + n \sigma_{\mu}^{2}$
More about variance components...
Often the variance component of a specific effect in the model is expressed as a percent of the total variation of the variation in the response variable.
Another common application of variance components is when researchers are interested in the relative size of the treatment effect compared to the within-treatment level variation. This leads to a quantity called the intraclass correlation coefficient (ICC), defined as: $ICC = \frac{\sigma_{\text{among trts}}^{2}}{\sigma_{\text{\text{among trts}}^{2} + \sigma_{\text{within trts}}^{2}}$
For single random factor studies, $ICC = \frac{\sigma_{mu}^{2}}{\sigma_{mu}^{2} + \sigma_{\epsilon}^{2}}$. ICC can also be thought of as the correlation between the observations within the group (i.e. $\text{corr} \left(Y_{ij}, Y_{ij'}\right)$, where $j \neq j'$. Small values of ICC indicate a large spread of values at each level of the treatment, whereas large values of ICC indicate relatively little spread at each level of the treatment: | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/06%3A_Random_Effects_and_Introduction_to_Mixed_Models/6.01%3A_Random_Effects.txt |
Consider a study of Battery Life, measured in hours, where 4 brands of batteries are evaluated using 4 replications in a completely randomized design (Battery Data):
Brand A Brand B Brand C Brand D
110 118 108 117
113 116 107 112
108 112 112 115
115 117 108 119
A reasonable question to ask in this study would be, should the brand of the battery be considered a fixed effect or a random effect?
If the researchers were interested in comparing the performance of the specific brands they chose for the study, then we have a fixed effect.
But if the researchers were actually interested in studying the overall variation in battery life, so that the results would be applicable to all brands of batteries, then they may have chosen (presumably with a random sampling process) a sample of 4 of the many brands available and tested 4 batteries of each of these brands. In this latter case, the battery brand would add a dimension of variability to battery life and can be considered a random effect.
Now, let us use SAS proc mixed; to compare the results of battery brand as a fixed vs. random effect:
1. Fixed Effect model:
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
Brand 3 141.687500 47.229167 Var(Residual) + Q(Brand) MS(Residual) 12 6.21 0.0086
Residual 12 91.250000 7.604167 Var(Residual) . . . .
2. Random Effect model:
Type 3 Analysis of Variance
Source DF Sum of Squares Mean Square Expected Mean Square Error Term Error DF F Value Pr > F
Brand 3 141.687500 47.229167 Var(Residual) + 4 Var(Brand) MS(Residual) 12 6.21 0.0086
Residual 12 91.250000 7.604167 Var(Residual) . . . .
Covariance Parameter Estimates
Cov Parm Estimate
Brand 9.9063
Residual 7.6042
We can verify the estimated variance component (arrow above) for the random treatment effect as: $s_{\text{among trts}}^{2} = \frac{MS_{trt} - MS_{error}}{n} = \frac{47.229 - 7.604}{4} = 9.9063$
With this, we can calculate the ICC as $ICC = \frac{9.9063}{9.9063 + 7.604} = 0.5657$
The key points in comparing these two ANOVAs are 1) the scope of inference and 2) the hypothesis being tested. For a fixed effect, the scope of inference is restricted to only 4 brands chosen for comparison and the Null hypothesis is a statement of equality of means. In contrast, as a random effect, the scope of inference is the larger population of battery brands and the Null hypothesis is a statement that the variance due to battery brand is 0.
Using R
R: Single Random Effect
• Load the battery life data.
• Obtain the ANOVA for a single random effect.
Show Detailed Steps
1. Load the battery life data by using the following commands:
setwd("~/path-to-folder/")
battery_data <- read.table("battery_data.txt",header=T)
attach(battery_data)
2. Obtain the ANOVA for a single random effect by using the following commands:
library(lmerTest)
library(lme4)
battery_anova<-lmer(lifetime ~ (1 | trt),battery_data)
summary(battery_anova)
Linear mixed model fit by REML. t-tests use Satterthwaites method ['lmerModLmerTest']
Formula: lifetime ~ (1 | trt)
Data: battery_data
REML criterion at convergence: 81.3
#Scaled residuals:
# Min 1Q Median 3Q Max
#-1.35317 -0.69070 0.07355 0.69665 1.34279
#Random effects:
# Groups Name Variance Std.Dev.
# trt (Intercept) 9.906 3.147
# Residual 7.604 2.758
#Number of obs: 16, groups: trt, 4
#Fixed effects:
# Estimate Std. Error df t value Pr(>|t|)
#(Intercept) 112.938 1.718 3.000 65.73 7.76e-06 ***
#---
#Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#confint(battery_anova)
# 2.5 % 97.5 %
#.sig01 0.6530752 7.166913
#.sigma 1.9371621 4.374014
#(Intercept) 109.1585596 116.716437
Note that the command lmer() gives the ANOVA table only for the fixed effects. Therefore, in this example, since there are no fixed effects, we won’t get the ANOVA table. In the "Random effects" section of the output, under the column variance we get the estimates for $\sigma_{\alpha}^{2}$ and $\sigma^{2}$, which are equal to 9.906 and 7.604 respectively. In the "Fixed effects" section under the column estimate, we get the estimate of $mu$, or the overall mean, which is equal to 112.938. With the command confint() we will get confidence intervals for the standard deviations and the overall mean. If you take the square of the lower and upper bounds, you will get a confidence interval for the model variances.
Alternatively, we can use the command aov() which gives a partial ANOVA table.
battery_anova1<-aov(lifetime~Error(trt),battery_data)
summary(battery_anova1)
#Error: trt
# Df Sum Sq Mean Sq F value Pr(>F)
#Residuals 3 141.7 47.23
#Error: Within
# Df Sum Sq Mean Sq F value Pr(>F)
#Residuals 12 91.25 7.604
detach(battery_data)
Note that both of these commands in R don't give the $F$-values and $p$-values for the tests. Therefore, these must be done manually. | textbooks/stats/Advanced_Statistics/Analysis_of_Variance_and_Design_of_Experiments/06%3A_Random_Effects_and_Introduction_to_Mixed_Models/6.02%3A_Battery_Life_Example.txt |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.