Datasets:

SouthernCrossAI
/

Arxiv

Dataset card Data Studio Files Files and versions Community

Split (1)

train · 132k rows

Filename stringlengths 22 64	Paragraph stringlengths 8 5.57k
Processed_Heightened_Faraday_Complexity_in_the_inner_1_kpc_o.txt	Another way of comparing the magnitude of our observed RM variations with other areas is with a structure function. We created a structure function of the Galactic Plane from the catalogue of Taylor et al. (2009) and a structure function for the set from O’Sullivan et al. (2017) to determine if our results shown in Figure 9 are unusual, compared with other RM catalogues. This was calculated using the same method as described in Section 3.3. We used a subsection of the Taylor et al. (2009) catalogue as shown in the top-left of Figure 10 to compute a Galactic Plane RM structure function (Stil et al. 2011). The region selected was a 20° × 20° box centred on 𝑙 = 60°, 𝑏 = 0°. The distribution of angular separations for our structure function, the Galactic Plane structure function and the structure function calculated from O’Sullivan et al. (2017) are shown in the top-right of Figure 10; the structure functions are presented in the bottom of Figure 10.
Processed_Heightened_Faraday_Complexity_in_the_inner_1_kpc_o.txt	log10 (SFPlane) ∼ 0.88 ∗ log10 (𝛿𝜃) + 2.2. The value of the Plane region structure function is 1 order of magnitude lower than the height of the structure function shown in Figure 9. The height slope of the O’Sullivan et al. (2017) SF is 3 orders of magnitude lower than the height of our structure function. This signiﬁcant diﬀerence suggests that the amplitude of the ﬂat structure function found in Figure 9 would not be expected if our surveyed region was away from the Galactic Centre.
Processed_Heightened_Faraday_Complexity_in_the_inner_1_kpc_o.txt	Based on the magnitude of RM ﬂuctuations in the direction of the Galactic Centre and the knowledge that only the Galactic Centre region itself is capable of producing such large RMs, we conclude that the Galactic Centre region is likely responsible for the large (cid:104)\| RM1 − RM2\|(cid:105) we observe. If the bulk of the Faraday rotation occurs at the Galactic Centre then these multiple lines-of-sight within a beam are probing the magneto-ionic environment at the distance of the Galactic Centre.
Processed_Heightened_Faraday_Complexity_in_the_inner_1_kpc_o.txt	To include small angular scales in the structure function, we use our modelled values of (cid:104)\| RM1 − RM2\|2(cid:105) (see Section 4.1 and Table 3). We note that the RM structure function is deﬁned as the mean square diﬀerence of RM across varying angular size scales (shown in Eq. 8). Measuring (cid:104)\| RM1 − RM2\|2(cid:105) provides an equivalent extension to scales below the synthesised beam-width. For these scales, we account for the intrinsic RM scatter of multiple screens along the line-of-sight by subtracting the mean squared, (cid:104)\| RM𝑖 − RM 𝑗 \|2(cid:105), from O’Sullivan et al. (2017) of 2700 ± 700 rad2 m−4 (see Section 4.1). As for the RM structure function (see Section 3.3) we subtract the intrinsic scatter of RM for the Milky Way ISM and extra-galactic contribution found by Schnitzeler (2010) of 64 rad−2 m−4 and 36 rad−2 m−4, respectively.
Processed_Heightened_Faraday_Complexity_in_the_inner_1_kpc_o.txt	For unresolved sources, the angular separation we use the average width of our beams of 17”, which serves as an upper limit.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Results for nucleon matrix elements (arising from moments of structure functions) and form factors from a mixture of runs using Wilson, clover and overlap fermions (both quenched and unquenched) are presented and compared in an eﬀort to explore the size of the chiral ‘regime’, lattice spacing errors and quenching artefacts. While no run covers this whole range of eﬀects the partial results indicate a picture of small lattice spacing errors, small quenching eﬀects and only reaching the chiral regime at rather light quark masses.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	at pseu- doscalar masses down to ∼ 560MeV in order to see if there are any discernable quenching eﬀects.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Note that the physical pion mass is about mπ ∼ 140MeV and we use r0 = 0.5fm ≡ (394.6MeV)−1 to set the scale.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Wilson or clover fermions, amq = (1/κ − 1/κc)/2 with κc to be determined). Note that these refer to the number of conﬁgurations used for ﬁnding the matrix elements; for the nucleon/pion mass sometimes a higher statistic was used.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	y-axis to be ∝ m2 N seems phenomenologically to lead to linear lines (although possibly the lightest point in a data set might be bending slightly up- wards). A pseudoscalar mass of ∼ 400MeV cor- responds to (r0mps)2 ∼ 1.0 while mps ∼ 600MeV gives (r0mps)2 ∼ 2.3. Most of the lattice data lies above the region mps ∼ 600MeV, with the ex- ception of the quenched Wilson fermion results. In what region might the leading order χPT be applicable? In Fig. 1 we also plot a representa- tive curve (for the quenched nucleon mass). A breakdown is seen at about (r0mps)2 ∼ 1 or mps ∼ 400MeV, as the gradient of the χPT result is too steep. Higher order χPT might improve the situation [8,9].
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Figure 1. (r0mN )2 versus (r0mps)2 for the four sets under consideration. The experimental nucleon mass is denoted by a star. The chiral limit m2 ps = 0 is shown as a short dashed line, while the physical pion mass is denoted by the long dashed √2mK . The (quenched) χPT line. Also shown as a dot-dashed line is the mass of a hypothetical ss meson calculated as line uses the numerical values from eq. (8).
Processed_Structure_functions_and_form_factors_close_to_the_.txt	be renormalised. Techniques for both steps are standard, eg [10]. In general, one calculates non- singlet, NS, matrix elements because then the diﬃcult to compute one-quark-line-disconnected part of the matrix element cancels (although for the vector current this additional piece is proba- bly very small, [11], so we shall ignore this point here).
Processed_Structure_functions_and_form_factors_close_to_the_.txt	the vector current is conserved however allows a more direct NP determination of ZV , [11,12]. For overlap fermions in particular this appears to indicate that there are signiﬁcant diﬀerences be- tween perturbative based results and NP results; however for Wilson or clover results there is of- ten only a couple of percent (or less) diﬀerence between TI and NP results.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Figure 2. of 2GeV versus (r0mps)2, upper picture, and v − (F v2 ps against (a/r0)2, lower picture, Filled circles, squares and diamonds represent β = 6.0, 6.2 and 6.4 re- spectively. The MRS phenomenological value of vNS;M S is represented by a star.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	rors. First we note that there seem to be very small a discretisation errors and indeed the nu- merical eﬀect of additional operators needed to ensure O(a) improvement seems to be negligi- ble, [15]. Secondly a linear chiral extrapolation seems to describe the data adequately. (Again, as for the nucleon mass, we might expect χPT to be only valid for mps ∼< 400MeV.) A result of vM S 2b (2GeV) = 0.25(1) is found. This is to be com- pared to the phenomenological result of ∼ 0.18, so we see about a 40% discrepency.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Turning now to the unquenched data, in Fig. 3 we show the results. Again similar conclusions hold as for the quenched case, although it should be noted that we are not so close to the continuum limit. We ﬁnd vM S (2GeV) = 0.27(2). So at least in this quark mass range quenching eﬀects seem to be small.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Figure 3. Same as Fig. 2 for unquenched clover fermions. Filled circles, squares, up triangle, diamonds and down triangle represent β = 5.20, 5.25, 5.26, 5.29 and 5.40 re- spectively.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	numerical results to reach the phenomenological value seems diﬃcult, as down to ∼ 300MeV they seem to be rather linear. Of course we would not necessarily expect complete agreement with the phenomenological result, however past expe- rience with the quenched approximation does sug- gest that it can lead to values only a few percent away from experimental values.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Figure 5. (2GeV) using quenched Wilson fermions at β = 6.0. The 163 32 results are shown as circles, the 243 48 results as diamonds. A linear ﬁt, eq. (6), is also shown as a dashed line. The (quenched) χPT line uses the numerical values from eq. (10).
Processed_Structure_functions_and_form_factors_close_to_the_.txt	with mV ∼ 0.82GeV, µp ∼ 2.79 and µn ∼ −1.91. (Present experimental results indicate however e(q2)/Gp that, [18], Gp m(q2) is decreasing with in- creasing −q2.) In Fig. 6 we show Gp m for the proton and in Fig. 7 for the neutron. Note that we have made simple linear chiral extrapo- lations here, see eq. (6), but at present we can do little more than this, as in particular on the large 323 × 48 lattice, we only have two quark mass values. The technique used here is the same as described in [19,20]. Previous lattice calcula- tions described in these papers seem to show little or no lattice artifacts. Due to the larger lattice size, we are now able to go to smaller momentum min ∼ (2π/32a)2 ∼ 0.17GeV2 than transfer −q2 was previously possible. We ﬁnd that the lattice result tracks the experimental values reasonably well, although with rather large errors. Finally we note that as usual the electric form factor of the neutron remains diﬃcult to measure on the lattice (see [21] for a recent attempt).
Processed_Structure_functions_and_form_factors_close_to_the_.txt	Figure 6. Gp q2 using quenched Wilson fermions at β = 6.0. The previous experimental results are given by the open circles.
Processed_Structure_functions_and_form_factors_close_to_the_.txt	The Wilson and clover numerical calcula- tions were performed on the Hitachi SR8000 at LRZ (Munich), the Cray T3Es at EPCC (Edinburgh), NIC (J¨ulich) and ZIB (Berlin) as well as the APE/Quadrics at NIC (Zeuthen). The overlap numerical calculations were per- formed at NIC J¨ulich, HLRN Berlin and NeSC Edinburgh (IBM Regattas), NIC Zeuthen and Southampton (PC Clusters), and HPCF Cran- ﬁeld (SunFire). We thank all these institu- tions for support. This work is supported by the European Community’s Human Potential Program under contract HPRN-CT-2000-00145 Hadrons/Lattice QCD and by the DFG un- der contract FOR 465 (Forschergruppe Gitter- Hadronen-Ph¨anomenologie).
Processed_Software_Engineering_in_Australasia.txt	Six months ago an important call was made for researchers globally to provide insights into the way Software Engineering is done in their region. Heeding this call, we hereby outline the position Software Engineering in Australasia (New Zealand and Australia). This article first considers the software development methods, practices and tools that are popular in the Australasian software engineering community. We then briefly review the particular strengths of software engineering researchers in Australasia. Finally, we make an open call for collaborators by reflecting on our current position and identifying future opportunities.
Processed_Software_Engineering_in_Australasia.txt	One such study surveyed practitioners working in Brazil, Finland, and New Zealand in a transnational study where similar findings were reported across the three territories [1]. Outcomes from this study revealed that the sample of practitioners surveyed focused on a small portfolio of projects that were of short duration. In addition, Scrum and Kanban were used most; however, some practitioners also used more traditional methods. Coding Standards, Simple Design and Refactoring were used most by practitioners, and these practices were held to be largely suitable for project and process management [1]. Interestingly, another study also found the same threats to software development success across Brazil, Finland, and New Zealand [2]. While evidence here suggests that there may be similarities in the software development three countries, and particularly in terms of the size of organisations, this evidence may also point to the global language that is ‘software engineering’. In fact, globally, software engineers now place increasing value on agility this process, and methods and practices notwithstanding the various hybrids that are typically implemented in practice [3].
Processed_Software_Engineering_in_Australasia.txt	Australian companies is Scrum, with a mean team size of 18 members and a mean iteration length of just under five weeks. The most followed Agile techniques are daily standups, Kanban boards, sprints, and the sprint backlog, whereas pair programming and self- assignment were the least followed practices.
Processed_Software_Engineering_in_Australasia.txt	Australasian software engineering researchers work in all knowledge areas. Figure 1 provides a word cloud depicting Australasian software engineering researchers’ focus, where the size of the word denotes the relative degree of research focus across all active researchers. Of the areas listed, empirical software engineering, formal methods, cloud computing, and artificial intelligence are particular strengths. Other significant research areas cover specific variances of cloud computing, formal methods, artificial intelligence and machine learning, programming languages and algorithms, program analysis and repair, mining software software engineering requirements engineering, software design and architecture, software engineering education, software quality and testing, software development methods and practices, process mining, software measurements, empirical software engineering, human factors in software engineering, green software engineering, NLP applications, distributed systems, empirical methods and data quality, theory building in software engineering, project management, and so on.
Processed_Software_Engineering_in_Australasia.txt	Our research outcomes have also led to impact outside of academia, underscoring the prominence of some of the topics instance, Chakkrit Tantithamthavorn’s work on developing analytics for predicting wait-time in the emergency department is currently deployed in many hospitals in Australia5.
Processed_Software_Engineering_in_Australasia.txt	Software engineering researchers across Australasia continue to participate actively in every area of research worldwide, and we are open for business. Across the spectrum of the esteem metrics, Australasian software engineering researchers continue to perform admirably, including with scientific acknowledgments and awards, editorial tasks, organising roles in academic events, PC memberships, peer review records, external refereeing (including for major grants), invited talks and lectures, research visits, and professional memberships. For instance, APSEC2016, ICSME2020, VL/HCC2020, and ASE2020 were all recently held in Australasia. ASE2021, RE2022, and ICSE2023 will be held in Melbourne, Australia. Software engineering researchers have also increasingly received funding; for instance, software engineering academics in New Zealand have received several research grant awards from the New Zealand government including a prestigious Marsden award which was received by Kelly Blincoe. Notable Science for Technological Innovation National Science Challenge (SfTI) funding was also awarded to Michael Homer, Steve Reeves, Sherlock Licorish and Amjed Tahir. Furthermore, Panos Patros is part of a team awarded $12.5M to decarbonize the New Zealand Industrial Sector via Self-Adaptive Software.
Processed_Software_Engineering_in_Australasia.txt	Much of the recent software engineering research funding success in New Zealand may be attributed to the formation of Software Innovation New Zealand (SI^NZ) in October 2016. The goal of this committee is to bring together leading software engineering research teams from New Zealand universities in a coordinated centre of research excellence with a strong industry focus and raise the level and profile of New Zealand software engineering research. SI^NZ has also facilitated increased collaboration among software engineering researchers in New Zealand.
Processed_Software_Engineering_in_Australasia.txt	In Australia, Professor John Grundy has been awarded an ARC Laureate Fellowship, one of the most prestigious research awards in Australia. Seven early-career academics received an Australian Research Council (ARC) Discovery Early Career Researchers Award (DECRA) over the last few years (i.e., Aldeida Aleti, Christoph Treude, Marcel Boehme, Li Li, Xin Xia, Chakkrit Tantithamthavorn and Patanamon Thongtanunam).
Processed_Software_Engineering_in_Australasia.txt	Notwithstanding the COVID-19 pandemic and the barriers to travel recently, we invite the world to travel to Australasia for research study leave to facilitate more exchanges. You may contact any of the authors of this article or others in the wider Australasian software engineering community about visiting. In addition, there is an open call for support for Australasia’s bids for hosting conferences, and also for colleagues to travel to the region for conference attendance and to forge partnerships. In the short term, we warmly look forward to hosting you for ASE20217.
Processed_Software_Engineering_in_Australasia.txt	We would like to thank ACM SIGSOFT and the instigators of the Software Engineering Worldwide column, especially Professors Marco Kuhrmann and Dietmar Pfahl, for the invitation to contribute to this article series.
Processed_Hiring_from_a_pool_of_workers.txt	Abstract. In many countries and institutions around the world, the hiring of work- ers is made through open competitions. In them, candidates take tests and are ranked based on scores in exams and other predetermined criteria. Those who satisfy some eligibility criteria are made available for hiring from a “pool of workers.” In each of an ex-ante unknown number of rounds, vacancies are announced, and workers are then hired from that pool. When the scores are the only criterion for selection, the procedure satisﬁes desired fairness and independence properties. We show that when aﬃrmative action policies are introduced, the established methods of reserves and pro- cedures used in Brazil, France, and Australia, fail to satisfy those properties. We then present a new rule, which we show to be the unique rule that extends static notions of fairness to problems with multiple rounds while satisfying aggregation independence, a consistency requirement. Finally, we show that if multiple institutions hire workers from a single pool, even minor consistency requirements are incompatible with varia- tions in the institutions’ rules. JEL Classiﬁcation: C78, J45, L38, D73 Keywords: public organizations, hiring, aﬃrmative action.
Processed_Hiring_from_a_pool_of_workers.txt	Abizada: School of Business, ADA University, 11 Ahmadbay Aghaoglu St., Baku AZ1008, Azerbaijan. Email:[email protected]. Bó (Corresponding author): University of York, Department of Economics and Re- lated Studies, York, United Kingdom; website: http://www.inaciobo.com; e-mail: ina- [email protected]. Date: December 2020. The authors thank Oguz Afacan, Samson Alva, Julien Combe, Janaína Gonçalves, Rosalia Greco, Rus- tamdjan Hakimov, Philipp Heller, Sinan Karadayi, Morimitso Kurino, Alexander Nesterov, Thibaud Pierrot, Bertan Turhan, Chiu Yu Ko, and two anonymous referees for helpful comments. Research as- sistance by Florian Wiek is much appreciated. Financial support from the Leibniz Association through the SAW project MADEP (Bó) is gratefully acknowledged.
Processed_Hiring_from_a_pool_of_workers.txt	Notice that the quote above refers to a “reserve list”. A reserve list is a group of candidates who are not hired initially but may (or may not) be hired later. Procedures that mention reserve lists are equivalent to those with a pool of workers: the pool consists of the ﬁrst set of workers hired together with the reserve list.
Processed_Hiring_from_a_pool_of_workers.txt	A vast number of hirings occur around the world following this type of procedure. Most developed countries use them when hiring public sector workers to the best of our knowledge. Below we provide three examples, which are informative about the number of jobs involved.
Processed_Hiring_from_a_pool_of_workers.txt	level, etc.8 When the workers’ scores constitutes the sole element for determining which workers to hire, a very natural rule, namely sequential priority, is commonly used: if q workers are to be hired, hire the q workers with the highest scores among those who remain in the pool. This rule is simple but has many desirable characteristics. First, it is fair in the sense that every worker who is not (yet) hired has a lower score than those who were hired. This adds a vital element of transparency to the process: if the worker can see, as is often the case, the scores of those who were hired (or at least the lowest score among those who were hired), then she has a clear understanding of why she was not hired. Secondly, it responds to the agency problem: an institution cannot arbitrarily select low-scoring workers before selecting all those who have a score higher than that worker. Finally, the selected workers’ quality and identity do not depend on the number of rounds and vacancies in each round. That is, selecting 20 workers in four rounds with ﬁve workers in each results in selecting the same workers as if 20 workers were selected at once. We denote this last property by aggregation independence. One implication of this requirement is that the set of selected workers is independent of the number of rounds and vacancies in each round: selecting 10 workers in two rounds of ﬁve workers in each results in the same selection as selecting two workers in each of ﬁve rounds.
Processed_Hiring_from_a_pool_of_workers.txt	While sequential priority satisﬁes those desirable properties, the criteria used for hiring workers often combine scores with other compositional objectives, in the form of a desired proportion of workers belonging to some subset of the population, such as ethnic minorities, people with disabilities, or women. In section 4 we formalize these objectives, noting that these cannot be achieved by the sequential priority rule even if scores are designed to incorporate them, and show that minority reserves, which is arguably the best method for implementing these objectives in static problems, does not satisfy desirable properties when used multiple times over a single pool of workers. In section 5 we present our main contribution, which is a new rule for hiring workers, that is the unique rule that satisﬁes natural concepts of fairness for this family of problems. We also show that it is essentially the only rule that extends static notions of fairness with compositional objectives to a problem with sequential hirings while being aggregation independent.
Processed_Hiring_from_a_pool_of_workers.txt	In section 6, we evaluate rules used in real-life applications in diﬀerent parts of the world, combining scores with compositional objectives. These include “quotas” for individuals with physical or mental disabilities in public sector jobs in France, for black workers in public sector workers in Brazil, and the gender-balanced hiring of ﬁreﬁghters in the Australian province of New South Wales. We show that these fail most of the time to satisfy natural concepts of fairness and aggregation independence.
Processed_Hiring_from_a_pool_of_workers.txt	Finally, in section 8 we consider the cases where there are multiple institutions (or locations, departments, etc.) hiring from a single pool of workers. While this scenario is widespread, our main result shows that a mild requirement, saying that the order in which ﬁrms hire workers should not change whether some of the institutions hire a worker, essentially leaves us with a single rule, which says that all institutions must hire workers following a single common priority over them.
Processed_Hiring_from_a_pool_of_workers.txt	Other than the sections described above, the rest of the paper is organized as follows. In section 2 we introduce the basic model of hiring by rules and justify the desirability of aggregation independence. In section 3 we restrict our focus to rules that are based on scores associated with each worker, in section 7 we evaluate the properties of the rules evaluated when they are used for a single round of hiring, and in section 9 we conclude. Proofs and formal descriptions of the rules absent from the main text are found in the appendix.
Processed_Hiring_from_a_pool_of_workers.txt	There are a few branches of the literature, however, that are related to our analysis. First, the description and analysis of methods for hiring public sector workers around the world and the incentives involved. Sundell (2014) evaluates to what extent the use of examinations constitutes a meritocratic method for recruiting in the public sector. The author observes that exams may not be the most adequate way to identify ﬁtness for each function, but that the patronage risk involved when using more subjective criteria such as interviews and CV screening often overcomes those losses. In fact, in an empirical analysis in diﬀerent ministries in the Brazilian federal government, Bugarin and Meneguin (2016) found a positive relationship between corruption cases and the proportion of employees hired by using subjective criteria.
Processed_Hiring_from_a_pool_of_workers.txt	notions of consistency have been used in other matching and allocation problems based on priorities as well (Ergin, 2002; Klaus and Klijn, 2013).
Processed_Hiring_from_a_pool_of_workers.txt	Finally, a big part of our analysis concerns what we denote compositional objectives: objectives regarding characteristics that some portions of the workers hired should have, such as a minimum proportion of workers with disabilities, ethnic minorities, or certain genders. Sönmez et al. (2019) evaluated the constitutionally mandated aﬃrmative ac- tion policy used in the hiring of public sector workers in India. Similar to the cases we study, workers are also selected based on open competitions that result in an order of merit of the candidates. They also identify some shortcomings that result from how the rules are used to implement aﬃrmative action objectives and propose solutions for them. However, they do not consider the cases in which hirings occur in multiple rounds,9 which, as we show in this paper, might have signiﬁcant consequences.
Processed_Hiring_from_a_pool_of_workers.txt	Most of the positive and normative literature on the market design consequences of aﬃrmative action policies focus on college admissions and school choice. Kojima (2012) and Hafalir et al. (2013) evaluate the use of maximum quotas (which limit the number of non-minority students who can be admitted in a school) with minority reserves in the context of a centralized school choice procedure. They show that majority quotas may paradoxically hurt minority students, while minority reserves improve upon this problem. However, these welfare results have no parallel in our analysis, in which workers are either hired or not. Several other papers also evaluate aﬃrmative action procedures currently used to select students into schools or universities, identifying shortcomings and proposing alternative procedures. Aygun and Bó (Forthcoming) show how the aﬃrmative action procedure used in university admissions in Brazil results in fairness and incentive problems. Dur et al. (2020) studied the allocation of students to Chicago’s elite public high schools and compared various reservation policies.10 All of them, however, treat the problem from a static perspective: either only one choice is made, or a complete allocation is produced once and for all.
Processed_Hiring_from_a_pool_of_workers.txt	A rule determines which workers an institution should hire, given a number of work- ers to hire, a pool of workers, and, potentially, the workers that the institution hired before. Each time an institution attempts to hire workers from the pool is denoted as a round.
Processed_Hiring_from_a_pool_of_workers.txt	Where A0 = ∅, W 0 = W and for i > 0, Ai = Ai−1 ∪ ϕ (W i−1, Ai−1, qi) and W i = W \Ai. Furthermore, for simplicity, we will use the shorter notation ϕ (W, q) when A = ∅. Unless stated explicitly, none of our results rely on situations in which there are not enough workers, either in general or with some characteristics, to be hired. That is, in all of our results, we will assume that the number of workers in W is at least as qi , and the same holds for the cases that we will evaluate in which some large as workers belong to minority groups.
Processed_Hiring_from_a_pool_of_workers.txt	One crucial property of the process of hiring by rules is that the sequence of hires λ = hq1, . . . , qti is ex-ante unknown. That is, every round may or may not be the last one. The total number of workers who will be hired is also unknown. Therefore, the In this properties that we will deem as desirable should hold at any point in time. context, a critical property that a rule should satisfy is aggregation independence. A rule is aggregation independent if the total set of workers hired after a certain number of rounds does not depend on how they are distributed among rounds.
Processed_Hiring_from_a_pool_of_workers.txt	Deﬁnition 1. A rule ϕ is aggregation independent if for any q ≥ q1 ≥ 0 and sets of workers W and A, ϕ (W, A, q) = ϕ (W, A, hq1, q − q1i).
Processed_Hiring_from_a_pool_of_workers.txt	Therefore, when the rule being used is aggregation independent, an institution that hires q1 workers in the ﬁrst round and q2 in the second will select the same workers that it would by hiring q1 + q2 in a single round. One can easily check that if a rule is j q′ aggregation independent, this extends to any combination of rounds: if j, a sequence of hires q1, . . . , qn will select the same workers as q′ m.
Processed_Hiring_from_a_pool_of_workers.txt	We now provide three reasons to justify aggregation independence as a strongly desired property for rules for hiring by rules: transparency, non-manipulability, and robustness.
Processed_Hiring_from_a_pool_of_workers.txt	Transparency One of the main reasons driving governments and institutions to use hiring by rules is that, for those who are not hired, the reason that happens is made clear and straight- forward. For example, take the rule that consists of always hiring the workers with the highest exam scores. By knowing the rule and observing the hired workers (and their scores), any worker who was not hired knows that there was no obscure reason why she was not yet hired: it is merely because her score was lower than all those who were hired.
Processed_Hiring_from_a_pool_of_workers.txt	not be able to easily understand why she was not hired, even understanding the rule that was used, because it would also be necessary for her to know the precise sequence of the number of workers that were hired in each round.
Processed_Hiring_from_a_pool_of_workers.txt	Non-manipulability While many times the rules which govern the hiring process are chosen in a way that reduces the ability of managers to make arbitrary choices of whom to hire, they may have freedom in choosing the sequence of hires. For example, instead of hiring four workers in one month, she may choose to hire two workers ﬁrst and then two additional workers.
Processed_Hiring_from_a_pool_of_workers.txt	If the rule is aggregation independent, diﬀerent choices of sequences of workers hired will not lead to diﬀerent sets of workers hired. However, if the rule is not aggregation independent, that may not be the case, and a manager may choose a speciﬁc sequence of hiring decisions, which will allow a particular worker to be hired, whereas she would not be, absent the speciﬁc sequence chosen. An aggregation independent rule, by deﬁnition, is not manipulable by the choice of the sequence of hires.
Processed_Hiring_from_a_pool_of_workers.txt	Robustness The third reason why aggregation independence is a desirable property is that the degree to which the set of workers hired satisﬁes the objectives represented in the rule is robust to uncertainty or bad planning on the part of the manager in terms of the number of workers that is needed. In other words, assuming that the criterion for choosing workers which is set by the rule represents the desirability of the workers it chooses (for example, it chooses the most qualiﬁed set of workers subject to some constraint), an aggregation independent rule will always choose the best set of workers, whether the manager makes hiring decisions all at once or continually re-evaluates the number of workers to be hired. Aggregation independent rules do not have that problem: managers may hire workers based on demand, and that will not result in a less desirable set of workers hired.
Processed_Hiring_from_a_pool_of_workers.txt	workers with the highest scores in W .12 A natural property for a score-based rule is for it to be fair. That is, after any number of rounds, if a worker w was hired and w′ was not, then sw > sw′.
Processed_Hiring_from_a_pool_of_workers.txt	Deﬁnition 2. A rule ϕ is fair if for any W , A and λ = hq1, . . . , qti, w ∈ ϕ (W, A, λ) and w′ 6∈ ϕ (W, A, λ) implies that sw > sw′.
Processed_Hiring_from_a_pool_of_workers.txt	A natural rule for these kinds of problems is what we denote by sequential priority. When hiring q workers, it consists of selecting the q workers with the highest score from the pool of workers. If the pool contains less than q workers, then hire all of them. The following remark comes immediately from the deﬁnition of the rule.
Processed_Hiring_from_a_pool_of_workers.txt	It is common for hiring processes based on rules to combine the use of scores with com- positional objectives, such as aﬃrmative action. Typically, the objective is to reserve some of the jobs for workers with a certain characteristic, sometimes those belonging to an ethnic minority or those who possess some type of disability. Denote by M the set of workers who belong to the minority group (that is, M ⊆ W ) and ω (W ) be the number of minority workers in W . The aﬃrmative action policy also has a minority ratio m, where 0 ≤ m ≤ 1, which represents the proportion of hires that should be based on aﬃrmative action.
Processed_Hiring_from_a_pool_of_workers.txt	As argued in Section 2, the desirable properties associated with aﬃrmative action should also hold after any number of rounds. Our ﬁrst requirement is that, when possible, the proportion of selected minorities should be at least m after each round.
Processed_Hiring_from_a_pool_of_workers.txt	Assume that there are three workers: one minority (call him K) and two non-minority (L and V), where the scores are as follows sL > sV > sK . If the rule needs to select two workers and m is 0.3, then in order to respect minority rights, the rule should select K and L, which is not fair, as fairness requires L and V to be selected.
Processed_Hiring_from_a_pool_of_workers.txt	Therefore, we deﬁne a weaker notion of fairness, which takes into account the minority restriction. A rule is minority fair if, conditional on respecting minority rights, the hiring decision is based on scores.
Processed_Hiring_from_a_pool_of_workers.txt	In words, a rule is minority fair if it (i) chooses between workers from the same group (minorities or non-minorities) based on their scores, (ii) does not hire low-scoring non- minorities while higher-scoring minorities are available, and (iii) only hires low-scoring minorities over higher-scoring non-minorities when that is necessary to bring the ratio of minorities closer to m from below.
Processed_Hiring_from_a_pool_of_workers.txt	One natural question one might have is whether a carefully designed “standardized” scoring system could turn the sequential priority rule into one that respects minority rights and/or is minority fair. W of the scores sW , for which the sequential priority rule under s′ W respects minority rights and is minority fair under sW ? The answer for the ﬁrst part is yes, and for the second, no. To see this, let s be the highest score any worker might have, and s′ W be a scoring system that replicates sW for non-minorities, and that adds s to the scores of minorities. That is, s′ W makes all minorities have higher scores than non-minorities but keep all relative scores the same otherwise. It is clear that the sequential priority rule under this scoring system respects minority rights but might not be minority fair since it will always hire minority candidates whenever some are still available, regardless of the value of m.
Processed_Hiring_from_a_pool_of_workers.txt	Proposition 1. The sequential use of minority reserves respects minority rights. How- ever, it is neither minority fair nor aggregation independent.
Processed_Hiring_from_a_pool_of_workers.txt	Example 1. Let W = {w1, w2, w3, w4, w5}, M = {w1, w2, w5}, and sW = (100, 90, 80, 50, 20). Let r = 2, q1 = q2 = 2 and m = 0.5. In the ﬁrst round, the top worker from M and the top from W \ {w1} are hired, that is, {w1, w2}. In the second round, the pools of remaining workers are W 2 = {w3, w4, w5} and M 2 = {w5}. The top worker from M 2 , that is, {w5}, and the top from W 2\ {w5} are hired. Therefore, {w3, w5} are hired in the second round and ϕSM (W, hq1, q2i) = {w1, w2, w3, w5}.
Processed_Hiring_from_a_pool_of_workers.txt	Now consider the case where q = q1 + q2 = 4. Then in the ﬁrst and unique round, the two top workers from M, {w1, w2}, and the top two workers from W \ {w1, w2} are hired, that is, {w3, w4}. Therefore, ϕSM (W, q1 + q2) = {w1, w2, w3, w4}. Therefore, the ϕSM rule is not aggregation independent. Moreover, note that w4 ∈ ϕSM (W, hq1, q2i), w5 ∈ ϕSM (W, hq1, q2i) and sw4 = 50 > 20 = sw5 while ω ϕSM (W, hq1, q2i) 0.75 > 0.5 = m, implying that the ϕSM rule is not minority fair.
Processed_Hiring_from_a_pool_of_workers.txt	The sequential use of minority reserves rule is a good rule for providing examples of problems associated with rules that are not aggregation independent. First, consider the issue of manipulability. Take the example 1 above and suppose that the manager prefers to hire the worker w5. If she hires the four workers that she needs all at once, w5 would not be hired. If, instead, she chooses to hire ﬁrst two workers, and then later two more workers, w5 will be hired. That is, by choosing a sequence of hires strategically, the manager can hire the person she wanted.
Processed_Hiring_from_a_pool_of_workers.txt	considered is more of a technical property that results from aggregation independence than a criterion that can be described as a scoring method for hiring before knowing who will apply for the jobs.
Processed_Hiring_from_a_pool_of_workers.txt	Example 2. Let W = {w1, w2, w3, w4, w5, w6, w7, w8}, M = {w5, w6, w7, w8}, sW = (100, 90, 80, 70, 60, 50, 40, 30) and m = 0.5.
Processed_Hiring_from_a_pool_of_workers.txt	Assuming that the scores are a good representation of the degree of desirability of a worker for a task, the example above shows that a lack of planning could lead to hiring a set of workers that are substantially less qualiﬁed.
Processed_Hiring_from_a_pool_of_workers.txt	Therefore, the sequential adjusted minority reserves adapts the set of workers hired according to those hired in previous rounds. This makes sense: if we do not take into account, for example, that after the last round, the number of minority workers greatly exceeded the minimum required, some high-scoring non-minority workers may not be hired, leading to a violation of minority fairness. The theorem below shows that this is essentially the only way of achieving these objectives.
Processed_Hiring_from_a_pool_of_workers.txt	Theorem 1. If ϕ is a rule that is minority fair and respects minority rights, then for every set of workers W and sequence of hires λ, ϕ(W, λ) = ϕSA(W, λ).
Processed_Hiring_from_a_pool_of_workers.txt	For simplicity, the description below assumes that the number of workers in M and W is large enough so that in every round there is a suﬃcient number of them to be hired. A more general description can be found in the appendix.
Processed_Hiring_from_a_pool_of_workers.txt	The sequential adjusted minority reserves is also the only rule that extends “static” notions of fairness to this dynamic setting while being aggregation independent. To see that, we ﬁrst deﬁne the static counterparts of deﬁnitions 3 and 4.
Processed_Hiring_from_a_pool_of_workers.txt	Deﬁnition 5. A rule ϕ respects static minority rights if, for any W and q > 0, (i) when \|M\| ≥ m × q we have ω (ϕ (W, hqi)) / \|ϕ (W, hqi)\| ≥ m, or (ii) when \|M\| < m × q we have M ⊂ ϕ (W, hqi).
Processed_Hiring_from_a_pool_of_workers.txt	In other words, a rule satisﬁes the static versions of these two notions if they hold when there is only one round of hiring. As a result, a rule that respects minority rights also respects static minority rights, and a rule that is minority fair is also static minority fair. However, the converse does not apply: deﬁnitions 5 and 6 restrict only the ﬁrst set of workers hired in a sequence of hires.
Processed_Hiring_from_a_pool_of_workers.txt	The following result shows that the sequential adjusted minority reserves is essen- tially16 the only rule that extends these static notions to multiple hires while being aggregation independent.
Processed_Hiring_from_a_pool_of_workers.txt	Theorem 2. If ϕ is a rule that respects static minority rights, is static minority fair, and aggregation independent, then for every set of workers W and sequence of hires λ, ϕ(W, λ) = ϕSA(W, λ).
Processed_Hiring_from_a_pool_of_workers.txt	Moreover, notice that the ﬁrst round of hiring in the sequential adjusted minority reserves is equivalent to the one in the sequential use of minority reserves. Therefore, the sequential adjusted minority reserves is also the aggregation independent extension of the static minority reserves rule (Hafalir et al., 2013).
Processed_Hiring_from_a_pool_of_workers.txt	Corollary 1. If ϕ is a rule for which ϕ(W, hqi) = ϕSM (W, hqi) and ϕ is aggregation independent, then for every set of workers W and sequence of hires λ, ϕ(W, λ) = ϕSA(W, λ).
Processed_Hiring_from_a_pool_of_workers.txt	Since the deﬁnition of a rule includes an arbitrary set of previous hires, there can be more rules that can satisfy those properties by deﬁning them diﬀerently, for example, for sets of previous hires that could not result from using the rule from the beginning. If we restrict ourselves to the case in which A = ∅ (the rule is used for every hire ever made), this is a uniqueness result.
Processed_Hiring_from_a_pool_of_workers.txt	The French law also establishes that at least 6% of the vacancies should be ﬁlled by Instead of incorporating the selection of people with physical or mental disabilities. those workers into the hiring procedure in a uniﬁed framework, the institutions instead open, with unclear regularity, vacancies exclusive for workers who have those disabili- ties.17 The hiring of workers over time continues following the same procedure as the open positions described above. However, nothing prevents workers with disabilities from applying for open positions. In fact, the authorities provide some accommodation for these workers during the selection, such as, for example, allowing for extra time to write down the exams. These are meant as an attempt to make up for some disadvan- tages that those workers have with respect to those without disabilities, and not to give any advantage.
Processed_Hiring_from_a_pool_of_workers.txt	Let W ∗ be the set of workers who applied for the open competitions, and M ∗ those who applied for the competitions reserved for candidates with disabilities. Workers in M ∗ must have a disability, but workers with disabilities might also apply for open competitions. Therefore, M ∗ ⊆ M, and in general M ∗ ∩ W ∗ 6= ∅. Since these constitute diﬀerent competitions, there are scores for the workers in each competition, and more speciﬁcally, a worker who applies to both competitions might obtain diﬀerent scores in each. Therefore, we denote by sO w the score obtained by worker w in the open competition and by sD w the score that worker w obtained in the competition for workers with disabilities. Since W ∗ and M ∗ are usually diﬀerent sets of workers, some workers might have a score in only one competition, but some workers with disabilities might have in both.
Processed_Hiring_from_a_pool_of_workers.txt	In some cases, diﬀerent procedures are also used, such as making candidates with disabilities compete for the same vacancies as those without disabilities, but giving them a “bonus” in their scores. Similar methods are used in aﬃrmative action policies elsewhere, such as in local universities in Brazil. In this paper, however, we focus on examples involving quotas and reserved vacancies.
Processed_Hiring_from_a_pool_of_workers.txt	will consider two alternative policies.18 In both cases, we will assume that there are two pools of workers, W ∗ and M ∗, and given a number of positions to be hired q, a total of q workers from these pools must be hired. In what follows, we consider an arbitrary sequence of hires hq1, q2, . . . , i.
Processed_Hiring_from_a_pool_of_workers.txt	Policy 1: This policy consists of ﬁrst hiring the top q1 workers from W ∗ and then adjusting the number of workers in M ∗ hired in later rounds. For example, say that q1 = 100, but only four workers among the top 100 workers in W ∗ (with respect to sO W ) hired have disabilities. Then, considering the objective of hiring at least 6% workers with disabilities, if q2 = 50, then open ﬁve vacancies exclusive for workers in M ∗ (selected with respect to sD W ). As a result, by the end of the second round, at least 9 workers with disabilities, or m × (q1 + q2),19 will be hired.
Processed_Hiring_from_a_pool_of_workers.txt	Policy 2: This policy consists of ﬁrst hiring m×q1 from M ∗ (selected with respect to sD W ), (1 −m) ×q1 workers from W ∗ (selected with respect to sO W ) and then adjusting the number of workers in M ∗ hired in later rounds. For example, say that q1 = 100. Then the policy will result in hiring six workers from M ∗ and 94 from W ∗ in the ﬁrst round. At least 6% of the workers hired would be among those with a disability, therefore, but potentially more. Suppose that eight workers with disabilities were hired in the ﬁrst round and that q2 = 50. Then in the second round, two vacancies exclusive for workers in M ∗ would be open, and the remaining 48 would be open for all workers in W ∗.
Processed_Hiring_from_a_pool_of_workers.txt	While the two policies above represent what we believe are the best eﬀorts to satisfy the objectives stated in the law under the current existing procedures, Policy 1 diﬀers in that when only one round of hiring is done, the proportion of workers with a disability hired might be below the minimal proportion stated in the law. This fact will more evident in Proposition 2.
Processed_Hiring_from_a_pool_of_workers.txt	Whenever necessary, we will refer to the rules deﬁned by policies 1 and 2 by ϕF1 and ϕF2. Since under the French assignment rule each worker may have one or two scores, what constitutes minority fairness is less clear in this context. However, the example below shows that policy 2 may lead to outcomes that clearly violate the spirit of minority fairness.
Processed_Hiring_from_a_pool_of_workers.txt	Example 3. Let W ∗ = {w3, w4, w5} and M ∗ = {w1, w2, w3}, with scores sO W = (50, 40, 30) and sD If q = 2, then ϕF2 ({W ∗, M ∗} , q) will select {w1, w3}. Worker w2, however, has a disability and a better score than w3 in the competition in which both participated.
Processed_Hiring_from_a_pool_of_workers.txt	For simplicity, here and in the rest of the main text we will assume that every expression involving numbers of workers or vacancies are integers. In the appendix, we relax that restriction and show that none of the results presented depend on that.
Processed_Hiring_from_a_pool_of_workers.txt	Notice that if worker w2 also applied for the open vacancies and in that competition obtained a score that is also better than the one obtained by w3, she would have been hired instead of w3. If the relative rankings of the workers in both competitions are diﬀerent, more intricate violations of the spirit of minority fairness can also occur. If we make the (strong) assumptions that all workers with disabilities apply to both competitions and that the relative rankings between those workers in both competitions are the same, we can obtain a clear distinction between both policies, as shown below.
Processed_Hiring_from_a_pool_of_workers.txt	Proposition 2. Suppose that M ∗ ⊆ W ∗ and that for every w, w′ ∈ M ∗, sO w′ ⇐⇒ sD w > sD w′. Policy 1 of the French assignment rule does not respect minority rights and is not aggregation independent. Policy 2 respects minority rights, is aggregation independent, and minority fair.
Processed_Hiring_from_a_pool_of_workers.txt	It is crucial to notice, however, that the result in proposition 2 depends on the relative rankings of the workers with disabilities being the same in both competitions, but perhaps most importantly, on workers with disabilities partic- ipating in both competitions. This is not a minor issue, since these competitions often involve a signiﬁcant amount of time and eﬀort.
Processed_Hiring_from_a_pool_of_workers.txt	Diﬀerently from France, the quotas for black workers are explicitly incorporated into the hiring process. More speciﬁcally, the rule currently used in Brazil (denoted the ϕB rule) works as follows. Let k be a number that is higher than any expected number of hires to be made.
Processed_Hiring_from_a_pool_of_workers.txt	from M, and O be the top k−⌈m×k⌉ workers in W \T M.21 Let T M 1 = T M and O1 = O.
Processed_Hiring_from_a_pool_of_workers.txt	Round r ≥ 1 The ⌈m × qr⌉ top scoring minority workers from T M r, and the top q − ⌈m × q⌉ workers from Or are hired. By removing these workers hired, we obtain T M r+1 and Or+1.
Processed_Hiring_from_a_pool_of_workers.txt	For the Brazilian law speciﬁcally, m = 0.2. The example below shows that the Brazilian rule is not minority fair.

Subsets and Splits

No saved queries yet

Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.