Model Of Statistical Discrimination Research Articles

University Admission: Is Achievement a Sufficient Criterion?

Abstract We analyse university admissions using a statistical discrimination model where students differ by ability and social group. In this university system, candidates are evaluated on the basis of their expected human capital, which includes both their innate abilities and the knowledge acquired during their schooling. Consequently, students determine their study effort based on the behaviour of universities. Interestingly, we find that students from a less advantaged group need a lower grade to gain admission to the best universities. If a university cannot discriminate between social groups, all students with the same grade will attend universities of the same quality, but with different levels of human capital.

Discrimination against gay and transgender people in Latin America: A correspondence study in the rental housing market

We assess the extent of discrimination against gay and transgender individuals in the rental housing markets of four Latin American countries. We conducted a large-scale field experiment building on the correspondence study methodology to examine interactions between property managers and fictitious couples engaged in searches on a major online rental housing platform. We find no evidence of discrimination against gay male couples, but we do find strong discrimination against heterosexual couples with a transgender woman partner (trans couples). Relative to heterosexual couples, trans couples receive 19% fewer responses, 27% fewer positive responses, and 23% fewer invitations to showings. We also assess whether the evidence is consistent with taste-based discrimination or statistical discrimination models by comparing response rates when couples signal being professionals with stable jobs (quality-job signal). While we find no significant effect of the signal for heterosexual or gay male couples, trans couples benefit from providing a strong labor market signal. Their call-back, positive-response, and invitation rates increase by 25%, 36% and 29%, respectively. These results suggest that discrimination against trans couples is consistent with statistical discrimination. Moreover, we find no evidence of heterosexual couples being favored over gay male couples, nor evidence of statistical discrimination for gay male or heterosexual couples.

Random and block architectures of N-arylitaconimide monomers with methyl methacrylate

Abstract “Itaconimide” is the members of imide (–CO–NH–CO–) family with reactive exocyclic double bond and it is easily obtained from the renewable resource i.e. D-glucose. The polymerization of various N-arylitaconimide (NAI) monomers with methyl methacrylate (MMA) have been reported to improve the glass transition temperature (T g) and thermal stability of poly(methyl methacrylate) (PMMA). In literature, these studies have been done mostly using conventional free radical polymerization methods, which restricts the architecture of copolymers to “random” only. The block copolymers of NAI and MMA are an important due to the combination of glassy PMMA and thermally stable poly(NAI), which offers its applications for higher temperature service. The architectural control of polymers in provisions of its topology, composition, and various functionalities is possibly obtained using reversible-deactivation radical polymerizations (RDRPs). In RDRPs, the concentration of free radical is controlled in such a way that the termination reactions are minimized (normally in range of 1–10 mol%), and not allowed to obstruct with the desired architecture. However, this is possible by achieving (or by establishing) a rapid dynamic equilibrium between propagating radical and dormant species (i.e. R–X). Among all RDRPs, the atom transfer radical polymerization (ATRP) is very popular and adaptable method for the synthesis of polymers with specifically controlled architecture. Two different architectures of NAI and MMA copolymers are reported using ATRP process. The effect of various pedant groups on the rate constants of propagation (k p) and thermal properties NAI and MMA copolymers is studied. The poly(NAI-ran-MMA)-b-poly(MMA) are stable up to 200 °C and degraded in three steps. Whereas, the poly(NAI-ran-MMA)-b-poly(NAI) are stable up to 330 °C and degraded in two steps. The density functional theory methods are used for calculation of equilibrium constants (K ATRP) for the ATRP process for the series of laboratory synthesized alkyl halides. A good agreement was observed between the experimentally determined and theoretically calculated K ATRP values. The mechanistic studies are carried for poly(NAI-ran-MMA) copolymer system using statistical model discrimination method along with 1H decoupled 13C NMR spectroscopy. For studying the mechanism of copolymerization of NAI and MMA via ATRP methods, “trimer model or penultimate model” will be more accurate than “dimer model or terminal model”.

Non-Bayesian Statistical Discrimination

Models of statistical discrimination typically assume that employers make rational inference from (education) signals. However, there is a large amount of evidence showing that most people do not update their beliefs rationally. We use a model and two experiments to show that employers who are conservative, in the sense of signal neglect, discriminate more against disadvantaged groups than Bayesian employers. We find that such non-Bayesian statistical discrimination deters high-ability workers from disadvantaged groups from pursuing education, further exacerbating initial group inequalities. Excess discrimination caused by employer conservatism is especially important when signals are very informative. Out of the overall hiring gap in our data, around 40% can be attributed to Bayesian statistical discrimination, a further 40% is due to non-Bayesian statistical discrimination, and the remaining 20% is unexplained or potentially taste-based. This paper was accepted by Marie Claire Villeval, behavioral economics and decision analysis. Funding: F. Mengel thanks the European Research Council for financial support [Starting Grant 805017]. P. Campos-Mercade acknowledges funding from the Danish National Research Foundation [Grant DNRF134 (CEBI)]. Supplemental Material: The online appendix and data are available at https://doi.org/10.1287/mnsc.2023.4824 .

Using the Discriminant Analysis to Classify the Income of Households in Sinnar State, Sudan (2021)

The purpose of this study was to distinguish between sufficient and insufficient income and to identify the most discriminating factors that influence income. The data was obtained from households in Sinnar through a structured questionnaire addressed to the heads of families, a sample of (800) households (417) had sufficient incomes, and (383) had insufficient incomes. Discriminate analysis and decision trees were applied with the help of the (SPSS) program. The results suggested that the discrimination model applied had a good fit with the data obtained from the sample and that 7 of the 24 variables used in discrimination were statistically significant. The most important discriminating variables were the evaluation of the standard of living and borrowing to cover the family's living expenses. The research showed that the possible error in discriminate function model specificity does not exceed 14.2% compared to decision trees where the possible error does not exceed 14.5%. The research study recommended the use of a statistical discrimination model to discriminate between a sufficient income and insufficient income and the use of decision trees to classify the administrative unit of Sinnar according to income.

Estimating ancient biogeographic patterns with statistical model discrimination.

The geographic ranges in which species live is a function of many factors underlying ecological and evolutionary contingencies. Observing the geographic range of an individual species provides valuable information about these historical contingencies for a lineage, determining the distribution of many distantly related species in tandem provides information about large-scale constraints on evolutionary and ecological processes generally. We present a linear regression method that allows for the discrimination of various hypothetical biogeographical models for determining which landscape distributional pattern best matches data from the fossil record. The linear regression models used in the discrimination rely on geodesic distances between sampling sites (typically geologic formations) as the independent variable and three possible dependent variables: Dice/Sorensen similarity; Euclidean distance; and phylogenetic community dissimilarity. Both the similarity and distance measures are useful for full-community analyses without evolutionary information, whereas the phylogenetic community dissimilarity requires phylogenetic data. Importantly, the discrimination method uses linear regression residual error to provide relative measures of support for each biogeographical model tested, not absolute answers or p-values. When applied to a recently published dataset of Campanian pollen, we find evidence that supports two plant communities separated by a transitional zone of unknown size. A similar case study of ceratopsid dinosaurs using phylogenetic community dissimilarity provided no evidence of a biogeographical pattern, but this case study suffers from a lack of data to accurately discriminate and/or too much temporal mixing. Future research aiming to reconstruct the distribution of organisms across a landscape has a statistical-based method for determining what biogeographic distributional model best matches the available data.

Racial disparities in the small business loan market

ABSTRACT We investigate patterns of racial bias in small business loans denial rates in the US across different credit risk scores. Using data constructed from the 1998 Survey of Small Business Finances and Kauffman Firm Survey, we find disparities in loan approval ratings between Black and White entrepreneurs in intermediate risk categories, but not for the best and worst categories. We explain these findings with a simple and generalizable statistical discrimination model where banks hold prior beliefs of repayment probability based on the applicant’s group and observe noisy signals of creditworthiness. Our model predicts that differences in loan denial rates across groups are more pronounced at middle range values and disappear at very high and very low credit scores. Identifying the likely cause of differential treatment in the market is an important first step in improving representation of minority groups as entrepreneurs. Our findings contribute to this critical effort by delineating the nature of racial bias in the credit market and informing remediation policies.

Understanding Discrimination: Outcome-Relevant Information Does Not Mitigate Discrimination

Abstract People experience discrimination across a variety of domains, including at work and in dealings with public institutions, but what makes some individuals discriminate against others? Two dominant scholarly approaches—“statistical” and “taste-based”—offer different explanations. Statistical discrimination models imply that discrimination occurs because of incomplete information (informational bias), whereas taste-based discrimination models emphasize more elusive and deep-rooted cognitive biases. Adding new insights into whether discrimination is “statistical” or “taste-based,” this article examines how providing information that reduces informational bias affects discrimination. Using a preregistered survey experimental design, a representative sample of Danish residents (n = 2,024) are exposed to three unique vignettes, each involving a choice of service provider (general practitioner, babysitter, and house cleaner). Relating to gender and nativity stereotypes, we manipulate the gender of the general practitioners and the babysitters, and the country of origin of the house cleaners. Moreover, we manipulate exposure to rating cues about the service providers’ task performance, thus mitigating informational bias to some extent. Contrasting the expectations of statistical discrimination models, the performance ratings cues do not mitigate discrimination. Across all three vignettes, the participants exhibit stereotypical preferences, and the performance rating cues do not affect these discriminatory biases.

Statistical discrimination and committees

We develop a statistical discrimination model where groups of workers differ in the observability of their productivity signals by the evaluation committee. We assume that the informativeness of the productivity signals depends on the match between the potential worker and the interviewer: when both parties have similar backgrounds, the signal is likely to be more informative. Under this “homo-accuracy” bias, the group that is most represented in the evaluation committee generates more accurate signals, and, consequently, has a greater incentive to invest in human capital. This generates a discrimination trap. If one group is initially poorly evaluated (less represented into the evaluation committee), this translates into lower investment in human capital of individuals of such group, which leads to lower representation in the evaluation committee in the future, generating a persistent discrimination process. We explore this dynamic process and show that quotas may be effective to deal with this discrimination trap. We show that introducing a “temporary” quota allows to reach a steady state equilibrium with a higher welfare when groups have similar size in the population. If instead the discriminated group is underrepresented in the workers’ population (for example because of his race), restoring efficiency requires to implement a “permanent” system of quotas.

Gender-oriented statistical discrimination: Aggregate fertility, economic sector, and earnings among young Chinese workers

Previous studies usually attribute residual gender pay gap net of human capital, job characteristics, and family responsibilities to discrimination against women, without directly exploring the underlying discriminatory mechanisms that generate gender inequalities in the labor market. By applying and extending a previously developed statistical discrimination model to post-socialist urban China, we examine the association between period age-specific fertility rate (ASFR), as a proxy for the likelihood of women taking maternity leave, and the earnings of young working women and men, as well as its variation across economic sectors. Using data compiled from a random sample of the 2005 one-percent mini-census of the People’s Republic of China, our analyses reveal gender-asymmetric associations between ASFR and earnings among young workers before their prime fertility age, where ASFR on average was negatively related to the earnings among young women, but not among young men. Moreover, such negative association was limited primarily to female workers in the private sector, negligible in state-owned and collective enterprises, and even reversed to positive among those working in government agencies and public institutions. Our findings suggest that employers in the private sector may use aggregate fertility to “evaluate” potential training and replacement costs and the productivity of female employees associated with maternity leave, thereby reinforcing and perpetuating stereotypes and discrimination against women in the workplace.

A Model of Occupational Licensing and Statistical Discrimination

We develop a model of statistical discrimination in occupational licensing with endogenous occupation selection and wage determination. We find a unique equilibrium with sharp comparative statics. Our key theoretical result is that the licensing premium is higher for workers who are members of demographic groups that face a higher cost of licensing. The predictions of the model can explain, for example, the empirical finding in the literature that occupational licenses that preclude felons close the racial wage gap among men by conferring a higher premium to Black men than to White men (Blair and Chung 2018).

Can the signaling game serve as a model of statistical discrimination in hiring

Can the signaling game serve as a model of statistical discrimination in hiring

A Model of Occupational Licensing and Statistical Discrimination

A Model of Occupational Licensing and Statistical Discrimination

Irrational Statistical Discrimination

Models of statistical discrimination typically assume that employers make rational inference from (education) signals. However, there is a large amount of evidence showing that most people do not update their beliefs rationally. We use a model and two experiments to show that employers who are conservative, in the sense of signal neglect, discriminate more against disadvantaged groups than Bayesian employers. We find that such non-Bayesian or ``irrational'' statistical discrimination deters high-ability workers from disadvantaged groups from pursuing education, further exacerbating initial group inequalities. Excess discrimination caused by employer conservatism is especially important when signals are very informative. Out of the overall hiring gap in our data, around 40\% can be attributed to rational statistical discrimination, a further 40\% is due to irrational statistical discrimination, and the remaining 20\% is unexplained or potentially taste-based.

Why are there Racial Disparities in the Small Business Loan Market

We investigate patterns of racial bias in small business loans denial rates in the U.S. across different credit risk scores. We motivate this inquiry with a simple and generalizable statistical discrimination model where banks observe noisy signals of creditworthiness and hold prior beliefs of repayment probability based on the applicant’s group. Our model predicts that differences in approval rating across groups are more pronounced at middle range values and disappear at very high and very low credit scores. Using data constructed from the 1998 Survey of Small Business Finances and the restricted access Kauffman Firm Survey we find disparities in loan approval ratings between Black and White entrepreneurs in intermediate risk categories but not for the best and worst categories.

Is Education in China Served for Productivity Augmenting? –––An Empirical Test about Educational Signaling Hypothesis

This study applies an empirical approach based upon Employer Learning and Statistical Discrimination Model to examine education’s role in China and analyzes difference in returns of schooling in context of urban rural segmentation. An Ordinary Least Square estimation and Quantile Regression are conducted with micro data from China Family Panel Study (CFPS). Evidence support education signaling hypothesis overall, and a longer lasting signaling effect is found in rural area. A sensitivity analysis by quantile regression confirms a universality of our result in all income quartiles in China but strength of education signaling varies across earning groups.

Race Discrimination: An Economic Perspective

We review the empirical literature in economics on discrimination in the labor market and criminal justice system, focusing primarily on discrimination by race. We then discuss theoretical models of taste-based discrimination, particularly models of frictional labor markets and models of statistical discrimination, including recent work on invalid statistical discrimination. We explore and evaluate the evidence for and against these theories. Although there is substantial evidence of the existence of discrimination, little is known about the extent to which disparities are driven by discrimination. Finally, we argue that economists miss the important self-enforcing relationship between disparities and discrimination and the effect of disparities in one domain on discrimination in other domains.

Replicator Evolution of Welfare Stigma: Welfare Fraud vs. Incomplete Take-Up

There are two important problems in welfare benefit programs: the prevalence of welfare fraud, in which ineligible people receive welfare benefits, and incomplete take-up, whereby eligible poor people are reluctant to claim welfare benefits. This study investigates both of these opposing phenomena using simple replicator models of statistical discrimination and the tax-payer resentment view welfare stigma suggested by Besley and Coate (1992). We find multiple stable equilibria in the long run, one of which entails low welfare fraud and 100% incomplete take-up and the other of which entails high welfare fraud and complete take-up in either model, and, moreover, that an interior stationary equilibrium that allows for the coexistence of welfare fraud and incomplete take-up is unstable in the model of statistical discrimination view welfare stigma, but it is stable in the model of the tax-payer resentment view welfare stigma. This difference arises from the different nature of stigma cost functions in these two models.

A Model of Occupational Licensing and Statistical Discrimination

A Model of Occupational Licensing and Statistical Discrimination

A Search Model of Statistical Discrimination

We offer a search-theoretic model of statistical discrimination, in which firms treat identical groups unequally based on their occupational choices. The model admits symmetric equilibria in which the group characteristic is ignored, but also asymmetric equilibria in which a group is statistically discriminated against, even when symmetric equilibria are unique. Moreover, a robust possibility is that symmetric equilibria become unstable when the group characteristic is introduced. Unlike most previous literature, our model can justify affirmative action since it eliminates asymmetric equilibria without distorting incentives.