Model Of Statistical Discrimination Research Articles

Endogenous Learning, Persistent Employer Biases, and Discrimination

I present a statistical discrimination model of the labor market in which employers are initially uncertain about the productivity of worker groups and endogenously learn about it through their hiring. Previous hiring experiences with groups shape subsequent incentives of employers to hire from these groups again and learn more about their productivity, leading to differential learning across employers and biased beliefs about the productivity of groups. Given a market-clearing wage, optimal hiring follows a cutoff rule in posterior beliefs: employers with sufficiently negative experiences with workers from a group stop hiring from the group, preserving negative biases and leading to a negatively-skewed distribution of beliefs about their productivity. When employers have noisier initial information on the productivity of one worker group, discrimination against the group can arise and persist without productivity differentials or prior employer biases, with market competition, and with or without worker signaling or investment decisions. The model generates steady state predictions analogous to the Becker (1957) taste-based model, in a statistical framework with beliefs replacing preferences, explaining apparent prejudice as the result of incorrect statistical discrimination. The model also generates additional predictions with policy implications that contrast with traditional models of discrimination.

Do Female Executives Make a Difference? The Impact of Female Leadership on Gender Gaps and Firm Performance

Abstract We investigate the effects of female executives on gender-specific wage distributions and firm performance. Female leadership has a positive impact at the top of the female wage distribution and a negative impact at the bottom. The impact of female leadership on firm performance increases with the share of female workers. We account for the endogeneity induced by non-random executives’ gender by including firm fixed-effects, by generating controls from a two-way fixed-effects regression and by using instruments based on regional trends. The findings are consistent with a model of statistical discrimination in which female executives are better at interpreting signals of productivity from female workers. This suggests substantial costs of women under-representation among executives.

A Note on School Quality, Educational Attainment and the Wage Gap

This paper explores the linkage between school quality, educational attainment and the wage gap. In a model of statistical discrimination based on both the quality and quantity of schooling, we show that lower-quality education can, on average, lead to lower human capital accumulation if agents anticipate future labor market discrimination. Because blacks in general have less access to good-quality schools compared to whites, this link provides a novel explanation for the differences in black–white educational attainment and the resulting wage gap.

Credit Reports as RRsumms: The Incidence of Pre-Employment Credit Screening

We study recent bans on employers' use of credit reports to screen job applicants – a practice that has been popular among employers, but controversial for its perceived disparate impact on racial minorities. Exploiting geographic, temporal, and job-level variation in which workers are covered by these bans, we analyze these bans' effects in two datasets: the panel dimension of the Current Population Survey (CPS); and data aggregated from state unemployment insurance records. We find that the bans reduced job-finding rates for blacks by 7 to 16 log points, and increased subsequent separation rates for black new hires by 3 percentage points, arguably contrary to the bans' intended effects. Results for Hispanics and whites are less conclusive. We interpret these findings in a statistical discrimination model in which credit report data, more so for blacks than for other groups, send a high-precision signal relative to the precision of employers' priors.

A question of degree: The effects of degree class on labor market outcomes

How does performance at university affect labor market outcomes? Employing a regression discontinuity design, we show that university degree class causally affects graduates’ industry, wages, and earnings. Our sample consists of students at the London School of Economics, and our data combine administrative records with the Destinations of Leavers from Higher Education survey. We estimate that receiving a First Class degree instead of an Upper Second increases the probability of working in a high-wage industry by fourteen percentage points, leads to three percent higher wages, and yields two percent higher annual salaries. For the comparison between Upper and Lower Seconds, the corresponding figures are ten, seven, and four. Effects are larger for males and graduates of math-intensive degree programs. We show that this is consistent with a model of statistical discrimination, in which employers attach more importance to the degree class signal if it is more informative about underlying ability.

Employment Discrimination in a Former Soviet Union Republic: Evidence from a Field Experiment

We provide the first experimental evidence about ethnic discrimination in the labor market in Georgia. We randomly assign Georgian and non-Georgian, male and female, names to similar resumes and apply for jobs as advertised in help-wanted web sites in Georgia. We find that gender has no effect on the probability of callback, but a job applicant who is ethnic Georgian is twice more likely to be called for a job interview than an equally skilled ethnic non-Georgian (Azeri or Armenian). The almost 100% gap in callbacks is statistically significant and cannot be abridged by having more experience or education. Both taste-based discrimination and statistical discrimination models are consistent with the evidence provided in this study. Labor market discrimination tends to aggravate in economic busts.

The heterogeneity of ethnic employment gaps

This paper investigates the heterogeneity of ethnic employment gaps using a new single-index based approach. Instead of stratifying our sample by age or education, we study ethnic employment gaps along a continuous measure of employability, the employment probability minority workers would have if their characteristics were priced as in the majority group. We apply this method to French males, comparing those whose parents are North African immigrants and those with native parents. We find that both the raw and the unexplained ethnic employment differentials are larger for low-employability workers than for high-employability ones. We show in a theoretical framework that this heterogeneity can be accounted for by homogeneous underlying mechanisms and is not evidence for, say, heterogeneous discrimination. Finally, we discuss our main empirical findings in the light of simple taste-based vs. statistical discrimination models.

Racial Discrimination in Grading: Evidence from Brazil

We investigate whether racial discrimination in the form of biased assessment of students is prevalent within Brazilian schools. Evidence is drawn from unique administrative data pertaining to eighth-grade students and educators. Holding constant performance in blindly-scored tests of proficiency and behavioral traits we find that blacks have lower teacher-assigned math grades than their white classmates. Heterogeneity in differentials provides evidence both of robustness with respect to omission biases and of compatibility with predictions from models of statistical discrimination. (JEL I21, I24, J15, O15)

Effective moving shadow detection using statistical discriminant model

Shadow detection is considered to be a challenging problem in computer vision and video analysis recently. In this paper, a novel moving shadow detection method based on statistical discriminant model for surveillance scenarios is presented. First, we extract various features from video frames to form a feature space for classification. Second, we adopt a statistical learning method that employs Partial Least Squares (PLS) and Logistic Discrimination (LD) to classify moving shadows and their corresponding objects. Finally, simple post processing is performed to obtain the refined classified results. Extensive experimental results indicate that the proposed method can adapt to different kinds of scenarios under various illumination conditions automatically. Specifically, compared with several well-known methods, our method exhibits much better performance for moving shadow detection.

Statistical Discrimination and the Implication of Employer-Employee Racial Matches

In this paper, I test the empirical validity of a statistical discrimination model that incorporates employer’s race. I argue that if an employer statistically discriminates less against an employee that shares the same race (matched) than an employee who does not share the same race (mismatched), then the correlation between the employee’s wage and his skill level (AFQT) is stronger for a matched employee than for a mismatched employee. Using data from the NLSY97, which includes information about the racial background of employees and their supervisors, I find evidence that is consistent with a statistical discrimination model for young male employees.

Human Capital Choice and the Wage Gap: The Role of Worklife Expectancy and Statistical Discrimination

This paper explores the link between worklife expectancy, educational attainment and its impact on the wage gap. A model of statistical discrimination that accounts for the achievement gap and differences in worklife expectancy demonstrates that (a) the achievement gap does not influence educational attainment directly through the choice of human capital, (b) discrimination lowers human capital directly through lower wage offers, and (c) lower worklife expectancy leads to lower human capital accumulation. While sufficient data on worklife expectancy by race and education is not available for a detailed empirical analysis, a simple calibration suggests the importance of the achievement gap and worklife expectancy relative to traditional labor market discrimination.

Is ethnic discrimination due to distaste or statistics?

Employing a vignette experiment, we test the empirical importance of key attitudes underlying the models of taste-based and statistical discrimination in explaining ethnic hiring discrimination. We find that employer concern that co-workers and customers prefer collaborating with natives drives discrimination.

Collective Reputation and the Dynamics of Statistical Discrimination

Economists have developed theoretical models identifying self-fulfilling expectations as an important source of statistical discrimination in labor markets (Arrow, 1973). The static models dominating the literature of statistical discrimination, however, may leave the false impression that a bad equilibrium is as fragile as a bubble'' and can burst at any moment when expectations flip. Such models thus understate the adversity that disadvantaged groups face in seeking to escape bad equilibria. By developing a dynamic version of a statistical discrimination model based on Coate and Loury's (1993) original setup, we clarify the limits of expectations-related fragility. We show that when a group is strongly affected by negative reputational externalities, the group cannot escape a low skill investment trap, regardless of how expectations are formed. By examining the evolution of stereotypes in this way, we also provide new insights into egalitarian policies.

Efectos del prestigio de los colegios de enseñanza media, de las universidades y el aprendizaje del empleador sobre los salarios en Chile

espanolEste estudio examina los efectos de haber asistido a un colegio privado, el ranking de universidades y el aprendizaje del empleador sobre los salarios en Chile. La estrategia empirica se basa en regresiones de salarios tipo Mincer. El analisis se realizo usando una muestra de egresados universitarios. Los principales resultados dan cuenta de un efecto importante y significativo de haber asistido a colegios privados y a universidades prestigiosas sobre los salarios de los trabajadores al inicio de su carrera laboral. Estos resultados son consistentes con un modelo de discriminacion estadistica y aprendizaje por parte del empleador dado que los efectos estimados disminuyen con la experiencia laboral de los trabajadores. Utilizando este modelo, se encuentra que los empleadores reducen en un 50 por ciento la ponderacion que le dan al ranking de universidades de sus trabajadores luego de tres anos. Mis hallazgos, ademas, indican que incorporar rankings basados en el percentil de la prueba de admision universitaria en las solicitudes de empleo puede mejorar significativamente la capacidad del mercado para asignar adecuadamente los sueldos, reduciendo la brecha informativa entre potenciales empleados y empleadores. EnglishThis paper studies the effects of attending a private high school, the university ranking and the employer learning on wages. The empirical strategy is based on the Mincer-type wage regressions. The analysis was carried out using individual-level data from Chile. I find a large and significant effect on wages at the beginning of workers’ careers from attending a private high school and from attending a highly ranked university. These findings can be rationalized by the statistical discrimination and employer learning model since the effects of attending a private high school or highly ranked university decrease with experience. I construct an employer learning model to explain these decreases and find employers decrease by 50 per cent the weight they place on university ranking when setting wages in three years. My findings further indicate that incorporating university admission test percentile rankings in employment applications can significantly improve the market’s ability to appropriately assign wages by decreasing the information gap between potential employees and employers.

Taste‐Based or Statistical Discrimination: The Economics of Discrimination Returns to its Roots

We briefly review the evolution of empirical work on discrimination. We discuss why traditional regression-based approaches neither convincingly measure market discrimination nor disentangle the relative importance of animus versus statistical discrimination in explaining such discrimination as exists. We describe the development of modern correspondence studies. We argue that these studies have the promise to credibly identify the presence of discrimination if not its magnitude, can inform us about the underlying mechanism generating discrimination and can also point to avenues for new theoretical and empirical work on discrimination. We discuss two articles with exemplary applications of these new methods. At least since Becker (1957), economists have been deeply interested in the presence of discrimination in markets. In the 1970s and into the 1980s, there was an active debate in the literature about whether discrimination was better described by Becker’s taste-based discrimination model, or by Arrow and Phelps’ statistical discrimination model. That debate gave way to a theoretical literature that advanced each model separately and to an empirical literature focused on documenting the presence of disparities and the effects of policies designed to counteract discrimination. Only recently has the literature returned to the question of whether taste-based or statistical discrimination is a more appropriate description of the phenomenon. The two articles that follow are good examples of the move towards an understanding of the mechanisms underlying discrimination. In this article, we attempt to place the literature’s recent direction into context and to show how it is a natural place for economists studying discrimination to arrive, given where we began. In The Economics of Discrimination, Becker explored ways in which individual tastes for discrimination might interact in a market setting to produce discriminatory outcomes for market participants. Becker described three models, each considering a different source of discriminatory tastes: employers, co-workers and customers. The Economics of Discrimination included deep insights about discrimination. Becker clarified the distinction between individual tastes for discrimination and market discrimination, he clarified the difference between segregation and market discrimination and he showed how the market might produce incentives for those with discriminatory tastes to avoid interaction with those towards whom they harbour animus. 1

Why do I like people like me?

Abstract In this paper we extend the standard model of statistical discrimination to a multidimensional framework where the accuracy of evaluators depends on how knowledgeable they are in each dimension. The model yields two main implications. First, candidates who excel in the same dimensions as the evaluator tend to be preferred. Second, if two equally productive groups of workers differ in their distribution of ability across dimensions group discrimination will arise unless (i) evaluators are well informed about the extent of these differences and (ii) evaluators can take candidatesʼ group belonging into account in their assessments.

Insights into the molecular mechanism of protein native-like aggregation upon glycation

Several human neurodegenerative diseases such as Alzheimer's disease, Parkinson's disease and Familial Amyloidotic Polyneuropathy, have long been associated with, structural and functional changes in disease related proteins leading to aggregation into amyloid fibrils. Such changes can be triggered by post-translational modifications. Methylglyoxal modifications have been shown to induce the formation of small and stable native-like aggregates in the case of the amyloidogenic proteins insulin and α-synuclein. However, the fundamental biophysical mechanism underlying such methylglyoxal-induced protein aggregation is not yet fully understood. In this work cytochrome c (Cyt c) was used as a model protein for the characterization of specific glycation targets and to study their impact on protein structure, stability, and ability to form native-like aggregates. Our results show that methylglyoxal covalently modifies Cyt c at a single residue and induces early conformational changes that lead to the formation of native-like aggregates. Furthermore, partially unfolded species are formed, but do not seem to be implicated in the aggregation process. This shows a clear difference from the amyloid fibril mechanisms which involve partially or totally unfolded intermediates. Equilibrium-unfolding experiments show that glycation strongly decreases Cyt c conformational stability, which is balanced with an increase of conformational stability upon aggregation. Data collected from analytical and spectroscopic techniques, along with kinetic analysis based on least-squares parameter fitting and statistical model discrimination are used to help to understand the driving force underlying glycation-induced native-like aggregation, and enable the proposal of a comprehensive thermodynamic and kinetic model for native-like aggregation of methylglyoxal glycated Cyt c.

Wage Growth and Job Mobility in the Early Career : Testing a Statistical Discrimination Model of the Gender Wage Gap

The paper focuses on the early career patterns of young male and female workers. It investigates potential dynamic links between statistical discrimination, mobility, tenure and wage profiles. The model assumes that it is more costly for an employer to assess female workers' productivity and that the noise/signal ratio tapers off more rapidly for male workers. These two assumptions yield numerous theoretical predictions pertaining to gender wage gaps. These predictions are tested using data from the 1979 cohort of the National Longitudinal Survey of Youth. As predicted by our statistical discrimination model, we find that men and women have the same wage at the start of their career, but that female wages grow at a slower rate, creating a gender wage gap. Also consistent with our model, we find that mean wages are higher for workers who keep their job, while wage growth is stronger for workers who change job.

Gender Differences in Academic Promotion and Mobility at a Major Australian University*

This paper analyses gender differences in faculty promotion and mobility at the University of New South Wales (UNSW) using personnel records for 1999–2010. It finds female lecturers less likely than men to be promoted or to leave UNSW and female associate professors more likely to be promoted. These results are consistent with a flipping statistical discrimination model. Leave‐taking did not account for gender differences. Instead, taking maternity leave indicates higher likelihood of staying at UNSW and being promoted. The paper finds that gender differences narrowed over time but not always as predicted by gender‐related policy changes.

Wage Growth and Job Mobility in the Early Career: Testing a Statistical Discrimination Model of the Gender Wage Gap

The paper focuses on the early career patterns of young male and female workers. It investigates potential dynamic links between statistical discrimination, mobility, tenure and wage profiles. The model assumes that it is more costly for an employer to assess female workers' productivity and that the noise/signal ratio tapers off more rapidly for male workers. These two assumptions yield numerous theoretical predictions pertaining to gender wage gaps. These predictions are tested using data from the 1979 cohort of the National Longitudinal Survey of Youth. As predicted by our statistical discrimination model, we find that men and women have the same wage at the start of their career, but that female wages grow at a slower rate, creating a gender wage gap. Also consistent with our model, we find that mean wages are higher for workers who keep their job, while wage growth is stronger for workers who change job.