Black-white Wage Differentials Research Articles

Isolating the Roles of Individual Covariates in Reweighting Estimation

SUMMARY A host of recent research has used reweighting methods to analyze the extent to which observable characteristics predict between-group differences in the distribution of an outcome. Less attention has been paid to using reweighting methods to isolate the roles of individual covariates. We analyze two approaches that have been used in previous studies, and we propose a new approach that examines the role of one covariate while holding the marginal distribution of the other covariates constant. We illustrate the differences between the methods with a numerical example and an empirical analysis of black–white wage differentials among males. Copyright © 2015 John Wiley & Sons, Ltd.

Black–white wage differentials: duration and probability unemployment effects in a multiple of sample selection bias model

The extent to which probability and duration of unemployment affect the black–white wage differentials is examined in this paper. The paper simultaneously incorporates in the wage equation the multiple sample selection bias that occurs as a result of individuals’ propensity to be in the labor force, and the firm’s hiring decisions. The results reveal a substantial contribution of the duration of unemployment variable to the black–white wage differential, but a small portion of the differential is explained by the probability of unemployment. The results also indicate a sizeable difference between the contribution of the duration of unemployment variable to the male’s wage differentials (26%) and to the female’s (35%). The study finds that an individual’s labor force decision as well as a firm’s hiring decision are important in the wage determination process and that failure to account for the sample selectivity bias due to these two decisions will result in either underestimating or overestimating the wage differentials between black and white workers. At the macro level, the results seem to suggest that promotion of racial wage equality should be associated with policies that will minimize blacks’ incidence of unemployment and duration of unemployment spells.

Black–White Wage Differentials in a Multiple Sample Selection Bias Model

This paper simultaneously incorporates two sources of selection bias in the black-white wage equations. It demonstrates that the biases due to an individual’s propensity to be in the labor force and the firm’s hiring practices are important in determining the black–white wage differential and failure to account for both biases will result in inaccurate estimation of the black–white wage differential. We found that adjusting for double selection bias in the wage equation, the black–white female wage gap is 26% larger than the black–white male wage gap, and 12.1% larger when we adjust for a single selection bias. The results seem to suggest that at the macro level, the enforcement of policies related to racial issues in the labor market will likely lead to a reduction in the black–white wage gap.

Measurement Error in Human Capital and the Black-White Wage Gap

Proxy variables are frequently used in economics to control for unavailable variables in a linear regression setting. For example, AFQT scores have been used to control for human capital accumulation in measuring black-white wage differentials. This practice may bias the coefficient estimates for the correctly measured variables as well. This paper models proxy variables as a measurement error process and derives bounds for the coefficients on the correctly measured variables under a variety of assumptions. The results show that the coefficient on race in a linear regression is an overstatement of the actual black-white wage gap. Sensitivity analysis suggests that if human capital could be correctly measured it would be unlikely that the coefficient on black would be negative.

Changes in relative wages in the 1980s Returns to observed and unobserved skills and black–white wage differentials

We assess the potential contribution of a rise in the return to unmeasured productivity correlated with education and race to the dramatic increase in the college–high-school wage differential and the stagnation of black–white wage convergence during the 1980s. A relatively unrestricted error-components panel data model is used to estimate the rise in the unobserved skill premium. Identification of the model is based on across-group variation in changes in within-group log-wage variances over time. In the absence of credible instruments for education and race, we calibrate the impact of time-varying `ability’ biases under various assumptions on the extent of non-random sorting of ability. Both between-cohort and within-cohort changes are examined using earnings data on men from multiple Current Population Surveys. There is systematic variation in changes in within-group wage variances over time, suggesting about a 10–25% rise in the unobserved skill premium during the 1980s. In addition, there are noticeable differences across cohorts in changes in the college–high-school wage gap. However, the model estimates imply that the rise in the return to ability can account for at most 30–40% of the observed rise in the college premium for young workers. Similarly, young, well-educated black men experienced at least a 0.13 log point decline in wages relative to their white counterparts between 1979 and 1991.

Understanding Black–White Wage Differentials: 1960–1990

Understanding Black-White Wage Differentials, 1960-1990 Author(s): James J. Heckman, Thomas M. Lyons, Petra E. Todd Source: The American Economic Review, Vol. 90, No. 2, Papers and Proceedings of the One Hundred Twelfth Annual Meeting of the American Economic Association, (May, 2000), pp. 344 -349 Published by: American Economic Association Stable URL: http://www.jstor.org/stable/117248 Accessed: 16/08/2008 00:56

HAS DISCRIMINATION LESSENED OVER TIME? A TEST USING BASEBALL'S ALL‐STAR VOTE

I. INTRODUCTION The objective of this analysis is to examine changes in racial discrimination over time. The labor market literature indicates that black-white wage differentials have declined substantially in the last 45 years, and most of that decline came before 1980.(1) However, the degree to which this was due to altered attitudes among whites, increased productivity among blacks, or corrective legislation is debated.(2) Most studies are not able to effectively test the hypothesis of changes in attitudes, as attitudes are very difficult to observe. Some activities may be very indicative of attitudes, though. Take voting for baseball's annual Game. Each vote is cast anonymously; the voter gets neither praise nor censure. The cost of voting is low - ballots are freely supplied at ballparks, and filling them out takes but a minute or two. To a large degree, the reward is in the voting itself; the voter is able to express his/her opinion on an issue of interest. There are no rules to be followed; a fan can vote for a player because of his ability, reputation, performance, team, age, appearance, color, or anything else. The vote is as close to a pure expression of preference as is likely to be found. If fans prefer white players to black, all else equal, black players should receive fewer votes than equivalent whites. If those preferences change, the vote differential should change in tandem. This is the test that this article applies, in an examination of voting from 1970 to 1996. Baseball has long been a useful source of discrimination studies - it has a clear and unambiguous history of discrimination in which owners, players, and fans have all been implicated.(3) There are two particular advantages in looking to baseball for evidence of changes in discriminatory attitudes: baseball has not been as directly influenced by the equal opportunity and affirmative action laws that are argued to have helped close the black-white wage gap, and baseball productivity can be controlled for in a relatively precise fashion. While changes in the wage may have a variety of causes, changes in an All-Star vote gap more plausibly reflect changes in attitudes. Therefore, if we find the variance in votes unexplained by other factors and correlated with race has shrunk over time, we will take it as support for the hypothesis that discriminatory attitudes have weakened. The article will proceed as follows: Section II discusses the vote and the underlying model. Section III presents the analysis. Section IV concludes. II. THE ALL-STAR GAME Baseball's Game has been played every year (and twice per year for a short period) since 1933, with the exception of the war year of 1945. It is held in the middle of the season, and pits the best players from each of Major League Baseball's (MLB) two leagues (the National and the American) against each other. Various methods of selecting teams have gone in and out of vogue. The game was first established through the efforts of Arch Ward, sports editor of the Chicago Tribune, as a special part of the 1933 Chicago World's Fair. Ballots were printed in the Tribune and in several other major newspapers, and fans voted for their favorite stars. Fan voting was dropped two years later in favor of selection by baseball's managers, only to be re-instituted in 1947, and abandoned a second time (after ballot-stuffing by Cincinnati Reds fans led to the election of seven Reds to the starting line-up) in 1957. From 1958 to 1969, the starting line-ups were chosen by a vote of the players themselves. In 1970, baseball commissioner Bowie Kuhn re-established the fan vote amid much ballyhoo. The new system was a popular success, and has remained in place ever since. This study analyzes and compares fan votes over 26 years. It begins with 1970, the year that the fan vote was re-established, ends with 1996, and fills in the middle with the votes of 1971, 1973, 1975, 1977, 1979, 1980, 1985, and 1990. …

African Americans and Post-Industrial Labor Markets.

This collective portrait documents the disproportionate vulnerability of African Americans to the ongoing transformations of the U.S. economy from industrial to service areas as the twenty-first century approaches. chapters have been previously published in Review of Political Economy between 1991 and 1996. This volume represents one of the best sources of up-to-date perspectives on the circumstances facing African Americans in post-industrial labor markets. African Americans and Post-Industrial Labor Markets is divided into seven sections: The Intersection of Race, Power, Culture, and Economic Discrimination, Black-White Wage Differentials, Occupational Crowding, Black Women in the Labor Market, Structural Unemployment and Job Displacement, Sectoral Analyses, and Strategies to Increase Employment. authors discuss such topics as: the impact of the general status of race relations on labor markets; increasing access to higher-paying occupations; the relationship between occupational segregation and local labor market dynamics; and the earnings of black women compared to white women and black and white men. chapters are connected by a common theme: black employment is highly sensitive to changes in both aggregate and local economic conditions. As a result, policy changes designed to promote macro-level economic stabilization could well have the unintended effect of further increasing job instability among blacks. African Americans and Post-Industrial Labor Markets is a momentous compendium and should be read by economists, African American studies scholars, sociologists, and professionals in the business world.

Wage dispersion, returns to skill, and black-white wage differentials

During the 1980s wage differentials between different age and education groups expanded rapidly. Wage dispersion among individuals with the same age and education also rose. A simple explanation for both sets of facts is that earnings vary with a one-dimensional index of skill and that the relative return to higher levels of skill rose over the decade. We explore a simple method for estimating and testing such a ‘single-index’ model of wages. Our approach integrates three aspects of skill: age, education, and unobserved ability. We find that a one-dimensional skill model provides a reasonably accurate, although oversimplified, description of changes in the structure of wages for white men and women between 1979 and 1989. We then use the estimated models for white women and white men to interpret recent changes in the relative wages of African American workers.

Returns to Schooling, Implicit Discount Rates and Black-White Wage Differentials

A simple econometric model of investment in schooling is developed and estimated. The measure of individual discount rates implicit in their educational investment decisions suggests no difference between individuals from different socioeconomic backgrounds. Differences in individual speeds of educational attainment, which do vary with background, explain most of the variation in levels of attainment that is attributable to family background.

Imperfect Signalling, Affirmative Action, and Black-White Wage Differentials

The existence of black-white wage differentials has received considerable attention in the economics literature. Imperfect information provides the underpinning of several models which attempt to explain such differentials [1; 2; 3; 4; 11]. Arrow's work, for example, shows that, when screening is costly and expectations of productivity differ across groups, wages may differ between individuals with identical productivity [2; 3]. Borjas and Goldberg point out that biased screens may lead to wage differentials even when the two groups have identical distributions of productivity [4]. It is important to note that these differentials arise in the absence of discriminatory tastes among market participants. Other researchers have investigated the impact of affirmative action programs on minority-majority wage differentials. Welch indicates the kinds of wage differences which can arise under certainty about productivity levels [12]. Lazear demonstrates that the apparent narrowing of blackwhite differentials under affirmative action may well be an illusion due to the failure to measure non-pecuniary income in the form of on-the-job training [7]. While these models provide an appealing explanation of the main outlines of minoritymajority wage differentials, they do not explain another stylized fact which has been observed recently. A study of academicians revealed that black males with a number of publications earned more than their white counterparts, whereas black male academics with no publications earned less than comparable whites [5]. Olneck showed that, in terms of occupational status, the return to college graduation was higher for non-whites than for whites, but the return to secondary education was lower [6, 159-90]. Similarly, while the income of college-educated black males rose relative to that of whites of similar qualifications between 1967 and 1978, the income of less educated black males fell relative to that of their white counterparts over the same period.' Sowell has argued that this skewing of the relative distributions is a result of affirmative action programs [9; 10]. Although the quotas seem to increase the demand for minority workers, their legislative framework increases the

Structural Equations and Path Analysis for Discrete Data

This article proposes a solution to the long-standing methodological problem of incorporating discrete variables inoto causal models of social phenomena. Only a subset of the variety of ways in which discrete data arise in empirical social research can be satisfactorily modeled by conventional log-linear or logit approaches. Drawing on the insights of several literatures, this article exposits a general approach to causal models in which some or all variables are discretely measured and shows that path analytic methods are available which permit quantification of causal relationships among variables with the same flexibility and power of interpretation as is feasible in models that include only continous variables. It presents methods of identifying and estimating these models and shows how the direct and indirect effects of independent varibles can be calculated by extensions of usual path analysis methods for continuos variables An important distincion developed here is that discrete variables can play two roles: (1) as measures of inherently discrete phenomena and (2) as indicators of underlying continous variables. The value of this distinction is shown in two empirical examples examined previously by other authors. In examining the effects of social background and parental enouragement on college plans of high school seniors, the article shows that modeling a discrete measure of encouragement as an indicator of a latent continous variable rather than as an inherently discrete variable (as has been done in previous analyses) provides a clearer interpretation and a superior fit to the data. In examining the effects of state Fair-Employment-Practices Legislation on black-white wage differentials, this study shows that two distinct effects on the relative wage can be detected: the direct ameliorative effect of the law itself and the effects of the popular progressive sentiment for racial equality of which the law is an indicator. The methods and models presented here are not only natural generalizations of structural equation and path analysis methods for continuous varibles to include discrete variables but also provide a means of investigating a richer variety of substantive hypotheses than is feasible with methods for discrete data commonly used in the sociological literature to date.

On the Relationship between Discrimination and Market Structure: Reply

In our paper [3], we test the Alchian-Kessel [1] managerial discretion hypothesis that the extent to which employers exercise tastes for discrimination should be systematically related to the extent of their market power. To motivate an empirical specification we develop a model of an industry labor market which demonstrates, under general conditions of supply and demand for labor, that tastes for discrimination will affect black-white wage differentials across industries as well as employment ratios. If labor supply curves are infinitely elastic the model clearly indicates that tastes for discrimination will affect employment alone. The assumption of an infinitely elastic labor supply curve is a convenient simplifying convention employed in a number, e.g. [2], of theoretical and empirical studies of discrimination. The validity of this assumption at the industry level can be tested empirically. Therefore the statistical analysis in our paper is designed to test the two interrelated hypotheses: (i) are labor supply curves to an industry upward sloping or infinitely elastic? and (ii) is discrimination within an industry systematically related to market power. To measure potential discrimination observed black-white wage differentials within several industries were corrected for productivity differences of black and white employees. Then, the corrected wage differentials were regressed on the industry characteristics, market concentration, percent union, and the elasticity of demand for labor. The theoretical model indicates that if labor supply curves are infinitely elastic then all of these industry characteristics should be unrelated to the corrected wage differentials. On the other hand, a significant coefficient on the elasticity of demand and percent union is persuasive evidence that labor supply curves are upward sloping. The managerial discretion hypothesis is tested by the significance of the market power variable. Our results confirm that industry labor supply curves rise with employment. Blackwhite wage differentials across industries are systematically related to industry characteristics such as percent union and the elasticity of demand for labor. The results, however, offer no support to the contention that market power and discrimination are related. The author [4] suggests that the latter result can be attributed to perfectly elastic supply curves of labor, an assertion which we have shown to be incorrect.