Abstract
Fifty years have passed since the release of the Kerner Commission’s findings, conclusions, and policy recommendations. This article first reviews recent trend and cross-section analysis on racial employment and earnings inequality before synthesizing the evidence on racial inequality’s causes and speculating how these factors might shape future African American outcomes. In conclusion, it offers a framework for addressing the nation’s persistent racial inequality.
Last year marked the fiftieth anniversary of President Lyndon Johnson’s Kerner Commission Report. The commission’s goal was to explain why the summer of 1967 riots occurred and what could be done to prevent them in the future. It concluded that “our Nation is moving toward two societies, one black, one white—separate and unequal” (Harris and Curtis 2018).
Since then, African American workers have experienced dramatic changes in the U.S. labor market. They continue to do so. On the one hand, they narrowed education and skill gaps with whites. On the other, a disproportionate share of blacks came into contact with the criminal justice system. During this half century, the U.S. economy experienced its strongest economic expansions, yet at the same time globalization (immigration and trade) and technological innovation changed how all Americans work, when they work, where they work, and with whom they work.1 Labor-market institutions changed dramatically. Enforcement of affirmative action and antidiscrimination laws was first supported, but then challenged.
Today, they are under renewed scrutiny. Organized labor’s ability to influence employment and earnings has weakened. Union membership fell from 20.1 percent in 1983 to 10.7 percent in 2017.2 Increases in wage floors, such as the minimum wage, became increasingly difficult to implement. The last time the federal minimum wage was increased was in 2009. Given these and other dramatic shifts in the economy and policy over this period, to what extent have African Americans improved their absolute and relative standing in the U.S. economy?
To answer this broad question, this article first reviews recent trend and cross-section analysis on racial employment and earnings inequality, synthesizes the evidence on inequality’s causes, offers a framework for addressing inequality, and concludes with thoughts on the future prospects of African Americans and their communities.
The main conclusion is that the relative earnings of African Americans are similar to what they were in 1979. If the widespread effects of incarceration are included, the relative position falls to what it was in 1950. These sobering comparisons are because we have and continue to choose policies that put less compensation and power in the hands of workers. Macroeconomic growth cannot compensate for the forces that generate overall inequality. A summary of the supporting evidence for this conclusion follows.
UNEMPLOYMENT AND EMPLOYMENT
Since the early 1970s, the African American unemployment rate has typically remained twice as large as the white, even among college graduates. The U-6 unemployment rate from the U.S. Bureau of Labor Statistics (BLS), which includes the underemployed, suggests a picture more synonymous with the historical ratio of black and white unemployment rates.3
The employment-population ratio provides the most comprehensive description of the labor-market prospects for blacks. Along with job search (unemployment), this measure captures labor-force participation decisions. It tells a different story than the official BLS unemployment rate. For the first five months of 2019, the African American employment-population ratio sat at 58.2 percent, relative to a record of 60.5 percent in 2000. Further, it has just recently made up ground lost during the Great Recession.
This lack of the macroeconomy’s ability to move African American outcomes past their pre-recession levels is problematic. Previous work shows that young black men’s labor-market outcomes are more responsive to improvements in the macroeconomy.4 However, a 1 percent decrease in the unemployment rate is associated with a smaller increase in the employment-population ratio of young, non-college-educated African American men and women (Freeman and Rodgers, forthcoming). This means that macroeconomic growth as a tool to reduce racial employment and wage inequality is not as effective as it was in the past.
The final major employment theme is the adverse impact that mass incarceration had and continues to have on African Americans. The consensus is that the increase in incarceration has had a major impact on young non-college-educated men and women. Criminal justice policy became more punitive, particularly for nonviolent drug offenders.
EARNINGS
Today, the black-white earnings gap is larger than it was in 1979. This expansion did not emerge gradually. During the 1960s, earnings inequality narrowed rapidly, followed by an expansion starting in the mid-1970s and throughout the 1980s. During the 1990s, especially the second half of the decade, racial inequality narrowed considerably. Since 2001, the racial wage gap expanded (Wilson and Rodgers 2016).
The primary factors for the narrowing of the gap during the 1960s were a closing of the gaps in human capital, and the passing, implementation, and enforcement of affirmative action and antidiscrimination laws. Almost fifty years ago, Richard Freeman (1973) undertook one of the first empirical studies to demonstrate the ability of “governmental and related anti-discriminatory activity” in the form of the 1964 Civil Rights Act to improve both the absolute and relative economic status of African Americans. The study also revealed a key role that the 1960s economic expansion played in narrowing racial inequality.
John Bound and Freeman (1992) show that the expansion of the earnings gap during the mid-1970s through the 1980s is attributable to a slowdown in the African American acquisition of human capital, an erosion in unions especially in the Midwest, and a failure to raise the federal minimum wage. Chinhui Juhn, Kevin Murphy, and Brooks Pierce (1991) took a different empirical approach and find that growing earnings inequality that disadvantaged all less-skilled workers contributed to racial inequality’s widening. The sources of the growth in earnings inequality are consistent with what Bound and Freeman (1992) identify as contributing to expansion of the wage gap.
Moving to the 1990s, Freeman and William M. Rodgers III (2000) attribute the relative improvement of African Americans to the period’s “tight” labor market and a slight improvement in institutions that help to raise the wages of all low- and moderate-wage workers. These race-neutral improvements had a disproportionate impact on blacks because they were more likely to be impacted by the policy changes. Valerie Wilson and Rodgers (2016) show that since the early 2000s, the erosion in the relative position of African Americans has primarily been because of discrimination (or racial differences in skills or worker characteristics that are unobserved or unmeasured in the data) and growing earnings inequality that has reemerged.
However, when racial differences in mass incarceration are taken into account, Patrick Bayer and Kerwin Charles (2016) show that the current economic status of African Americans is roughly equivalent to that in 1950. Criminal justice policies that led to widespread incarceration swamp the gains of African Americans over the period. They too exacerbate the growth in discrimination and factors such as trade, technology, and policies (such as the minimum wage) that led to greater general earnings inequality.
Documenting and explaining racial inequality is a complex task. Racial differences in labor supply, labor demand, and institutional factors all explain a portion of racial employment and earnings inequality. The reasons can depend on educational attainment, region of residence, age, and gender. They also depend on type of work. Elsewhere in this issue, Katharine Abraham and Susan Houseman (2019) show that a larger fraction of respondents who report working in informal jobs to earn money are less educated, minority, low income, unemployed, or experiencing financial distress. As Bound and Freeman (1992) conclude, no single causal explanation accounts for the nation’s persistent racial inequality. Given the lack of a single narrative, no “silver bullet” policy will substantially improve the absolute and relative standing of African Americans. A comprehensive set of policies need to be not only developed but also committed to for a long time.
TREND ANALYSIS: EXPLAINING RACIAL DIFFERENCES IN EMPLOYMENT AND UNEMPLOYMENT
Figure 1 shows that during the 1970s, the African American unemployment rate fell, rose dramatically during the 1981–1982 recession, fell sharply between 1982 and 1989, rose slightly during the 1991–1992 recession, and then trended downward until the Great Recession, when it rose to above 15 percent. Since the end of the recession, it has fallen to a series record of 6.5 percent in 2018. The key observation is that even during times of prosperity (the 1980s, 1990s, and present) the African American unemployment rate typically remained twice as large as the white unemployment rate. At the time of writing, during the first five months of 2019, the black-white ratio sits at 2.1.
U.S. Unemployment Rates
Source: Author’s compilation based on BLS unemployment rates from the Labor Force Statistics of the Current Population Survey (BLS 2019).
Regardless of educational attainment, black unemployment rates exceed white unemployment rates. For example, as of April 2019, adult African American college graduates (bachelor’s degree only) have an unemployment rate of 2.7 percent, which corresponds to that of adult white high school graduates. This relationship has not changed over time.5
Given the prevalence of informal work (Abraham and Houseman 2019; Katz and Krueger 2019) and the presence of “fissured” labor markets (Weil 2019), the BLS U-6 unemployment rate and the employment-population ratio provide better descriptions of African American labor-market prospects.6
The U-6 unemployment rate measure does a better job of documenting underemployment (such as part-time work for economic reasons) that emerged prior to and during the recession and has lingered during the expansion.7 BLS does not publish the U-6 statistic by race. The Economic Policy Institute does, however. Its estimate for December 2018 indicates a rate of 11.9 percent for African Americans, 10.1 percent for Latinos, and 6.3 percent for whites.
The employment-population ratio captures job search and participation decisions. Figure 2 reports the ratio by race. Both tell a different story than the official BLS rate. For the first five months of 2019, the African American employment-population ratio was 58.2 percent, relative to a record of 60.5 percent in 2000. Further, it took eight years of expansion before the ratio in figure 2 made up the ground lost during the Great Recession.
U.S. Employment-Population Ratio
Source: Author’s compilation based on BLS unemployment rates from the Labor Force Statistics of the Current Population Survey (BLS 2019).
The employment-population ratio is important for understanding the challenges that young out-of-school and less-educated minorities face. Pamela Loprest, Demetra Nightingale, and Shayne Spaulding (2019) report the trends in participation by race and education for young adults. Of all race-gender-age demographic groups, young less-educated men typically have the lowest ratio.8
A large portion of racial inequality research focuses on explaining why young less-educated black men have the lowest labor-force attachment. Loprest, Nightingale, and Spaulding (2019) provide an excellent summary of the major reasons. To date, the chief factors are contact with the criminal justice system, structural racism and neighborhood effects of poverty, educational attainment differences, discrimination, and the challenges of being a youth parent (whether mother or father). The consensus is that mass incarceration has the worst impact on the employment prospects of all youth, but especially African American youth.
The following studies provide convincing evidence on incarceration’s role in creating racial inequality. Bruce Western, Jeffrey Kling, and David Weiman (2001) show that the nonwhite prison population increased from 48 percent in 1979 to 67 percent in 1997, much more than their share of the overall population. One-third of African American male high school dropouts between the ages of twenty and thirty-five were in jail during the late 1990s. Western and Becky Pettit (2002) show for men in their thirties, in 1989, approximately 2 percent of white men had ever been incarcerated relative to 13 percent of black men. A decade later, these rates had jumped by 50 percent. The risks of incarceration are about three times higher for high school dropouts in their thirties. In 1999, 14 percent of white and 59 percent of black male high school dropouts in their early thirties had been incarcerated. Harry Holzer, Paul Offner, and Elaine Sorensen (2005) show that during the 1980s and 1990s, prior incarceration and child support enforcement can explain half or more of the erosion in employment among twenty-five-to-thirty-four-year-old African American males. Prior incarceration also explains some of the decline among younger cohorts.
How do these differences affect employment? Bayer and Charles (2016) show that mass incarceration even explains the overall black-white male employment gap, not just racial inequality among young less-educated men. They show that since 1960, the fractions of both African American and white men who are not working rose dramatically. A major divergence in the odds of employment between black and white men occurred, the former working less than the latter.9
Bayer and Charles decompose the racial employment gap’s growth from 1960 to 2010 into three gaps: unemployment, incarceration, and labor-force participation. They conclude that 22 percent of the growth in the probability of work racial gap is due to the increasing unemployment gap, 43 percent to the incarceration gap’s widening, and 34 percent to the expansion in the labor-force participation gap.10
Kim Blankenship and Ana Maria Del Rio Gonzalez (2018) take a more granular approach and document the racially different experiences of ex-offenders who were jailed for drug-related offenses. Relative to whites, African Americans have fewer convictions but are more likely to be incarcerated. They are also more likely to experience incarceration as juveniles and to be incarcerated for drug charges. Whites are more likely to be let go when arrested. Incarceration had more adverse impacts on blacks’ education, dropping out of school, job loss, eviction, and homelessness. Incarceration also negatively affected the personal relationships of blacks more than whites.
Finally, even a booming economy that generates extremely tight labor markets is unable to offset the adverse effects of incarceration’s growth. William Darity and Samuel Myers explain this observation with their test of the Rusche and Kirchheimer hypothesis (Darity and Myers 2000; Rusche and Kirchheimer 1939). The hypothesis is that when superfluous labor exists, incarceration rates rise to drain off unwanted or unneeded workers. If true, during extremely tight labor markets such as in the 1990s, criminal justice policy should have adapted to release needed workers into the labor market. Darity and Myers find no empirical support for this explanation. In fact, they conclude that the continued surge in African American incarceration in the 1990s is due to the intersection of prison privatization and black males becoming a permanent surplus population. The criminal justice system’s punishment became more punitive, particularly for nonviolent drug offenders.11
Earnings
Figure 3 reports BLS data on the black-white ratio of usual median weekly earnings of full-time workers. It reveals the widely documented pattern: a fall in the ratio during the 1980s, an increase during the 1990s boom, and a decline in the ratio since the early 2000s. Figure 4 presents the U.S. Census Bureau’s data on the black-white ratio of median income for full-time, year-round workers. All comparisons are relative to white men. Switching to the census allows racial comparisons to extend back to 1955. The figure shows the strong relative gains of African Americans during the 1960s, especially women.
Ratio of Usual Median Weekly Earnings (Relative to White Men)
Source: Author’s compilation based on BLS unemployment rates from the Labor Force Statistics of the Current Population Survey (BLS 2019).
Ratio of Full-Time, Year-Round Workers’ Median Income
Source: Author’s compilation based on U.S. Census Bureau medium income data for full-time and year-round workers (U.S. Census Bureau 2018, table P-36).
Many studies have sought to explain the ratio’s movement over time. Narrowing gaps in educational attainment explain much of the gains prior to 1980. No single explanation dominates why earnings inequality expanded during the 1980s. Young, less-educated men were hurt by regional shifts in employment, the fall in the real minimum wage, deunionization, and increased crime. The expansion among college graduates is due to a large increase in the supply of black college graduates, plus heightened discrimination. Evidence indicates a retrenchment in the enforcement of affirmative action and antidiscrimination laws. Gains during the 1990s are largely due to a reduction in discrimination as labor markets tightened and public policy became more favorable for reducing racial inequality. Continued progress in narrowing education gaps between young black and white men also contributed to improvements during this period. Yet sizable gaps remain.
Since the 1990s, the evidence has been more consistent with a worsening in discrimination, or the increased contribution of racial differences in unobservable characteristics, such as cognitive skills. As in the 1980s, the explanations depend on the background (for example, educational attainment, age, region of residence) of the African American worker. The racial gap in college completion has grown. Black-white wage inequality among less-educated workers has become less of a regional issue but a greater problem for Americans overall. Black-white wage gaps across regions have converged, but at higher levels of inequality.12 Declining unionization has played a role in the growth of the black-white wage gap, particularly among new-entrant men. The union wage premium has also diminished.
For the period from 1960 to 1980, David Card and Alan Krueger (1992) as well as others show that relative gains in the educational attainment of African Americans contributed heavily to the narrowing in the black-white wage gap among men. Specifically, they find that over this period, gains in the relative quality of black schools explain 20 percent of the convergence of black-white earnings.
Many of the early studies focus on black men, arguing that they face the greatest labor-market challenges. Two of the few studies that estimate racial wage gaps among women are Francine Blau and Andrea Beller’s (1992) and Cecilia Conrad’s (2001). Blau and Beller (1992) show wage gains were similar for both black men and women, but female annual earnings relative to men improved due to increased work time inputs. Because the gender earnings gap among whites narrowed during the period, black female’s wages rose relative to white females as well as white males. The black female-to-white male wage ratio increased by 32.7 percent and the corresponding male ratio by 8.6 percent.
As in the employment trends, this improvement ended during the 1980s, and the gains dissipated for African American men and women. The wage gap among men expanded during the 1980s (Bound and Freeman 1992; Juhn, Murphy, and Pierce 1991). One study, Bound and Freeman (1992), shows that the wage gap’s widening results from shifts in demand for subgroups due to shifting regional employment, the falling real minimum wage, and deunionization. Those authors attribute the expansion of the gap among non-college-educated men to increased crime among this group and the expansion among college graduates to a large increase in the supply of black college graduates.
Juhn, Murphy, and Pierce (1991) find that race-neutral factors were a more important contributor to racial inequality’s expansion in the 1980s than race-specific factors such as discrimination. These could be the factors that Bound and Freeman also identify. Agreement is less strong as to why the wage gap expanded between black and white women. Elaine McCrate and Laura Leete (1994), Deborah Anderson and David Shapiro (1996), and Blau and Beller (1992) attribute the expansion to a divergence in pay structures, especially in the returns to education.
Darity and Myers (1998) argue that retrenchment efforts to enforce affirmative action during the Reagan administrations play a key role in explaining racial inequality’s expansion during the 1980s. The government’s primary tools for implementing affirmative action and antidiscrimination laws is done through the Equal Employment Opportunity Commission (EEOC). Jonathan Leonard (1990, 1996) asserts that enforcement of EEO regulations ceased in the late 1970s, and discrimination worsened in the 1980s due to these retrenchment efforts.13
The literature is smaller on the effects that the EEOC and the Federal Contract Compliance Program have on employment diversity. Economists are not as active in this area. In a review of employment-based programs, Holzer and David Neumark (2000a, 2000b) find that affirmative action opens and expands the applicant pool for minorities without a meaningful reduction in the quality of minority candidates, and that it can be done with relatively little loss in economic efficiency. The statistical case against affirmative action based on efficiency has little merit. Ward Thomas (2000) also finds that the employment share of African Americans increases because federal contractors modify their search and application processes such that more African Americans are considered. Conrad Miller (2017) demonstrates that affirmative action increases the share of black workers and that the share of blacks continues to rise after a company is regulated.
To explain trends since the 1980s, two studies—Rodgers (2006) and Wilson and Rodgers (2016)—use the mean-based wage decomposition approach (Juhn, Murphy, and Pierce 1991). Meanwhile, a quantile distributional approach demonstrates that the general inequality explanation is limited to less-skilled workers and explains none of the erosion in the earnings of black college graduates (Rodgers 2006). The evidence in both studies is more consistent with a worsening in discrimination, or the increased contribution of racial differences in unobservable characteristics (such as cognitive skills).
Wilson and Rodgers (2016) describe and explain the black-white earnings gap by age, gender, education, and region of residence.14 The black-white wage gap expanded more among women, but it is still largest among men. Today, black men’s average hourly wages are 22.2 percent lower than those of white men in 1979. Black women were near parity with white women in 1979, but by 2015 faced a disadvantage of 19 percent.
Causes are numerous, but discrimination has consistently played a major role.15 Since 1995, the narrowing of new-entrant racial wage gaps has largely been due to a reduction in discrimination as labor markets tightened and public policy became more favorable for reducing racial wage inequality. Continued progress in narrowing education gaps between young black and white men also contributed to improvements during the period. Yet, similar to earlier periods, the earnings gap remained large, especially among college graduates.
Since the Great Recession, gaps among new-entrant women have expanded more than among any other experience or gender group. As it was before 2000, growing discrimination is the primary source of the erosion. The growing racial gap in college completion is also a contributing factor.
Black-white wage inequality among less-educated workers is becoming less of a regional issue but more of a problem for all Americans. Since 1979, black-white wage gaps across regions have converged, but at higher levels of inequality. The gaps range from 18 to 20 percent.
Declining unionization has played a role in the growing wage gap, particularly for new-entrant males. Between 1983 and 2015, the gap grew 1.6 percent among new entrants and 3.0 percent among experienced men. Unionization’s decline accounts for about one-fourth to one-fifth of this growth. Among new-entrant men and experienced men, a diminishing union wage premium accounts for 43 and 33 percent of the total growth in the men’s racial wage gaps.
A long-standing concern with evidence in some studies is the widespread labor market exiting of young men, especially African American men, due to their contact with the criminal justice system (see, for example, Wilson and Rodgers 2016). The concern is that earnings gaps may be too small or the trend analysis for the 1990s overstates the relative gains that African Americans made because it is assumed that if incarcerated black men were employed, they probably would be earning wages at the lower end of the wage distribution. Thus, an adjustment of this nature would lower the average wage of African American men.
For example, Juhn (2003) shows that from 1969 to 1993, accounting for men who are out of the labor force, inflation-adjusted wage growth for prime-age black men falls by 40 percent and wage convergence between blacks and whites is reduced by one-third.16 In a more recent trend analysis, Bayer and Charles (2016) include individuals who report zero earnings. Their use of quantile regression allows them to describe the black-white annual earnings gap at different points in the earnings distribution. Their analysis uncovers three key insights. First, incorporating nonemployment substantially changes the path of the earnings gap over the past thirty years. Instead of stagnating, the median earnings gap has expanded again, so much so that it is now as large as in 1950. Second, the narrowing of the median earnings gap is largely due to the narrowing of the earnings distribution and decrease in the returns to education. The earnings gap expansion since the 1980s is fully explained by the stretching of the earnings distribution.17
Third, the Bayer and Charles distributional analysis allows the authors to describe the differential experiences of African American workers at different points of the earnings distribution. At the bottom, sharp increases in incarceration, labor-force nonparticipation, and unemployment since 1970 have severely retarded the ability of black men to compete in the labor market. At the 90th percentile, black men made significant “positional” gains. Their relative improvement accounts for the majority of the earnings gap’s narrowing at the 90th percentile from 1940 to 2014. These relative gains were largely due to blacks’ catching up in their college completion rates during the 1960s and 1970s.18
To summarize, blacks made significant gains during the 1960s and 1990s. Losses occurred during the late 1970s and into the 1980s. Losses also occurred after the early 2000s. The collective nature of the losses outweighed the gains during the 1990s boom and current expansion. The conclusion in Wilson and Rodgers (2016) is sobering. The typical African American’s relative earnings are where they were in 1979. The conclusion in Bayer and Charles (2016) is even more sobering. When mass incarceration is taken into account, the relative earnings of the typical black male are what they were in 1950.
CROSS-SECTION STUDIES
A wealth of analysis seeks to explain racial inequality at a point in time. These cross-section studies are the source of the greatest controversy about why racial wage inequality exists. One strand of the evidence tends to support the view that racial differences in pre-labor-market experiences (such as family, school, and neighborhood) are the key driver of the black-white gap.19 These studies all use the National Longitudinal Study of Youth and the composite Armed Forces Qualification Test (AFQT) score to estimate how much of the black-white wage gap is explained by cognitive skill differences (O’Neill 1990; Maxwell 1994; Neal and Johnson 1996; Carneiro, Heckman, and Masterov 2005). All or most of the gap vanishes when the test score is added.20 Premarket factors are not independent of racism and discrimination. The observed differences in premarket factors could be due to housing segregation, poverty, crime, poor public services, and schools with inadequate resources, all linked to persistent racist policies.
Another strand of evidence supports a view that these studies overstate the importance of pre-labor-market experiences and places greater emphasis on labor-market experiences. Rodgers and William Spriggs (1996, 2002) were the first to show that an AFQT score adjusted for racial differences in age at time of the exam and promotion rates explains much less of the earnings gap. The Rodgers and Spriggs adjusted score removes the large unexplained test score gap between blacks and whites.21 What is the source of the large unobservable gap in test scores? The answer depends on how one interprets the results in Rodgers and Spriggs (2002). Derek Neal and William Johnson (1996) would interpret it as racial differences in premarket factors, but both Rodgers and Spriggs (1996) and Darity and Patrick Mason (1998) label the unobservable gap in test scores as a racial bias in the AFQT’s ability to predict wages.
Two studies pursue this debate further, describing the framework on which this finding stands as one based on the notion that blacks have more pre-labor-market or cultural deficiencies than whites (Darity and Mason 1998; Myers and Chung 1998). Family, school, and neighborhood differences between black and white families (such as the larger share of black families led by single women) contribute to earnings inequality. They estimate earnings gaps between black and white family heads and find that rising female headship among black families explains a small portion of the 5 percent increase in the racial earnings gap between 1976 and 1985.
Alan Krueger (2017) adds additional support to the Darity, Myers, and Chung (1998) critique of the “deficiencies” hypothesis. Krueger links the opioid crisis to the decline in the participation of prime-age men. He shows that a larger share of prime-age white men are more likely to report having at least one of six functional disabilities (35.8 percent) than prime age, African American men are (32.3 percent).
In Krueger’s regressions, blacks have a 2.8 to 4.2 percentage point lower odds of taking pain medication. Krueger goes on to connect the high rate of pain medication among men out of the labor force to Anne Case and Angus Deaton’s (2015, 2017) findings of a rise in mortality for middle-age whites due to accidental drug poisonings, especially from opioid overdoses. Unfortunately, Krueger does not estimate a model with just race such that we can see whether the black-white male participation gap narrows when opioid use is added. This is not an issue with prime-age black and white women because they have the same participation rate.
The bottom line here is that the deficient family hypothesis is not as strong a factor as it was in the past for explaining racial differences (Darity, Myers, and Chung 1998). Black and white families are both suffering from excessive drug use and addiction. The opioid epidemic has brought to whites many of the problems that the crack epidemic brought to black families. Clearly, additional data on premarket and labor-market factors are needed so that social scientists can do a better job of explaining the large unexplained racial test score gap. The truth lies somewhere in the middle.
In fact, Roland Fryer, Devah Pager, and Jörgen Spenkuch (2013) show that discrimination and premarket factors both play key roles in explaining racial inequality. In a study of New Jersey workers at the end of the Great Recession, they find that discrimination accounts for at least one-third of the black-white wage offer gap of 33 percent. That is, after controlling for previous wage and a host of other characteristics, African Americans’ offered wages are approximately 16 percent less than whites’ offered wages. Premarket factors play a role as well. They conclude that “alleviating racial inequality may take a combination of policies to both eliminate barriers to investing in education and other premarket skills and enforce antidiscrimination policies, so that minorities are rewarded for those skills” (2013, 638).
Researchers use field and quasi experiments to test for discrimination. In-person audit studies provide the most direct and compelling evidence that discrimination remains a key feature of the black experience. They compare the odds of receiving a callback or job offer between matched pairs of black and white testers. These testers pose as real-world applicants (see, for example, Turner 1993; Bendick, Jackson, and Reinoso 1994; Pager 2003; Pager, Bonikowski, and Western 2009).22 The consensus of these studies is that black testers face greater difficulty in the hiring process, which is attributed to discrimination. However, as James Heckman (1998) discusses, the methodology depends on the crucial assumption that the tester pairs not only are similar in observables but also share the same distribution of unobservable characteristics. Heckman also questions whether the marginal employer treats black and white candidates differently.
Correspondence studies address the first critique (Firth 1981; Esmail and Everington 1993; Bertrand and Mullainathan 2004). For example, Bertrand and Mullainathan distributed approximately five thousand fictitious resumes with randomly assigned black- or white-sounding names to more than 1,200 Boston and Chicago help-wanted ads. They found that resumes with white-sounding names received about 50 percent more follow-ups. These studies still do not address whether the marginal employer treats black and white candidates differently.
Similar to the trend analysis, a great deal of research focuses on incarceration’s impact on labor-market outcomes. In fact, numerous studies find significant and persistent adverse impacts on the post-incarceration earnings of all individuals, especially blacks. In an early study, John Lott (1992) estimates the post-incarceration total monetary penalty associated with a drug conviction. Using District of Columbia data on heroin and cocaine distribution, importation, and possession violations, Lott finds that higher income is associated with higher fines but lower prison time.23
From the mid-1980s to late 1990s, the prison and jail population increased by 7.3 percent (Western and Pettit 2002). By 1996, a larger number of black male high school dropouts between the ages of twenty and thirty-five were in custody than in paid employment. Over this same period, wage inequality increased, especially among minority men. To estimate the wage impacts, Western and Pettit regress wages on age, dummy variables that capture time incarcerated, and incarceration status. The evidence indicates that ex-offenders earn 7.0 percent less than non-offenders. When individual effects are included, ex-offenders earn 19.0 percent less than their counterparts. Incarceration wages after incarceration increase at a slower rate, causing the wage gap between ex-offenders and non-offenders to increase with age. Further, incarceration expands wage inequality among blacks and Hispanics and between blacks and whites by 8.0 to 9.0 percent.
Several Devah Pager studies are excellent examples in revealing the existence of a large adverse employment and earnings impact associated with having a criminal record (Pager 2003, 2007; Pager, Bonikowski, and Western 2009). Data from Washington State Department of Corrections and Unemployment Insurance Records reveal the compounding impact that incarceration has on wage growth (Lyons and Pettit 2011). Based on growth curve models, post-incarceration wage growth was 21 percent slower for blacks than for whites. Race differences in wage growth were insignificant before incarceration. Focusing on incarcerated men from 1990 to 2000, Christopher Lyons and Becky Pettit compare before and after jail experiences. A key finding is that wage returns to work experience after incarceration are about 40 percent lower. This may be due to the fact that African Americans tend to work in lower wage retail and service occupations after incarceration. Sentence length has an adverse impact on wage growth.
Blankenship and Del Rio Gonzalez (2018) continue to find that blacks are more likely to report negative effects from incarceration on well-being. They study the experiences of 302 men and women on probation or released from jail in New Haven, Connecticut. They find that blacks have fewer convictions but are incarcerated twice as often as whites. Juvenile incarceration rates and other juvenile infraction rates are higher for blacks and they are more likely to be incarcerated for drug charges. Whites are more likely to be released when caught. Blacks report more adverse impacts on education, on dropping out of school, and loss of longest job. Incarceration leads to home evictions and greater homelessness.
THE POLICY FRAMEWORK
This review confirms the seminal work by Freeman and his colleagues (Freeman 1973; Bound and Freeman 1992). The labor-market experiences of African Americans are not monolithic. In fact, the African American experience in the U.S. labor market has become increasingly more diverse. No single narrative fully explains persistent racial employment and earnings inequality.
For example, young out-of-school blacks have weaker cognitive or soft skills, face discrimination, and are more likely to come into contact with the criminal justice system. Those in the Midwest were hurt by the region’s de-industrialization, and the use of robots is shown to have adverse impacts on employment of less-educated minority Midwestern workers (Freeman and Rodgers, forthcoming). Third, black college graduates faced growing discrimination in the 1980s and have been unable to recover from it, largely because of a retrenchment in policy after the Obama administration.
Thus no silver bullet policy or policy agenda will eradicate or substantially reduce the discrimination. A coordinated and comprehensive set of labor supply, labor demand, and institutional policies must be implemented and a long-term commitment to the framework and its policies is needed. Loprest, Nightingale, and Spaulding (2019) come to the same conclusion for addressing declining labor-force participation of young adults. Thus overlap in the policies we recommend for consideration is substantial.
Based on this heterogeneity of policy needs, the following section presents a policy scaffolding for addressing the nation’s persistent racial employment and earnings inequality—a menu of race-neutral and race-specific approaches that will assists all Americans, especially African Americans.24
BLUEPRINT FOR NARROWING RACIAL INEQUALITY
Building Human and Social Capital
Provide children and parents with the resources and support such that they have a “high quality” preschool experience.
Professionalize preschool teachers. Raise their status to that of K–12 teachers.
Expand community schools and strengthen economic and financial literacy.25
Expand teenagers’ employment, internship, and volunteer opportunities such that they acquire soft skills. A large literature shows a strong link between teenage employment and education, wages, and contact with the criminal justice system.26
Increase investment in career technical occupations, such as auto technicians, HVAC technicians, and plumbers. Simultaneously reduce the stigma that youth and parents have with working in such careers. Stress that these jobs are less likely to be outsourced, providing greater economic security that many Americans crave.27
Convene policymakers, researchers, practitioners, current high school students, current college students, and recent graduates to identify and recommend ways to narrow the black-white wage gap among college graduates.
Create a balanced policy that not only reduces child support arrearages, but also preserves and builds a noncustodial parent’s connection to their children.28
Improving Opportunity
Expand the demand for workers in “innovation” areas of the economy and increase education and training investments in these sectors.29 Use the forfeiture funds program of the U.S. Department of Justice (DOJ) to fund workforce development in these sectors. Explore whether the DOJ’s victim compensation and equitable sharing programs can be an additional source of funding for providing education and training to communities affected by crime. This approach is consistent with the program’s strategic goal no. 3, which is to compensate victims with recovered assets.30
Strengthen social safety nets and rebuild the capacity of nonprofits to assist vulnerable families.31 The Great Recession was so severe that a new set of families needed assistance. Families and communities are resilient, but recent experiences suggest that in many communities, a job loss, a natural disaster, or failed public water sources now seem to have larger individual, family, and community impacts than in the past.
Reduce the difficulty African Americans face with either commuting or moving to suburban employment opportunities.32
At the time of incarceration, assist ex-offenders, especially nonviolent drug offenders with returning to school, obtaining employment and training, driving, and voting. The First Step Act includes a few of these elements.
An infrastructure initiative might be implemented. If so, urge elected officials to run the procurement process through federal and state governments and add provisions for education and job training. Doing so will ensure that all workers, but especially minorities and women, have a chance to compete for the infrastructure jobs (Rodgers 2016).
Strengthen the ability of workers to bargain with their employers by combating state laws that restrict public employees’ collective bargaining rights or the ability to collect “fair share” dues through payroll deductions, pushing back against the proliferation of forced arbitration clauses that require workers to give up their right to sue in public court.
Secure greater protections for freelancers and workers in gig employment relationships.
Lessening Inequality
Urge Congress and the president to increase the minimum wage to $15.00 per hour, create new work scheduling standards and rigorously enforce wage laws aimed at preventing wage theft.
Urge elected officials to ensure that the Federal Reserve pursues monetary policy that targets full employment, with wage growth that matches productivity gains.
Counter the Trump administration’s Labor Department efforts to chip away at its practices that help to ensure safe and fair workplaces for all Americans.33
Respond to the Trump administration’s efforts and build a New Labor Policy based on the four features that Thomas Kochan and William Kimball (2019) lay out.34
Relatedly, the 2020 presidential election has sparked a public conversation about reparations. Again, this is not new. On January 3, 2017, Representative John Conyers Jr. introduced House Resolution 40, which calls for the creation of a commission that would study and consider a national apology and proposal for reparations for the institution of slavery. This would be an interesting and important national conversation.35
Fighting Discrimination
Under the leaderships of the U.S. Census Bureau and BLS, identify the “unobservable measures” that impact racial inequality and include them in national surveys.
Urge Congress to get the EEOC to develop metropolitan area measures of discrimination that could be linked to individual records in the federal surveys so that researchers could directly assess the role that local area discrimination plays.
Urge members of Congress to request a reversal in the decision of the Trump administration’s Justice Department to withdraw the Obama administration guidance documents that encouraged schools and colleges to promote diversity through their admissions process.36 Although affirmative action policy has not changed, the pulling back of the documents sends the message to colleges and universities that the administration is placing less value on the existence of diverse learning environments.
Advocacy groups need to gear up for a Supreme Court fight because Supreme Court Justice Anthony Kennedy, who sided with the court’s liberal block in 2016 Fisher v. University of Texas, retired and conservative justice Brett Kavanagh took his place.
In 2018, the EEOC experienced its first budget increase in eight years. Urge members of Congress to request more resources for the EEOC. The administration’s FY18 EEOC’s budget would have continued flat at $365 million, but advocacy groups’ pressure to address sexual harassment led to Congress adding an additional $13 million to the budget that the president signed in early spring of 2018.
Reduce the large backlogs and the number of days for an EEOC claim to get resolved. They remain much higher than a decade ago.
LOOKING FORWARD
Implementing any type of comprehensive agenda to address racial inequality will be a challenge. President Trump’s behavior toward women, minorities, immigrants, and people with disabilities has empowered and emboldened not only conservatives, but also neo-Nazi and alt-right groups to publicly express themselves. The president has also framed the national debate on income inequality solely in terms of national and personal security. He is capitalizing on the fact that many Americans feel bullied by globalization and technology, and are nervous with how the country is becoming more racially and ethnically diverse. If left unchecked, public support for even race-neutral remedies that address overall income inequality will weaken.
Although there are no signs of recession,37 the economy has peaked, thus “maxing out” its ability to narrow racial inequality.38 In fact, the policy direction that the Trump administration is taking has the potential of further expanding black-white employment and earnings gaps. There seems to be little appetite by this administration and the Republican Congress to increase the federal minimum wage, or provide workers with greater voice, such as strengthening the ability of unions to collectively bargain. Evidence is now emerging that the tax reform benefits that were implemented in early 2018 have gone to corporations in the form of stock buybacks, and not capital investment. Many analysts predict that the household tax cuts will favor high-income families and thus not have large micro effects on middle- and lower-income families (see, for example, Gale et al. 2018; Tankersley 2018). Further, congressional support to significantly invest in what the United Nations calls “human priority” investments is scant. These are investments in human and social capital. In fact, the Senate majority leader pointed to “entitlements,” an important social capital investment, as the “key cause of rising federal deficits, and blamed Democrats for refusing to go along with proposals to cut spending by Medicare, Medicaid and Social Security” (Miller 2018).
These examples of the apparent inability to compromise and support productivity-enhancing investments do not bode well for the future. The signing of the First Step Act into law may provide a glimmer of hope. The measure is significant because it is one of the few substantive bipartisan pieces of public policy that have become law during the Trump administration. The law has four main components. It makes retroactive the reforms enacted by the Fair Sentencing Act of 2010, which lessened the disparity between crack and powder cocaine sentences at the federal level. The Marshall Project estimates that this could affect nearly 2,600 federal inmates. Second, the law reduces mandatory minimum sentences under federal law. Third, it increases “good time credits” that federal inmates can accumulate. Fourth, it allows inmates to get “earned time credits” by participating in more career, technical, and rehabilitative programs. The goal is to address prison overcrowding and use education programs to reduce the likelihood that an inmate will commit another crime once released and, as a result, reduce both crime and incarceration in the long term.
The act has its limitations. It does not reduce the number of offenders sent to prison. It does not address length of sentences or mandatory minimums. It will have limited impact because the reform focuses on federal prisons, which house only 13 percent of the nation’s prisoners. The rest are in state facilities. Undocumented immigrants are not able to cash in their time credits for early release (see Haynes 2018).
Finally, the system will use an algorithm to determine who can cash in earned time credits. Critics are concerned that this algorithm could perpetuate racial and class discrimination. For example, an algorithm that prohibits offenders from earning credits due to their past criminal history may pass over African Americans and poor offenders, who are more likely to be incarcerated for crimes even though they are not more likely to actually commit those crimes. Clearly, a second chance act is needed.
Along with legislative gridlock, the economy and public policy continue to evolve in ways that are adversely affecting the income of all Americans. Bound and his colleagues (2019) find a major decline in appropriations to public universities. This decline in public resources could adversely impact the ability to narrow racial employment and wage gaps because it might reduce blacks’ access to its major source of higher education opportunities. Unfortunately, Bound and his colleagues do not directly answer this question. Future research perhaps needs to explore whether the decline in appropriations has impacted black enrollment.
Shifting to the labor market, the adoption and diffusion of technology in the workplace has many workers frightened. Employers are implementing robotics and artificial intelligence in their production processes. Little is known about whether it is having disparate effects across race and ethnicity. To begin to assess the impacts, Freeman and Rodgers (forthcoming) estimate the impact that industrial robots have on the wages and employment on American workers by race, ethnicity, age, and gender.39 First, they find that industrial robot intensity does not explain racial, ethnic, and gender differences in employment and wages. Second, their measure of a metropolitan area’s robot intensity is associated with lower employment and wages in the East North Central (ENC) region. The negative effects are the worst for all young less-educated ENC men and women.
Based on these estimated relationships, Freeman and Rodgers (forthcoming) predict that without the current economic expansion, changes in metropolitan area industrial robot intensity would have led to ENC youth employment-population ratios to be 10 to 15 percentage points lower than their actual 2017 values.40 Even though the effects are concentrated in a particular region and demographic group they are nontrivial. Like George Borjas and Freeman (2019), they conclude that the continued rapid growth in the use of robots could have adverse effects in the future.
Support is growing for public policies that reduce inequality. Abraham and Houseman (2019) show that informal work has become quite prevalent throughout the economy and that minorities are more likely to report that earnings from informal jobs are important to household income. A major policy concern with these employment arrangements is the lack of benefits. Future work needs to examine how informal work contributes to the racial wage gap.
Many families, especially minority families, face the challenges of balancing work and family responsibilities. One way to assist all families, black families in particular, is to implement family-friendly policies such as paid family leave, expanded public or subsidized childcare, and employer mandates for scheduling control and flexibility. Elizabeth Doran, Ann Bartel, and Jane Waldfogel (2019) show that these polices can assist all families, especially low-income families, and help them create a balance between work and home responsibilities. Although race is not a focus of their study, they do find that blacks have greater access to employer provided or subsidized childcare than whites. They also have more access to paid parental leave than whites. Blacks do have less access to regular schedule flexibility, but this seems to be fully explained by racial differences in marital status, education, household type, and type of employment.
More of the public support making changes to strengthening the Fair Labor Standards Act, such as increasing the federal minimum wage or adjusting overtime laws. Supporters argue that these policy shifts are good race-neutral ways to address racial inequality. Charles Brown and Daniel Hamermesh (2019) conclude, however, that the effect of these efforts would yield small improvements. This is consistent with my view that if as a society, we are serious about reducing persistent racial inequality, a coordinated and comprehensive set of labor supply, labor demand, and institutional policies must be implemented and that a long-term commitment must be made to the framework and its policies.
SUMMARY AND CONCLUSIONS
This article demonstrates that employment and earnings inequality between blacks and whites continues to be large and persistent. By one measure, African American relative earnings are where they were in 1979. If incarceration and nonparticipation are included, the relative status takes us back to 1950. Further, since the Great Recession, the gaps have expanded. Discrimination continues to contribute to the lives of women and minorities, but other factors also drive the disadvantages that African Americans so persistently face. Racial differences in education, experience, type of work, marital status, industry of employment, and coverage by collective bargaining agreements still contribute to the large and persistent employment and earnings gaps. However, the erosion in the relative position of all low-wage workers regardless of race and ethnicity plays a key role. The effects of incarceration on ex-offenders and their communities, especially on less-educated and less-skilled African Americans, continues to be quite substantial. The pay gap among black and white college graduates also remains largely unexplained. A variety of political, cultural, and economic headwinds, if left unchecked, will cause inequality to expand.
Continuing the current path and pace at which we are addressing the nation’s racial inequality will have serious consequences. A few of these, just a few, follow.
Larger class sizes, fewer teacher resources, and continued parental hurdles to participating in the development of their children.
Lower labor-force participation, elevated worker idleness, higher crime and incarceration rates, and higher recidivism rates.
Elevated levels of hunger, food insecurity, stress, anxiety, and clinical depression.
Lower individual and community productivity, and thus competitiveness.
Slower city, state, regional, and national economic growth.
In the aggregate, these consequences mean lower U.S. productivity and lower economic growth, that is, a less competitive economy, and a competitive economy is a must in today’s globalized world. All of this adds up to a lower standard of living for all Americans.
As former National Urban League president Hugh Price said, “We have many of the solutions at our finger tips. We need the political courage to make the necessary investments that make workplaces safe and fair for all workers.”41 Building on Price’s statement and the Kerner Commission’s call to action from fifty years ago, these solutions need a moral recommitment to their strengthening. Like Hugh Price, Kochan and Kimball (2019) believe that we have all the ideas needed to improve policy, but we need a major public awareness campaign that educates Americans about the weak state of labor policy. The evidence in this article and the rest of the articles in this issue provides the content.
FOOTNOTES
↵2. See U.S. Bureau of Labor Statistics, “Union Members 2018,” News Release, USDL-19-0079, https://www.bls.gov/news.release/pdf/union2.pdf (accessed July 18, 2019).
↵3. The Economic Policy Institute (EPI) estimates that as of December 2018, the U-6 rates are 11.9 percent for blacks, 10.1 percent for Latinos, and 6.3 percent for whites. The estimates for African Americans and Latinos are the lowest since 1989, when estimates were first made. The series record for whites is 5.6 percent from July 2000 to January 2001. Economic Policy Institute, State of Working America Data Library, “Underemployment,” 2019.
↵4. See, for example, Freeman 1973; Freeman and Rodgers 2000.
↵5. William Spriggs and Rhonda Williams conclude that “economic growth and black human capital acquisition are not sufficient to eradicate the two-to-one black white unemployment gap” (2000, 203).
↵6. Some studies on youth (sixteen-to-twenty-four-year-olds) use a measure of idleness—the percentage of the civilian population that is not employed, searching for a job, or enrolled in school.
↵7. The U-6 unemployment rate includes individuals working part time who want to work full time, as well as those who have actively stopped searching for work but say if they were offered a job, they would take it.
↵8. Freeman and his colleagues show that young less-educated black men’s employment-population ratios are quite sensitive to macroeconomic conditions (Freeman 1973; Freeman and Rodgers 2000). Freeman and Rodgers (forthcoming) continue to find that a 1 percent decrease in the unemployment rate is associated with a bigger increase in the employment-population ratio of young, non-college-educated African American men.
↵9. For black men, the rate jumped from 18 percent in 1960 to 37.8 percent in 2010; for white men, the rate moved from 7.9 to 18.6 percent.
↵10. The relative disadvantage that African American men face could be biased downward. Bayer and Charles can only infer current incarceration from their institutionalization variable. Because of that limitation, it is not possible to identify men who are unable to find work because of prior incarceration. As a result, a nontrivial share of the increase in the racial gaps in labor-force participation and unemployment could also be due to the effects of mass incarceration.
↵11. Derek Neal and Armin Rick (2016) also provide a detailed analysis of the causes of the increase in the criminal justice system’s severity of punishment. They show that changes in sentencing policy are the primary driver for prison population growth. The movement to harsher sentencing policies disproportionately affected African Americans. The latter is startling given that most of this shift did not target blacks or crimes they commit relatively more than whites.
↵12. The deterioration of regional economies has been bad news for less-educated blacks. Bound and Holzer (1993) show that the contraction of the nation’s manufacturing sector had a more adverse impact on the employment of black men. Inasmuch as they could move to better regions, at least from a wage standpoint, this is no longer possible, or as feasible. Racial wage gaps have grown most in the Midwest and least in the South.
↵13. Over the past fifty years, the EEOC’s jurisdiction has grown and now includes the Equal Pay Act of 1963 (included in the Fair Labor Standards Act), Pregnancy Discrimination Act of 1978, Age Discrimination in Employment Act of 1967, Sections 501 and 505 of the Rehabilitation Act of 1973, Titles I and V of the Americans with Disabilities Act of 1990, Title II of the Genetic Information Nondiscrimination Act of 2008, and Lilly Ledbetter Fair Pay Act of 2009. These laws protect individuals from employment discrimination on the basis of race, color, religion, sex, national origin, age, disability, genetic information, and family medical history. They make it illegal to retaliate against a person for opposing employment discrimination, filing a discrimination charge, or participating in an employment discrimination investigation or lawsuit (https://www.eeoc.gov/eeoc/plan/2017budget.cfm#_Toc442168099, accessed September 7, 2019).
↵14. On the Great Recession’s impact by race, see Vuolo, Uggen, and Lageson 2017.
↵15. The estimates for 1979 to 1985 in Wilson and Rodgers (2016) are similar to those in Bound and Freeman (1992) and Rodgers (2006). During the first half of the 1980s, the main sources of expansion of black-white wage gaps among new-entrant men and women were a worsening in discrimination or growing differences in unobserved skills, and the decline in relatively well-paying jobs for workers with less than a college degree. Among new-entrant men, these effects far outweighed the positive effects from narrowing the education gap. Increased discrimination was the sole cause of growing racial wage gaps for college graduates during this period. Between 1979 and 1985, racial wage gaps widened most in the Midwest and among men working in the manufacturing industry. Shifting patterns of employment across industries and occupations also contributed to growing racial wage gaps among new-entrant men during the period.
↵16. Charles Brown (1984) developed a technique for adjusting for the greater censoring in African American men’s wages. He also finds that the improvement is dampened when the correction for African American labor-market dropouts is incorporated in the estimates. Western and Pettit (2000) also develop an adjustment technique. It too reduces the employment rates of African Americans. They show that from 1982 to 1996, employment of young black male high school dropouts falls despite its being a period of low overall unemployment. They conclude that the standard CPS data which do not include incarceration data downward bias the black-white employment gap among young high school dropouts by 45 percent.
↵17. They do not find this result. Numerous studies show that increased earnings inequality contributed to a decline in the economic prospects of low- and middle-wage African Americans (Darity and Mason 1998; Juhn, Murphy, and Pierce 1991; Bound and Freeman 1992; Rodgers 2006; Wilson and Rodgers 2016).
↵18. Bayer and Charles speculate that these relative gains are due to the elimination of exclusionary practices that existed during the 1940s and 1950s in many professions, occupations, and higher education institutions. Since the 1980s, they argue, “affirmative action in college admissions may have better equalized effective college quality for high-ability black students, shrinking racial differences in unobserved skills within the upper part of the earnings distribution” (2016, 29–30).
↵19. In early work, Cordelia Reimers (1983) concludes that although no significant labor-market discrimination was found, it occurred in the acquisition of characteristics, such as education.
↵20. Nan Maxwell (1994) concludes that the main source of black-white wage differential is the racial difference in quality rather than quantity of schooling. The racial gap in basic skills learned in school explains two-thirds of the wage gap. After correcting for the error in Maxwell, school quality’s importance is reduced.
↵21. Rodgers and Spriggs (1996) do not say that the composite test score is culturally biased. In a parallel study, Rodgers is explicit that when the math and verbal scores are used to predict wages, the verbal score predicts only the wages of African Americans, and math scores predicts only the wages of whites.
↵22. In related work, Ian Ayres and Peter Siegelman (1995) conducted more than three hundred paired audits at new car dealerships. They find that car dealers gave white men significantly lower price quotes than to African American or female test buyers who used identical, scripted bargaining strategies. David Neumark, Roy J. Bank, and Kyle D. Van Nort (1995) perform an audit study of sex discrimination in hiring. They had matched pairs of men and women apply for waiters and waitresses in a variety of Philadelphia restaurants. Women had lower probabilities of a job offer and receiving an interview.
↵23. For example, a two-standard deviation income increase leads to a monetary penalty that is five times higher. If inflation-adjusted income pre-sentence equals $19,601 then income falls by 12.0 percent; however, if pre-sentence income equals $35,468 then income falls by 48.0 percent. Finally, conviction further reduces income.
↵24. Many of these recommendations were presented in Wilson and Rodgers (2016).
↵25. According to the Coalition for Community Schools, a community school uses a public school as the anchor and then creates partnerships between the school and other community resources. It is a one-stop approach to academics, health and social services, youth and community development, and community engagement. The vision is that this integrated approach leads to improved student learning, stronger families and healthier communities (see “About,” http://www.communityschools.org/aboutschools/what_is_a_community_school.aspx, accessed July 19, 2019).
↵26. For evidence on the links, see Pager 2003. On June 26, 2018, the Department of Labor’s Employment & Training Administration awarded $84.4 million to forty-one nonprofits and state and local governments with the goal of helping Americans transition from the justice system to the workforce. The projects assist eighteen-to-twenty-four-year-old adults who have been involved in the juvenile or adult justice system, or adults age twenty-five and older who were formerly incarcerated (see https://www.dol.gov/newsroom/releases/eta/eta20180626, accessed July 19, 2019).
↵27. I learned about this stigma when I served on the boards of the Williamsburg–James City County Public Schools and the New Horizons Regional Education Centers. The latter provides career technical education to students from six surrounding public school districts. Students can study automotive technology, construction technology, engineering–manufacturing technology, health sciences, human services, information technology, and public service. Staff had to work with local school district counselors to sell the programs to students. Many in these communities thought that the “best” path for their children was a four-year degree. In the past, the perception was that many children sent to the once-called vocational school were trouble makers. Career technical was seen as a last resort. A 2017 Brookings report provides an outstanding overview of the challenges with promoting career technical education (Jacob 2017). The federal government has a wealth of resources that seek to help youth make decisions about careers (see, for example, “Career Exploration & Skill Development,” https://youth.gov/youth-topics/youth-employment/career-exploration-and-skill-development, accessed September 7, 2019).
↵28. States are required to update and evaluate their child support guidelines every four years. Recently, the federal government required the use of labor-market data to assist in setting “reasonable” orders, especially for low-income obligors. The regulation was amended in December 2016 and reads as follows: “a State must consider economic data on the cost of raising children, labor market data (such as unemployment rates, employment rates, hours worked, and earnings) by occupation and skill-level for the State and local job markets, the impact of guidelines policies and amounts on custodial and noncustodial parents who have family incomes below 200 percent of the Federal poverty level, and factors that influence employment rates among noncustodial parents and compliance with child support orders” (U.S. Department of Health and Human Services 2016).
↵29. The economic historian Joseph Schumpeter developed the concept of creative destruction. As one industry contracts, other industries expand. Workers with the assistance of the public, private, and nonprofit sectors shift from contracting to expanding industries.
↵30. This was a suggested during a presentation at Newark’s Institute for Social Justice. The Asset Forfeiture Program’s mission is to punish and deter criminal activity by depriving criminals of property used in or acquired through illegal activities; promote and enhance cooperation between federal, state, local, tribal, and foreign law enforcement agencies; recover assets that may be used to compensate victims when authorized under federal law; and ensure the program is administered professionally, lawfully, and in a manner consistent with sound public policy (U.S. Department of Justice 2018). The federal government should consider using recovered assets to further fund education and training programs in areas where criminal activity has occurred. This would be a way of compensating and increasing the safety of communities with large drug and gang activity.
↵31. With respect to the Unemployment Insurance system, see, for example, O’Leary and Kline 2016.
↵32. This phenomenon is called spatial mismatch. Low-income urban residents, especially minorities, have difficulty commuting or moving to the suburbs, where job growth is occurring.
↵33. EPI developed the Frances Perkins Project, which tracks the administration’s efforts to erode wages and worker protections (https://www.epi.org/blog/how-president-trump-and-congressional-republicans-are-undercutting-wages-and-protections-for-working-people, accessed September 7, 2019).
↵34. The new labor policy framework should be based on four foundations. First is freedom of association. Second is a labor policy integral to a set of national economic policies that support high and growing productivity. This growth accompanies wage growth and economic security. Third is a worker voice that has no single approach or set of strategies. Fourth is labor policies that support high-quality labor-management relationships that generate worker voice and strong economic performance.
↵35. The argument for providing African Americans reparations is not new. Robert Browne (1990) presents a variety of present value estimates for the slavery period but also during Jim Crow and to the present. Thomas Craemer (2015) reviews these and other estimates and concludes that reparations debt ranges from $2.5 to $15.0 trillion dollars. He compares the amount to the Bush-Obama stimulus package of $1.6 trillion and the size of the U.S. economy of $15.0 trillion. Even if the estimates are placed in per capita terms, the reparations estimates are larger than the stimulus package and GDP per capita.
↵36. “Attorney General Jeff Sessions Rescinds 24 Guidance Documents,” press release, July 3, 2018, https://www.justice.gov/opa/pr/attorney-general-jeff-sessions-rescinds-24-guidance-documents (accessed July 19, 2019).
↵37. Some economists are citing evidence that the yield curve has begun to flatten, inflation has ticked upward, home sales are starting to decline in key markets, and credit card debt and late payments are rising (see Sean Motley, “6 signs we’re closer to the next recession that you think,” USA Today, September 5, 2018, https://www.usatoday.com/story/money/markets/2018/09/05/recession-signs-getting-closer/37630241, accessed July 19, 2019).
↵38. Freeman and Rodgers (forthcoming) find that for metropolitan areas where jobless rates fell by at least 5 percentage points from 2009–2010 to 2016–2017, men benefit the most. All men experience an increase in their employment-population ratios; women in these areas have modest employment gains. None experience increases in their inflation-adjusted wages. Further, the employment-population ratios of young black men increase only from 52.5 to 59.5 percent. Those for young black women increase from 50.4 to 54.7 percent, which leaves considerable room for improvement.
↵39. They find that prior to the Great Recession, industrial robot intensity trended upward. However, during the recession robot intensity plummeted. Since 2009, robot intensity has sharply increased. The East North Central Census region which contains, Michigan, Ohio, Indiana, Illinois, and Wisconsin have the highest robot intensities. Minorities and youth with no more than a high school diploma live in metropolitan areas that have similar industrial manufacturing robot intensities as whites and adults. The top ten metropolitan areas with respect to robot intensity are (1) Los Angeles-Long Beach-Santa Ana, California; (2) Chicago-Naperville-Joliet, Illinois; (3) Houston-Baytown-Sugar Land, Texas; (4) Phoenix-Mesa-Scottsdale, Arizona; (5) Detroit-Warren-Dearborn, Michigan; (6) Milwaukee-Waukesha-West Allis, Wisconsin; (7) Philadelphia, Pennsylvania, Camden, New Jersey, and Wilmington, Delaware; (8) San Jose-Sunnyvale-Santa Clara, California; (9) Indianapolis, Indiana; and (10) Cleveland-Elyria, Ohio.
↵40. Their estimates do not control for the impact of productivity on manufacturing employment. Without robots, fewer firms might have been able to be competitive against internal suppliers, leading to less manufacturing employment. Thus their model may overstate the impact of robots on overall employment levels of different populations.
↵41. Price said this in a conversation about the National Urban League’s annual publication, “The State of Black America.”
- © 2019 Russell Sage Foundation. Rodgers, William M., III. 2019. “Race in the Labor Market: The Role of Equal Employment Opportunity and Other Policies.” RSF: The Russell Sage Foundation Journal of the Social Sciences 5(5): 198–220. DOI: 10.7758/RSF.2019.5.5.10. I thank Peter Ni and Jenna Bendinelli for their research assistance. Direct correspondence to: William M. Rodgers III at wrodgers{at}ejb.rutgers.edu, Heldrich Center for Workforce Development, Rutgers University, New Brunswick, NJ 08901.
Open Access Policy: RSF: The Russell Sage Foundation Journal of the Social Sciences is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
REFERENCES
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.