The critical issue that persists in understanding ethnic and racial differences in employment outcomes (possession of a job, compensation for the job, likelihood of advancement or promotion) is the relative role of discrimination in explaining disparities. The Samuel DuBois Cook Center on Social Equity aims to design effective procedures for preventing or mitigating the effects of employment discrimination. To design the best policies requires an understanding of how the mechanism of discrimination operates.

Discrimination in employment can be detected on the basis of self-reports, a content study of job advertisements in newspapers or other media, statistical decomposition methods, or field experiments that audit for its presence. Self-reports are subject to respondent self-censorship, cognitive dissonance or error. A comparison of independently derived estimates of the degree of wage discrimination individuals face with their self-reports of discrimination, finds that blacks substantially underreport their exposure to discrimination in employment while whites substantially over-report their exposure. Measuring discrimination is complicated if group distinctions are based exclusively on self-reports of race or ethnicity. For example, Latinos, regardless of skin shade, report their race as white at high rates and rarely report their race as black. It leads to inaccurate estimates of the magnitude of discrimination.

Therefore, data sets that enable researchers to make distinctions based upon skin shade will prove more useful in locating evidence of discrimination than data sets that rely exclusively on self-reported race.

Properly designed field experiments are the gold standard for detection of discrimination. However, they do not afford a straightforward procedure for calculating the magnitude of the effects on wages or lifetime earnings. 

  • Correspondence tests that involve the submission of artificial letters of application with faux resumes to advertised positions are especially powerful but only effective in settings where names and addresses of applicants signal group identity. 
  • Trained actor studies involving face to face contact between testers and employers also are potentially powerful, but always are subject to the complaint that the testers’ beliefs about the presence or absence of discrimination can affect their behavior and distort the results of the experiment (Darity 2010). 

The first task for the Cook Center will be application of these methods -- the self-report, statistical decomposition, and field experiment -- at a common set of geographical locations to assess the consistency of information about discrimination generated by each approach. Careful attention will be given to how identity is expressed and understood by job applicants, employees and by others who can affect their employment outcomes. Researchers will gather detailed information on respondents’ or testers’ phenotypical characteristics, including their skin shade. 

Ultimately, Cook Center researchers will design a context-specific research protocol for detection of discrimination in employment. They will also design a measurement of the magnitude of discrimination directed against people with multiple stigmatized identities. Do multiple identities have an adverse effect on earnings and employment outcomes?


Stratification Economics and a Gendered Analysis Around Closing the Wealth Gap

With support from the Nathan Cummings Foundation, Duke’s Samuel DuBois Cook Center on Social Equity is devoted to developing a research program that explores the intersection of gender with race and ethnicity and prospects for economic security and well-being. Our first entry in our new Research Brief Series, published in collaboration with the Insight Center for Community Economic Development (Insight CCED) demonstrates graphically the importance of examining racial differences in wealth outcomes within a shared gender group. We find vast differences in net worth between black and white women regardless of age, family structure, educational attainment, or marital status. View and download the Women, Race and Wealth Research Brief here.

Project Implementation

At a time when the economy has sent a flood of jobless young adults back home, perhaps surprisingly, older people are quietly moving in with their parents at twice the rate of their younger counterparts. According to the UCLA Center for Health Policy Research and the Insight Center for Community Economic Development for seven years through 2012, the number of Californians aged 50 to 64 who live in their parents' homes grew by 67.6% to about 194,000, The jump is almost exclusively the result of financial hardship caused by the Great Recession rather than for other reasons, such as the need to care for aging parents. In collaboration with Insight CCED, Cook Center researchers examine the gender and racial/ethnic dimensions of this pattern of older Americans moving in with their even older parents. Given our findings about the stark racial differences in wealth among women, this phase of the study is important in determining whether particular social groups and whether women, in particular, are facing greater stress associated with returning to living with their own parents at a later stage in life.

Cook Center research scholars also build upon the Center's groundbreaking research that demonstrates the great importance of parents’ transfer of wealth in influencing adult children’s economic outcomes. Researchers focus on mature adults as means of gaining a better understand their future prospects and that of their adult children. Cook Center researchers dually aim to gain a better understanding of how factors like student debt, kinship and employment affect these households ability to accumulate wealth and the profound implications for multiple generations. In addition, Cook Center researchers investigate whether these processes of wealth transfer vary for sons and daughters and whether any gender differences in the receipt of inheritances and gifts from the prior generation are linked to race and ethnicity.

The traditional focus on an individual’s circumstance or the nuclear family potentially overlooks important constraints on black families in particular. Because black women bear a disproportionate role as heads of households, these constraints may impinge more heavily on their lives and well-being. These stressors also may contribute to racial and gender differences in health outcomes, another dimension that we will investigate under the research program on Gender, Race, Ethnicity, and Social Mobility.

Research Questions

Cook Center researchers examine the cumulative effects of gender and race and/or ethnicity on economic outcomes in the U.S. population. This project explores the following broad research questions:

  1. Are the consequences of being black and female additive, non-additive, or exponential in comparison with being white and female or being black and male for wealth or earnings or employment?
  2. What are the magnitudes of the economic penalties associated with the intersection of race and gender in the American economy?
Aims of the Project
  • Provide substantial detail on the relative asset and debt position at the intersection of gender, ethnicity and identity in the United States.
  • Afford, via a combination of strategies, insights about the typical life-cycle or life course pattern of asset and debt accumulation of the target population.
  • Produce a National Asset Scorecard that can be used to evaluate the comparative status of the target population at a point in time as well as over time.
  • Provide rich insights into the sources of the disparities in wealth outcomes for the target population that will be useful in crafting policies to remedy asset inequality and asset poverty.
  • Incorporate a more nuanced treatment of the positives and negatives associated with the accumulation of particular types of indebtedness.
  • Provide data and research support for individuals and organizations committed to advancing a more economically inclusive society.


Research Program on Race, Ethnicity, Gender and Self-Employment

With support from JP Morgan Chase, the Cook Center is committed to the comprehensive development of a research program focused on racial and ethnic inequality and rates of self-employment. Of particular concern is the relationship between engagement in self-employment and racial-cum-gender differences in the accumulation of wealth. In the course of this research program, careful attention will be given to the distinction between the sometimes overlapping categories of self-employment and business ownership and the consequences for household net worth.

Cook Center researchers address the following research questions:
  1. How do patterns of self-employment vary across multiple racial and ethnic groups over the past 35 years? Here, instead of the conventional use of broad Census-based categories like Hispanic, Asian, or black, Cook Center researchers examine self-employment outcomes for specific national origin groups. For example, in place of Hispanics as an aggregate group, researchers look at small business involvement on the part of Cubans, Puerto Ricans, Mexicans, Dominicans, etc. This also facilitates examination of the impact of different proportions of recent immigrants – since immigrants typically have higher rates of self-employment than non-immigrants – on each community’s respective self-employment. Cook Center researchers are especially interested in learning more about differences across immigrant populations in financial resources they possess upon entry into the United States. By utilizing a combination of decennial Census data for the years 1980, 1990, and 2000 when the ancestry question was asked and the semi-annual American Community Survey, thereafter, where the ancestry question is also asked, researchers can investigate small business participation on the part of narrowly defined racial and ethnic groups. The Cook Center will extend the study further back in time by using earlier Censuses to identify precise racial and ethnic groups by exploiting information about the respondent’s country of birth or their parents’ country of birth.
  2. Does wealth lead to self-employment or does self-employment lead to wealth? Wealth and self-employment are correlated. However, the direction of the relationship is unclear. It could be the case that self-employment leads to higher net worth, or that higher net worth facilitates becoming self-employed, or both. By utilizing a longitudinal data set with excellent information about household net worth, the Panel Study of Income Dynamics (PSID), Cook Center Researchers can explore whether self-employment is important to wealth-building or initial access to wealth is important to self-employment – or whether both causal patterns operate interdependently. To explore the relationship between self-employment and wealth, researchers will analyze employment and wealth data from the PSID from 1968 to 2012. Using baseline samples of never previously self-employed heads of households, Cook Center researchers also will examine subsequent self-employment and wealth outcomes, while also asking how this relationship varies by race and ethnicity? Preliminary findings from a study Cook Center researchers conducted using data from 1968 to 1999 led to the following conclusions: Higher levels of net worth do not increase the likelihood of a head of household becoming self-employed in the future. In testing the effect in the reverse direction, researchers found that becoming self-employed correlates with higher wealth outcomes when those individuals are compared against all persons who never became self-employed, including the unemployed. However, when those who became self-employed were compared exclusively with other employed heads of households, their wealth outcomes generally were no better. This suggests that with respect to wealth outcomes, being self-employed is better than no employment, but is not necessarily better than working for hire.
  3. Cook Center researchers will use original data from the National Asset Scorecard for Communities of Color (NASCC) survey to examine the relationships between immigration, race, ethnicity, gender, wealth endowment and entrepreneurial activity described above and below. The NASCC survey is designed to gather asset and debt experiences for ethnic and racial groups defined based on specific ancestral origin, e.g. Chinese, Cuban, Haitian, etc., as opposed to broadly defined categories such as Asian, Latino, and black. The ongoing survey was initiated in 2014 and has been administered in six metropolitan areas – Baltimore, MD (currently in the field), Boston, MA, Miami, FL, Los Angeles, CA, Tulsa, OK, and Washington, DC. These plural cities were strategically chosen to offer variation across and within broadly defined racial and ethnic groups, and nativity status across a range of contexts. The NASCC detailed modules on asset and debt accumulation will compliment both our census and PSID data analysis of self-employment activity. Unlike the NASCC, the census does not collect for specific asset and debt information, such as business equity, household net worth disaggregated by specific types of asset and debt, type of financial products and services used, or remittances and other forms of financial transfers to and from family and friends. And unlike the NASCC, the PSID does not generally permit detailed information on respondent national origin or enough statistical power to examine this inequality in specific geographical contexts across which asset prices and products are known to vary a great deal. Ultimately, the use of the census, PSID and NASCC will permit Cook Center researchers to triangulate our findings across all three surveys in order to boost the robustness of what is found.
  4. How do patterns of small business engagement vary at the intersection of race/ethnicity and gender? Does being female as a member of any national origin group consistently result in lower rates of self-employment in comparison with all men or in comparison with males who share the same racial/ethnic identity? What are the implications of increasing rates of single female headed households on self-employment and are these implications different across race? Is there a need for gender specific policies to promote greater self-employment for all women or for women in particular national origin communities? Are existing federal subsidy programs for female owned small business development sufficient?
  5. What is the historical record of national policies that have inhibited or stifled business development among marginalized ethnic groups? Does reversal of those policies stimulate business development among marginalized groups or do additional steps need to be taken?
  6. What is the relationship between closing the racial wealth gap increased black participation in small business activity? One approach to closing the racial wealth gap is the universal provision of a trust fund for all young people in the United States that they can access upon adulthood. It can have a major positive impact on the racial wealth gap because the amount of the trust fund will vary with the wealth position of the young person’s family, and since black families have considerably lower wealth levels than white families, the program would have a disproportionate benefit for blacks. So it would be universal but not uniform. Cook Center researchers propose to conduct a simulation exercise based upon an estimated structural model to project the impact of a policy of this type on the black-white differential in self-employment rates.
  7. Cook Center researchers also propose to design an experiment that will enable the examination of the relationship between sharply reducing black-white wealth disparities and prospects for significant increases in black participation in self-employment. This phase of the project will involve development of a test for a Universal Adulthood Trust Fund. This experimental approach will complement the simulation exercise described under point 5. The scale of the experiment will be contingent upon the amount of resources provided for conducting the exercise. Cook Center researchers gauge that the minimum number of participants needed to produce a meaningful experiment is 250 young people.


Federal Job Guarantee - National Investment Employment Corps

America’s workers would not be subjected to low-wage jobs if they were assured employment at non-poverty wages. The conventional way to achieve that objective is old-fashioned Keynesian pump priming or through the application of stimulus expenditures.

The federal job guarantee allows any American, 18 years or older, to be able to find work through a federally funded public service employment program – a National Investment Employment Corps (NIEC). While providing a particular benefit – such as raising the minimum wage – for those Americans in the most desperate straits, a universal job guarantee would benefit all Americans who experience joblessness now or in the future.

Moreover, a universal job guarantee would mean that all Americans would have access to jobs promising an income above the poverty threshold.

Each NIEC job would offer individuals nonpoverty wages: a minimum salary of $23,000, plus benefits including federal health insurance. The types of jobs offered could address the maintenance and construction of the nation’s physical and human infrastructure, from building roads, bridges, dams and schools to staffing high-quality day care.The program would include a training component to equip employees with the skills necessary to fill state and municipal needs.

The program would be cost effective, too. If the program put 15 million Americans to work – the total number of people out of work at the nadir of the recent recession – at an approximate cost of $50,000 per employee, the bill for the program would be $750 billion. In 2011, the total cost of the nation’s anti-poverty programs was about $740 billion.

NIEC would work similarly to the Works Progress Administration and the Civilian Conservation Corps developed in response to the unemployment crisis of the Great Depression.

Since the National Investment Employment Corps would function simultaneously as an employment assurance and anti-poverty program, the existing anti-poverty budget could be slashed drastically, with those savings going to finance the job guarantee.