Dr. Vedder Testifies before the United States Commission on Civil Rights



Expanding College Access to Minorities: A Cautionary Note

By Richard Vedder

It is conventional wisdom that greater participation in higher education is necessary for socioeconomic advancement and the achievement of the American Dream. And it is true that, on average, Americas with four year degrees earn dramatically more than those with a high school education, and that the college earnings differential is a good deal larger today than it was at the time that the Civil Rights Act of 1964 passed. In 2013, males with bachelor’s degrees working full-time year round earned on average 78.3 percent more than those with high school diplomas; for females, the college earnings differential was smaller, but a still sizable 62.5 percent.[1] That said, however, my message today is that greater higher education participation is no panacea for eliminating disparities in income and wealth between individuals based on group characteristics such as race and gender. A fervent drive to increase educational attainment among minority groups will likely lead to disappointment, as, in some sense, it already has. Let us look at African-Americans.  Chart 1 shows that the number of African-Americans participating in higher education has nearly tripled since 1970.

In 1970, for every 100 whites enrolled in higher education there were about 11 blacks. By 2013, there were 25– a dramatic growth in educational access by African-Americans. Yet the narrowing of income differentials between blacks and whites has been very modest, as indicated in Table 1. For example, black household income rose by roughly 2 to 5 percentage points relative to whites from 1980 to 2013, depending whether mean or median measures are used, from roughly 60 percent to closer to 65 percent, eliminating perhaps 10 or 12 percent of the income gap between the races. When racial differences in household composition are controlled for, measured progress is greater, but the fact remains increased black educational attainment has succeeded in eradicating only a very small portion of racial income differentials, and future prospects of it doing so in the future are not particularly good.

African-American Income  as a Percent of White Income, 1980 to 2013

Living Unit              Statistical Measured Used                   1980                            2013

Households                             Median                                    57.61%                       62.61%

Households                             Mean                                       63.75                          65.44

Families                                   Median                                    59.02                          63.15

Families                                   Mean                                       63.85                          67.42

Individuals (persons)              Median                                    70.98                          74.13

Individuals                              Mean                                       70.37                          72.98

Source: U.S. Bureau of the Census, author’s calculations.


Why is this so? First of all, the evidence is clear that the proportion of important minority groups like African-Americans and Hispanics entering college that actually graduate within six years is below the already abysmal national average of about 60 percent. For example, looking in 2005 at a cohort of students entering a sampling of flagship state universities, Bowen, Chingos, and McPherson observed that 75 percent of the whites had graduated within six years, compared with only 66 percent of Hispanics and 59 percent of blacks. At less selective non-flagship state schools, the white six year graduation rate was 49 percent, compared with 38 percent for blacks.[2] Schools, under pressure to admit minorities, often accept students with low prospects for success. Special remedial education programs have relatively low success rates. We have many urban universities with high minority participation where far more students dropout than graduate within six years.[3] A contributing factor, no doubt, is the generally inferior quality of inner city public secondary education, leading to students being admitted to college who are at best marginally qualified. Colleges brag about high minority enrollments, but often are guilty of luring students with very low realistic probabilities of success. They gain bragging rights and tuition revenues, but leave many students deep in debt with no degree or high paying job.

Second, merely graduating from college provides no assurance of a good future income. Growing evidence shows that a large portion of recent college graduates are underemployed, performing jobs where a majority of jobholders have high school diplomas. The latest study, published just weeks ago by Accenture Strategy, is of interest. Some 46 percent of 2012 and 2013 graduates with jobs report they are significantly underemployed, holdings job in which college degrees are unnecessary, and over 40 percent are also making less than $25,000 a year –less than the average student loan debt of the sample of over 2,000 recent graduates.[4] These are remarkably similar results to those of other studies using alternative data sources.[5] Arum and Roksa in Aspiring Adults Adrift found that nearly one-fourth of college graduates are living with their parents two years after graduation, and a majority still receive financial support from their parents.[6]  After assessing all the evidence, the authors reach a strong but I believe valid conclusion: “Graduating large numbers of students who have achieved high grades with little effort and achieved limited improvement in competencies such as critical thinking, complex reasoning, and writing is a disservice to the students who enroll in these schools, the families who put trust in these institutions, and the larger society that will be dependent on the productivity and citizenship of thee graduates in the future.”[7] Moreover, as the proportion of adult Americans with bachelor degrees or more approaches one-third, the mere receipt of a degree no longer necessarily indicates a person with above average skills and abilities. Employers are becoming more particular. The high college earnings premium still applies to the graduates of the elite mostly private schools who get good managerial, technical, and professional jobs, but those earnings premiums apply far less so to graduates of schools of lesser reputation, schools where minority representation is historically very high.

Using day from Payscale.com, I looked at the median average earnings of alumni of two groups of 22 schools —the first highly selective, elite private universities, most with fairly modest minority presence (other than Asians), and the second a group of 15 randomly selected rather low reputation schools with generally low admissions selectivity, along with five urban universities and two historically black colleges and universities.[8] The median of the 22 average salaries for early career workers for the elite schools was $58,350 –some 33.5 percent higher than for the comparable workers at the less selective schools. Moreover, the earnings differential grew as careers progressed, to 51.8 percent by mid-career. This suggested that those from the elite schools were more likely to get major promotions to better positions over time. The payoff from attending an elite school was dramatically higher than for the lower reputation institutions. To be sure, tuitions were lower at the non-elite schools on average, but many of the top schools engaged in deep tuition discounting based on need, so unquestionably the rate of return of the college investment was greater in the first group of schools.

Moreover, earnings of college graduates vary considerably with the field of major study.  As of this writing (May 2015), PayScale lists 16 majors with mid-career average salaries of $100,000 or above, all in the STEM disciplines where minorities such as African-Americans and Hispanics are historically seriously underrepresented, and 15 majors with mid-career salaries below $50,000.[9] Some minorities disproportionately major in fields (e.g., education, and social work) whose graduates have relatively low post- graduate earnings.

While some writers lament what they term “under-matching”—persons going to schools below their capability, my bigger worry runs in the opposite direction. Too many students are unaware of the risks associated with college attendance.

The Adverse Effects of the Law of Unintended Consequences

I think the Law of Unintended Consequences has operated as an outgrowth of public policy in ways that have hurt low income persons with minority status. For example, The Griggs v. Duke Power Supreme Court case emanating from the 1964 Civil Rights Act unintentionally increased the value of college diplomas by reducing the ability of firms to use alternative ways of certifying worker competency, thereby allowing colleges to raise fees more aggressively.[10] A potentially even bigger problem emerged as an inadvertent result of the various federal student financial assistance programs emerging out of the Higher Education Act of 1965. As with Griggs v. Duke Power, they expanded the demand for education relative to supply, raising the cost of attendance and disproportionately turning off minorities with below average incomes that were particularly sensitive to the magnitude of tuition fees.[11] The FAFSA form enacted to help disburse financial aid has disproportionately turned off minority group members bewildered by the form’s complexity.

Another unintended consequence of the civil rights legislation mid-twentieth century relates to historically black colleges and universities (HBCUs). These colleges were mostly created long before the civil rights revolution of the 1950s and 1960s. It can be argued that in a society that wishes to assess individuals on “the content of their character,” having schools especially devoted to educating students of one race is an undesirable anachronism. That said, however, many in the African-American community are strong supporters of the HBCUs, which receive special federal financial support. Yet as college enrollment stagnate (total enrollments in the spring of 2015 were lower than they were four years earlier), non-selective admissions schools often are struggling to maintain enrollments and are facing very substantial financial challenges that in some instances threaten their very existence. Schools like Florida A and M, Elizabeth City State and Norfolk State have had very significant enrollment declines and/or financial shortfalls in recent years. Ambitious and qualified African-Americans are wooed by non-HBCU schools aggressively, in some cases offering financial packages far better than those offered by poorly endowed HBCUs.

Other Protected Groups

This short paper has largely concentrated on African-Americans. The growth in enrollments over recent decades has been even more extraordinary for some other groups, especially Hispanics (see Chart 2). The white proportion of college enrollments was 84.3 percent in 1976, but only 59.3 percent in 2013, and probably will fall below 50 percent at some point in the not too distant future. Expanded university participation has not led to an eradication or even a reduction in the white-Hispanic earnings differential; by many measures, the differentials have actually widened since 1980.

The record with respect to women is somewhat different. As Chart 3 indicates, female access to higher education has not only risen, but it now dominates. The biggest gender issue in higher education is explaining the disappearing male from campus. In this case, rising female involvement in higher education has been accompanied by fairly substantial increases in compensation relative to men. To cite just one statistic, in 1991, full-time year round female workers earned on average 63.3 percent of that earned by men; by 2013, women on average earned 72.6 percent the male compensation –more than one-fourth of the differential was eliminated.

The statistics on earnings differentials by race or gender, of course, need to be treated with caution. Large variations in earnings between groups are often the result of factors that have nothing to do with racial or gender discrimination. June and Dave O’Neill have meticulously demonstrated using multivariate statistical techniques that almost all gender and racial earnings differentials are explained by various determinants of productivity, such as years of experience, the type of job involved, etc.


To be sure, as a general proposition education positively impacts on earnings –more is better. But public policy is concerned with life at the margin: does incremental amounts of education improve the economic standing of a group of people? Remember, always, that the mere participation in postsecondary education is far different than successfully completing a degree program, and that the risks associated with that education are particularly high for those with mediocre secondary educational preparation, relatively little knowledgeable parental encouragement, etc. –characteristics that are on average disproportionately present among some minority groups including African-Americans and Hispanics. Others, such as Stephen and Abigail Thernstrom and Richard Sander and Stuart Taylor, have written about some of the negative consequences that affirmative action recruitment of minorities can have on student outcomes.[14] I worry that we are on balance burdening African-Americans and Hispanics by overselling the gains and understating the risks associated with going to college.  The “we” here includes the colleges themselves. They gain revenue from admitting minority students with low prospects for success, not to mention secure public relations advantages associated with appearing to be inclusive and promoting access for all. But when students dropout, who faces the financial consequences? The students themselves, who have often incurred debt and have little to show for it. The taxpayers who subsidize college attendance. But not the colleges themselves. There has been severe criticism of for-profit institutions in this regard, but the problem is by no means confined to them. Institutions of higher education should have “skin in the game,” sharing in the adverse financial consequences associated with college dropouts falling delinquent on large amounts of college debt.

Noble intentions were behind the civil rights legislation of the 1960s. We have seen enormous changes in attitudes towards individuals based on race and gender. African-Americans who fought to vote in the 1950s have seen a member of their racial group attain the presidency of the United States but a half century later, not to mention top jobs in entertainment, business and politics. Women have likewise achieved levels of acceptance in the workplace and society unimaginable in, say, 1960. Indeed the progress is so great, and the evidence of overt discrimination so small relative to the past, that a case can be made that we no longer need to maintain an elaborate apparatus to enforce laws designed to promote equal opportunity.[15] In the area of higher education, the promotion of “college for all” has contributed to overinvestment and underperformance –high underemployment of workers and high student debt loads. This has disproportionately hurt members of minority groups. Proposed ideas such as free tuition for community college students and ever more generous loan forgiveness policies merely worsen the problem, doing nothing to reduce, not increase, the demand for higher education. Where are proposals for, say, vouchers for low income youths to attend postsecondary vocationally oriented programs to learn how to be, say, a welder, or drive an eighteen wheel truck?   In short, we need to rethink our higher education policies generally, and with respect to disadvantaged groups in particular.


[1] Data are from the Census Bureau, accessible at www.census.gov/hhes/www/income/data/historical/people . There are alternative ways the data are classified, such as for all workers or all individuals, and also median as well as mean data are presented. Similarly, data are available for households and families. In all cases, college degree holders earn a good deal more than those with high school diplomas.

[2] See William G. Bowen, Matthew M. Chingos, and Michael S. McPherson, Crossing the Finishing Line: Completing College at America’s Public Universities (Princeton, NJ: Princeton University Press, 2009), pp.46-50 for race and gender completion data.

[3] Five examples: the University of District Columbia, 16 percent; Wayne State University, 32 percent; Chicago State University, 21 percent; University of Texas at San Antonio, 30 percent; and California State University at Los Angeles, 36 percent.

[4] David Smith, Katherine LaVelle and Anthony Abbatiello, Great Expectations: Insights from the Accenture 2014 College Graduate Survey, accessible at www.accenture.com/us=en/pages/insight-2014-accenture-college-graduate-employment-survey.aspx 

[5] See, for example, Richard Vedder, Christopher Denhart, and Jonathan Robe, Why Are Recent College Graduates Underemployed? University Enrollment and Labor-Market Realities (Washington, D.C.: Center for College Affordability and Productivity, January 2013).

[6] Richard Arum and Josipa Roksa, Aspiring Adults Adrift: Tentative Transitions of College Graduates (Chicago, IL: University of Chicago Press, 2014).

[7] Arum and Roksa, p. 135.

[8]  The data were accessed at www.payscale.com/college-salary-report/bachelors , on May 15 and 16, 2015.The highly selective elite private schools included the entire Ivy League (Harvard, Yale, Princeton, Columbia, University of Pennsylvania, Dartmouth, Brown and Cornell), along with M.I.T., Stanford, University of Chicago, Northwestern, Duke, Cal Tech, Notre Dame, Emory, Washington U. in St. Louis, Vanderbilt, Georgetown, Johns Hopkins, Rice, and Carnegie-Mellon. The less selective, urban and HBCU schools included: Wayne State University, Chicago State University, University of Texas at San Antonio, California State University at Los Angeles, City University of New York: Hunter College, Florida A & M, Howard, West Virginia University, Immaculata University, California State University at San Bernardino, Portland State University, Oral Roberts University, Millersville University, University of New England, Northern Illinois University, Kennesaw State University, Millikin University, Carthage College, Embry-Riddle Aeronautical University-Daytona Beach, Nova Southeastern University, Wright State University and Fort Lewis College.

[9] www.payscale.com/college-salary-report/majors-that-pay-you-back/bachelors, accessed on May 16, 2015.

[10] See Bryan O’Keefe and Richard Vedder, Griggs v. Duke Power: Implications for College Credentialing (Pope Center for Higher Education Policy and the Center for College Affordability and Productivity, 2008). For commentary on the O’Keefe and Vedder study, see George F. Will, “Inadvertent Consequences of Rights Expanded,” Washington Post, January 4, 2009, accessed at www.washingtonpost.com/wp-dyn/content/article/2009/01/02/AR20090102098.html .

[11] This assertion is often called the Bennett Hypothesis, after former Secretary of Education William Bennett. See his “Our Greedy Colleges,” New York Times, February 18, 1987, accessed at www.nytimes.com/1987.02/18/opinion/our-greedy-colleges.html. For a supportive analysis of the Bennett hypothesis, see Andrew Gillen, Introducing Bennett Hypothesis 2.0 (Washington, D.C.: Center for College Affordability and Productivity, February 2012). There are numerous studies that analyze the validity of this hypothesis, often at least partly supporting it. See for example, Dennis Epple, Richard Romano, Sinan Sarpca, and Holger Sieg, “The U.S. Market for Higher Education: A General Equilibrium Analysis of State and Private Colleges and Public Funding Policies,” National Bureau of Economic Research Paper No.19298, August 2013. Bennett, along with David Wilezol, returned to the fray with their Is College Worth It? (New York: Thomas Nelson, 2013), reaffirming the Bennett Hypothesis.

[12] For example, the median income of Hispanic American families in 1980 was 67.6 percent of whites; it was 66.0 percent in 2013. See footnote one for links to the relevant data. There has been a large immigration of Hispanics to the U.S., no doubt influencing the statistical earnings estimates for that group of the population.

[13]June O’Neill and Dave O’Neill, The Declining Importance of Race and Gender in the Labor Market (Washington, D.C.: AEI Press, 2013).

[14] Stephen and Abigail Thernstrom: America in Black and White: One Nation, Indivisible (New York: Simon and Schuster, 1997), and Richard Sander and Stuart Taylor, Jr., Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It (New York: Basic Books, 2012).

[15] As the O’Neill’s and others have observed, black Americans made significant gains in earnings in the era 1940 to 1960, before the passage of Civil Rights legislation. Economic theory suggests those who discriminate on the basis of group characteristics pay a price –a loss of income. There are powerful market forces that work over time to reduce the discriminatory impulse. It is therefore not clear to what extent the advances of minorities in America since the 1950s are the result of market forces as opposed to legislative and regulatory efforts.