 |
Chairman Reynolds, members of the commission, my name is Douglas Besharov, and I am a resident scholar at the American Enterprise Institute for Public Policy Research and a professor at the University of Maryland School of Public Policy. Thank you for inviting me to testify today on this important topic. (My presentation will follow the PowerPoint presentation that I understand has been distributed to you.)
Essentially, my message today is that, since the early 1980s, the black middle class has hardly grown. You may be surprised when I say that, since just about everyone assumes that it is growing. That is because it is growing in absolute numbers (as it is in the general population), but not as a percentage of the black population.
In my testimony, I will trace the reasons why I have come to this conclusion and then offer a partial list of remedies. I will focus on what I consider the most achievable remedy: increasing the college graduation rate of African Americans by focusing on academic achievement (at all levels of schooling), the mismatch between achievement and school attended, the need for a supportive school culture, and the reform of financial aid formulas.
Slide 1
My topic today is "The Economic Stagnation of the Black Middle Class (Relative to Whites)." That last parenthetical is important, for, as the first slide shows, black economic progress continues. Between 1980 and 2003, the percentage of black households with incomes greater than the 1980 median income of all American households grew by a third, going from about 29 percent to about 40 percent. But that is a comparison with past white incomes, and white incomes have also continued to rise-- while African Americans have not been able to close the gap.
Slide 2
I believe that the best way to see this lack of black progress relative to whites is to examine the percent of African Americans in different income quintiles. I should mention, however, that my colleague Bart Landry, who will address you in a moment, will describe how sociology explores this same question.
The second slide portrays, for the year 2003, the income distribution of all American households, and then indicates both the median income (about $43,000) and the quintile divisions ($0 – about $18,000, $18,000 – $34,000, $34,000 – $54,000, $54,000 – $87,000, and $87,000 and above).
Slide 3
The next slide shows a number of things about income trends. First, we have added the bottom decile, that is, the bottom 10 percent of incomes. Thus, at the very bottom, African Americans are twice as likely to be in that group than are whites, and the proportion has hardly changed over the past 45 years.
Second, at the bottom quintile, there has been progress, with a quarter fewer blacks in that group-- although they are still 50 percent more likely to be in that income group than are whites.
Third, and also good news, the proportion of African Americans in the top quintile is about 66 percent higher.
Slides 4-6
The next three slides show the distribution of households in the middle and top quintiles.
The first one shows the distribution in the middle three quintiles. Looking at it, the distributions seem equal, but that is only because there are so many African Americans in the second quintile, as opposed to the top three.
For, as slides 5 and 6 show, in the top two and the top three quintiles, the lower incomes of African Americans are quite apparent. Just as importantly, these slides show that-- after substantial progress between 1959 and 1979-- black incomes relative to white incomes have essentially held constant.
In other words, if one considers the middle class to be, essentially, either the top two or the top three quintiles, the percentage of blacks in the middle class has been stuck at about the same percent since the early 1980s.
Slide 7
I have added slide 7 on self-identification as middle class to show the impact of the absolute and relative income measures on African American attitudes. On the one hand, the continued absolute progress in black incomes is demonstrated by the African American responses through the 1990s, but the lack of relative progress is mirrored in the higher white self-identification as middle class.
Slide 8
Let me turn now to the most likely culprits.
Slide 8 is simply a reminder that discrimination, at some level, is surely a factor, although the other factors that I will discuss are just as surely more important.
Slides 9 – 10
Family structure helps explain some of the differences between whites and blacks, although there is some argument about the direction of causation. Slide 9 shows how much higher the incomes of married-couple families are compared to female-headed families-- of either race. Slide 10 shows that blacks are simply less likely to be in married-couple families.
Study after study has shown that black poverty would be much lower if family structures had not weakened beginning in the 1960s. Examining the effect of family structure on child poverty, Isabel Sawhill and Adam Thomas of the Brookings Institution find that "had the proportion of children living in female-headed families remained constant since 1970, the child poverty rate in 1998 would have fallen by 1 percentage point, rather than rising by 3.4 percentage points, relative to the 1970 rate. Thus, the poverty rate is 4.4 percentage points – or 24 percent – lower as a result of assuming marriage patterns similar to those that existed in 1970."[1] Sawhill and Thomas continue, research by Eggebeen and Lichter suggests that, "had there been no changes in family structure between 1960 and 1998, the black child poverty rate in 1998 would have been 28.4 percent rather than 45.6 percent."[2]
Slides 11 – 13
Slides 11 – 13 portray trends in government employment. They are only a minor part of what has happened in the last forty years, but they demonstrate the complexity and changing nature of the factors influencing the incomes of African Americans.
Slide 11 shows two things. First, government employment was an important element of black economic progress from 1959 to 1979. At a time of much greater racial discrimination, government was an important source of jobs for African Americans. Although government employment has declined, it continues to be higher than that of whites. In fact, after 1979, the higher proportion of blacks working in government may help to explain their income stagnation, as government salaries have not kept up with those in other sectors.
Slide 12 shows that the same story was even more pronounced for the black middle class-- with a somewhat higher rise and somewhat larger decline.
Slide 13 quite dramatically shows the declining importance of government employment for all college graduates-- blacks, whites, and Hispanics.
Slide 14
This brings me to the crucially important role of education. As Harry Holzer will describe in a moment, economists agree that the U.S. labor market now puts a much greater premium on education and on skills generally than in the past, and that, since the early 1980s, the differences in earnings for workers with different levels of education have grown greatly.
Let me be clear about where I am going: All things being equal, since the early 1980s, differences in educational attainment and in work skills generally have resulted in often substantial differences in earnings. Hence, the main proximate cause of the lack of economic progress among African Americans is the continuing difference in educational attainment between whites and blacks.
In the context of today’s hearing, that is to say, the black middle class is not growing (compared to whites) because of high drop-out rates in middle and high school, low college attendance rates, and even lower college graduation rates (again compared to whites).
Slide 14 portrays the thirty-year rise in high school "completion" rates, especially for African Americans (now around 90 percent). (Note that we try to capture current patterns by restricting the analysis to 25- to 29-year-olds.) This slide is somewhat misleading, however, because, since 1992, completion includes those who have obtained a General Education Degree (or GED). Hence, the figures for 1992 and before are probably a better indicator of high school completion, and they suggest much less progress. In addition, sadly, many think that even finishing high school does not provide the education and skills that it once did, especially in many inner-city schools. (The even lower completion rates for Hispanics are partly a result of large numbers of immigrant children, but they have a similar dampening impact on earnings.)
Slides 15 – 16
Slides 15 and 16 have to be read togther. Slide 15 portrays an impressive increase in college attendance rates, for whites and blacks, at least. For instance, the college attendance rate for blacks males increased from about 16 percent in 1971 to about 50 percent in 2003, while the attendance rate for white males went from about 42 percent to about 61 percent.
However, as slide 16 shows, the graduation rate for African Americans is not nearly as rosy. Although the rate for African American males increased from about 7 percent to about 18 percent, that is still only 18 percent compared to the white male graduation rate of 31 percent. (Again, we try to capture current patterns by restricting the analysis to 25- to 29-year-olds.)
Moreover, the college drop-out rate for blacks has actually risen since the 1980s. Let me say that again: Even though the graduation rate has risen for African Americans, since the 1980s, so has the college drop-out rate. Since 1971, the white college drop-out rate (as a percentage of those who attend) has decreased from about 50 percent to about 47 percent in 2003. The Hispanic drop-out rate has hovered at about 66 percent. But the black drop-out rate has risen some, going from about 63 percent in 1971 to about 64 percent in 1980, and then to about 67 percent in 2003.
Slide 21
Everyone, I think, appreciates the tragically low academic achievement of many young people of all races and ethnicities, but with the greatest concentrations among disadvantaged minorities. Why this disappointing level of educational attainment among blacks and Hispanics?
The usual response from the left is to blame discrimination, poverty, and other social ills. The right tends to blame family breakdown and dysfunctional behavior in general. Both sides blame the low quality of the schools many disadvantaged minorities attend. Each of these explanations is to some extent valid, and they all deserve our attention.
But I would like to highlight three other factors that, because they are institutional (or programmatic), could be addressed more easily by government or other social institutions than could many other causes.
- Poor high school counseling, especially for disadvantaged minorities. I believe the evidence that too many young people go to the wrong college and that the resulting "mismatch" helps explain poor performance in college and high drop-out rates in general.[4] If this research is correct, African American graduation rates would rise if there were a better match between the incoming students’ academic achievement and the school’s level of academic rigor. Students from disadvantaged families, especially, need better information and advice about where (and how) to apply to college and, if accepted, which schools to attend and how to navigate the student aid process. And, yet, it is just such students who get the least counseling (or inadequate counseling).
- Unsupportive college cultures that often seem oblivious to the special challenges faced by young people coming from severely disadvantaged backgrounds. There are clearly right ways and wrong ways to provide support, and this is too important an area to be as unexamined as it is.
- Financial aid formulas tilted in favor of the middle class. Many might wish that college was free to all students, or at least to all but the most affluent. But that is not likely in our lifetimes. In the meantime, the limited aid available is inappropriately tilted toward the middle class as a result of the need analysis formula by which aid is determined. Let me explain.
The aid formula disregards all family assets when parental income is less than $49,999 and, regardless of family income, ignores the home equity (however great) in the family’s principal residence.[5] (According to Susan Dynarski of Harvard University, those with the highest home values saw the greatest boost to their aid eligibility when this disregard was implemented.[6]
Slide 22
As slide 22 dramatically shows, disregarding assets and home equity obscures important wealth differences between whites and blacks. This might not be a problem if there were enough funds and more to go around, but there are not. Hence, the effect of these rules is to decrease the amount of aid available for the truly needy. In fact, grants to low-income students (such as Pell Grants) have failed to keep pace with college costs, and now cover less than half of the cost of tuition, fees, and room and board at the average public four-year college. This forces the lowest income students to supplement their grants with substantial loans, illustrated by the fact that between 1981–1982 and 2001–2002, the share of all student aid (federal, state, and other) coming in the form of grants fell from 52 percent to 39 percent, while the share in the form of loans rose from 45 percent to 54 percent. In addition, unmet need often proves an insurmountable barrier for low-income students. Hence, the current approach to financial aid discourages college attendance and increases the drop-out rates of low-income (often minority) students.
Thank you for giving me this opportunity to describe my research.
Douglas J. Besharov is the Joseph J. and Violet Jacobs scholar in social welfare studies at AEI.
Notes
[1] Isabel Sawhill and Adam Thomas, "For Richer or for Poorer: Marriage as an Anitpoverty Strategy," Journal of Policy Analysis and Management 21, no. 4, October 8, 2002, pp. 587-599.
[2] Isabel Sawhill and Adam Thomas, "For Richer or for Poorer: Marriage as an Anitpoverty Strategy," Journal of Policy Analysis and Management 21, no. 4, October 8, 2002, pp. 587-599.
[3] Prior to 1992, the Census Bureau asked respondents to report the highest grade they had attended and whether or not they completed that grade. It was assumed that a person who had completed four years of college had a bachelor’s degree; however, students can complete four years of college without receiving a degree and drop-out before they do receive one. Since 1992, the questionnaire has asked for the highest grade of school completed or the highest degree obtained. Thus, the drop-out rates prior to 1992 could possibly be understated (if there were many drop-outs after four years) or, less likely, overstated (if there were many students who obtained a degree in less than four years).
[4] See Richard H. Sander, "A Systematic Analysis of Affirmative Action in American Law Schools," Stanford Law Review 57 (2004): 367.
[5] See U.S. Department of Education, Office of Federal Student Aid, "The Expected Family Contribution Formula, 2005–2006," available from: http://studentaid.ed.gov/students/attachments/siteresources/EFCFormulaGuide0506.pdf, accessed July 12, 2005.
[6] Susan Dynarski, "Loans, Liquidity, and Schooling Decisions," Harvard University, Kennedy School of Government and NBER, Draft, February 2002, p. 2, available from: http://ksghome.harvard.edu/~.SDynarski.Academic.Ksg/Dynarski_loans.pdf, accessed June 23, 2003.
