Advertisement

The Review of Black Political Economy

, Volume 24, Issue 4, pp 13–46 | Cite as

What does the AFQT really measure: Race, wages, schooling and the AFQT score

  • William M. Rodgers
  • William E. Spriggs
Articles

Keywords

School Quality Labor Market Discrimination Family Background Characteristic Racial Wage Basic Human Capital 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Notes

  1. 1.
    For recent work that describes wage differentials between black and white males, see J. Bound and R. Freeman, “What Went Wrong? The Erosion of Relative Earnings and Employment Among Young Black Men in the 1980s,” TheQuarterly Journal of Economics 107 (February 1992): 201–232; W.M. Rodgers HI, “Male Black-White Wage Gaps, 1979-1991: A Distributional Analysis” (unpublished manuscript, 1993); Samuel Myers, Jr., “‘The Rich Get Richer And.’ The Problem of Race and Inequality in the 1990s,”Journal of Law and Inequality 11 (June 1993): 369–389; and C. Juhn, K. Murphy, and B. Pierce, “Accounting for the Slowdown in Black-White Wage Convergence,” in M. Kosters, ed.,Workers and Their Wages: Changing Patterns in the United States (Washington, D.C.: AEI Press, 1991). For work that utilizes decennial censuses, see J. Smith and F. Welch,Closing the Gap: Forty Years of Economic Progress for Blacks (Santa Monica, CA: The Rand Corporation, R-3330-DOL, June 1986); J. Smith and F. Welch, “Black Economic Progress After Myrdal,”The Journal of Economic Literature 27 (June 1989): 519-564; and G. Jaynes, “The Labor Market Status of Black Americans: 1939-1985,”The Journal of Economic Perspectives 4 (Fall 1990): 9-24 For recent work on females, see L. Leete and E. McCrate, “Black-White Wage Differences, Among Young Women, 1977–1986,”Industrial Relations 33 (April 1994): 168–183, F. Blau and A. Bellar, “Black-White Earnings Over the 1970s and 1980s: Gender Differences in Trends,”The Review of Economics and Statistics 74 (May 1993): 276-286, and A. K. Fosu, ’Trends in Relative Earnings Gains by Black Women: Implications for the Future,”Review of Black Political Economy 17 (Summer 1988): 31-45.CrossRefGoogle Scholar
  2. 2.
    This has not gone unnoticed by economists. Earlier attempts to address these missing variables can be found in P. J. Taubman,Sources of Inequality in Earnings. Personal Skills, Random Events, Preferences Towards Risk, and Other Occupational Characteristics (New York: North-Holland Publishing, 1975), and P. J. Taubman, ed.,Kinometrics: Determinants of Socioeconomic Success Within and Between Families, (New York: North-Holland Publishing, 1977).Google Scholar
  3. 3.
    O’Neill, J. 1990, “The Role of Human Capital in Earnings Differences Between Black and White Men,”The Journal of Economic Perspectives 4 (Fall): 25–45.Google Scholar
  4. 4.
    Ferguson, R. 1994, “New Evidence on the Growing Value of Skill and Consequences for Racial Disparity and Returns to Schooling,” unpublished paper, John F. Kennedy School of Government, Harvard University (September). Ferguson, R. 1995, “Shifting Challenges: Fifty Years of Economic Change toward Black-White Earnings Equality,”Daedulus 124 (Winter): 37–76.Google Scholar
  5. 5.
    Maxwell, N. 1994, “The Effect on Black-White Wage Differences of Differences in the Quantity and Quality of Education,”Industrial and Labor Relations Review, 47 (January): 249–264.CrossRefGoogle Scholar
  6. 6.
    Neal, D. and W. Johnson, 1994. “The Role of Pre-Market Factors in Black-White Differences,”The Journal of Political Economy 104 (1996): 869–895.Google Scholar
  7. 7.
    The NLSY contains two scores. The original score is used by J. O’Neill, “The Role of Human Capital in Earnings Differences Between Black and White Men,”The Journal of Economic Perspectives 4 (Fall 1990): 25–45; N. Maxwell, “The Effect on Black-White Wage Differences of Differences in the Quantity and Quality of Education,”Industrial and Labor Relations Review 47 (January 1994): 249-264; and R. Ferguson, “New Evidence on the Growing Value of Skill and Consequences for Racial Disparity and Returns to Schooling” (unpublished manuscript, September 1993); it is called the AFQT80 score. It is the sum of the raw scores from the word knowledge, paragraph comprehension, arithmetic reasoning, and the numerical operations components. The numerical operations score is divided by one-half. In 1989, a new measure of skill called the AFQT89 score was created. It is the sum of the standard scores from the arithmetic reasoning, math knowledge, and verbal composite components. The verbal composite score is multiplied by 2. D. Neal and W. Johnson, “The Role of Pre-Market Factors in Black-White Differences” (see note 6) use this score. The Department of Defense felt that this construction better measures an individual’s skills.Google Scholar
  8. 8.
    David Card and Alan Krueger, “School Quality and Black-White Relative Earnings: A Direct Assessment,”The Quarterly Journal of Economics; 107 (February 1992): 151–200, show that school quality in 18 southern segregated states was quite similar even by the mid-1960s. The pupil-teacher ratio was 90 percent in 1966, and the black-white differences in term length and teachers pay were gone by the mid-1950s. J. Grogger, “Does School Quality Explain the Recent Black-White Wage Trend?”Journal of Labor Economics (forthcoming), shows that in the High School Beyond Survey of the Class of 1980, pupil-teacher ratios show a slight advantage to whites, term length was the same for each race, and 30 percent or more of teachers in each environment had advanced degrees. African Americans attended slightly larger schools. In terms of expenditures and salaries, the black-white ratio of direct expenditures per pupil was 96 percent, school expenditures per pupil was 95 percent, base teacher salary was 99 percent, and there was racial parity in computer instruction. Also see J. Smith, and F. Welch, 1986Closing the Gap: Forty Years of Economic Progress for Blacks, The Rand Corporation, Santa Monica, R-3330-DOL (February), and J. Smith, and F. Welch 1989 “Black Economic Progress After Myrdal,”The Journal of Economic Literature 27 (June): 519-564.CrossRefGoogle Scholar
  9. 10.
    Although researchers such as R. Hermstein and C. Murray,The Bell Curve (Free Press 1994), interpret the AFQT as a measure of “innate ability,” study guides for the ASVAB exam are available from independent sources that purport to be able to raise one’s score. If, indeed, the test can be coached, then it is not unreasonable to assume that the test partially reflects previous school quality.Google Scholar
  10. 11.
    Bock and Moore, 1986.Google Scholar
  11. 12.
    Barron’s Educational Series, 1988.Google Scholar
  12. 14.
    Wigdor and Green, 1991, Volume 1, p.149.Google Scholar
  13. 20.
    Wigdor and Green, 1991.Google Scholar
  14. 21.
    The earlier work of G. Chamberlain, “Education, Income, and Ability Revisited,” in D. J. Aigner and A. S. Goldberger, eds.,Latent Variables in Socio-Economic Models (New York: North-Holland Publishing, 1977), and Z. Griliches and W. Mason, “Education, Income, and Ability,”Journal of Political Economy 80 (1972): S74-S103, excluded the AFQT composite score from the log wage equation, and instead used a latent variable model approach. The result is the use of an instrumental variable based on the AFQT composite score. That is different from our approach used here. Our approach allows for a direct comparison to the work of O’Neill, Neal and Johnson and Ferguson. However, Chamberlain and Griliches and Mason argued that a basis for the exclusion of the AFQT composite score from the log wage equation includes that the AFQT weights knowledge to predict performance in a different arena than wages. In that sense, the approach taken by Chamberlain is similar to the argument made here.Google Scholar
  15. 22.
    The conclusion on p. 179 was as follows: To the extent that we have confidence in the hands-on criterion as a good measure of performance on the job, these findings strongly suggest that scores on the ASVAB exaggerate the size of the differences that will ultimately be found in the job performance of the two groups.Google Scholar
  16. 23.
    The fact that the AFQT score differentially predicts success in military occupations by race is not unique to the AFQT. For example, the U.S. Employment Service of the Department of Labor developed the General Aptitude Test Battery (GATB) in the late 1940s. This test was designed to help employment services in job referrals. J. Hartigan and A. K. Wigdor,Fairness in Employment Testing: Validity Generalization, Minority Issues, and the General Aptitude Test Battery (Washington, D.C.: National Academy Press, 1989, p. 697) concluded that: … there were differences in both the validities and the prediction equations for blacks and nonminorities. First, the average correlations between test score and supervisor ratings were.12 for blacks and .19 for nonminorities. Second, the formula that best predicts black performance is somewhat different from that predicting the performance of majority-group applicants. … the disproportionate impact of selection error provides scientific grounds for the adjustment of minority scores so that able minority workers have approximately the same chances of referral as able majority workers. 24. The critical value for a 5 percent level of significance where there are 3 restrictions and 9 parameter estimates in the unrestricted model equals 2.60.Google Scholar
  17. 25.
    J. Currie and D. Thomas, 1995, “Race, Children’s Cognitive Achievement andThe Bell Curve, “ NBER Working Paper #5240 (August).Google Scholar
  18. 26.
    The critical value for a 5 percent level of significance equals 2.09.Google Scholar
  19. 27.
    Our results differ from Maxwell (1994). She focuses exclusively on white and black males and uses the number of siblings, parent’s education, a human capital background measure, whether the respondent lived in a female headed household, and whether the respondent’s parents were professionals to predict AFQT scores. She obtains R2’s of .21 and .22 for white and black men. Maxwell’s sample restrictions drive her results. She limits her sample to black and white men who have wages six years after leaving school. What is the implication of this sample restriction? We constructed a sample quite similar to hers, and estimated her specification. We obtained R2’s of .28 and .23. However, when we estimated her specification with all black and white men, we obtained R2’s of .32 and .20. The ability to explain black test scores worsens, while the ability to explain white test scores improves. These differences emerge because her sample restrictions exclude white men whose family background characteristics explain the variation in white scores quite well, and exclude black men whose family background characteristics explain the variation in black scores quite poorly. The R2’s for the men that Maxwell excludes are .38 and .16 for white and black men. Our results are available upon request. 28.We exclude five black men and one black women from the sample because their instrumental variable test score is two standard deviations below the mean score. We examined their data and found that they had implausible values for the logarithm of teacher salary or a combination of extremely low values for the logarithm of the following variables: the number of full-time equivalent instructors, the number of books in the school’s library, and the number of full-time counselors. We also estimate models that exclude black men and women whose instrumented test score is 3 standard deviations below the mean. This restriction only excludes one black male respondent. Doing this has no impact on our conclusions. 29.The obvious criticism of our AFQT specification is that we have omitted family background and school quality variables that are correlated with race, through nonlinearities in education and interactions between education and family background. To respond to the first criticism, we estimate specifications that enter parent’s education as dummy variables. To respond to the second criticism, we estimate specifications that interact the parental education dummy variables with the other family background variables. The estimates are not different from our original specification.Google Scholar
  20. 30.
    D. Neal and W. Johnson, 1994, “The Role of Pre-Market Factors in Black-White Differences,” (see note 6).Google Scholar
  21. 31.
    D. W. Grissmer, K. Nataraj, M. Berends, and S. Williamson, 1994,Student Achievement and the Changing American Family MR-4888-LE, Santa Monica, CA: RAND.Google Scholar
  22. 32.
    R. Haveman and B. Wolfe, 1994, “Intergenerational Determinants of the Education Level of Young Adults: Who Finishes High School and Who Goes Beyond?,” unpublished paper, University of Wisconsin-Madison.Google Scholar
  23. 33.
    M. Boozer, A. Krueger, and S. Wolkon, 1992 “Race and School Quality SinceBrown v. Board of Education,”Brookings Papers on Economic Activity: Microeconomics, pp. 269–338.Google Scholar
  24. 34.
    A.S. Cancio, T. D. Evans, and D. J. Maume Jr., “Cognitive Skills Test Scores and Racial Wage Inequality: A Reply to Farkas and Vicknair,”American Sociological Review 61 (August 1996): 561–564. A. S. Cancio, T. D. Evans, and D. J. Maume Jr., “The Declining Significance of Race Reconsidered: Racial Differences in Early-Career Wages,”American Sociological Review 61 (August 1996): 541–556. G. Farkas and K. Vicknair, “Appropriate Tests of Racial Wage Discrimination Require Controls for Cognitive Skill: Comment on the Paper by Cancio, Evans, and Maume,”American Sociological Review 61 (August 1996): 557–560.Google Scholar

Copyright information

© Springer 1997

Authors and Affiliations

  • William M. Rodgers
  • William E. Spriggs

There are no affiliations available

Personalised recommendations