PDF PDF    printable version of page Printer-friendly page

Focus On Basics

Volume 7, Issue A ::: June 2004

Skills Matter in the Types of Jobs Young Dropouts Will First Hold

by John Tyler
Do basic cognitive skills matter for the least educated? They may not, if computerization of the workplace has "deskilled" the types of jobs in which young dropouts are first employed. This is, therefore, an empirical question that has received little attention and less systematic research. The ideal way to examine this question would be to begin with a pool of school dropouts, randomly assign individuals in this pool different levels of cognitive skills, and then follow them into the labor market to see if those with higher levels of cognitive skills were employed more and earned more than those with lower skill levels. To approximate this situation in a recent research study, I used data on General Educational Development (GED) candidates who attempted the GED exams in Florida between 1995 and 1998, when all were 16 to 18 years of age.

I used the scores of these individuals on the math portion of the GED battery as a measure of their basic cognitive skills. To score well on the GED math exam, you have to know basic math, you have to be able to read the problems, and you have to be able to follow basic instructions. These data are well suited for determining the economic importance of cognitive skills for dropouts for several reasons. First, the GED exams are a high-stakes test for these dropouts and so we think that individuals bring their best effort to the exams. As a result, test scores on the GED exams are likely a better measure of underlying cognitive skills than test scores on standardized tests with no stakes attached. Second, these data contain very recent labor market information on a large sample of dropouts. Following the dropouts for three years after they last tested on the GED, I observed earnings as recently as 1998 and 2001. Third, these data allow me to control at least partially for confounding factors that could limit what we can learn about the returns to skills. For example, if, in a random sample of dropouts, we see a dropout who has a higher test score earning more than a lower-scoring dropout, we do not know how much of the observed earnings premium is a result of higher cognitive skills and how much is a result of unobserved (to the researcher) factors correlated with both higher test scores and greater earnings.

Confounding Factors
Unobserved motivation is an example of such a confounding factor. If we assume that motivation is rewarded in the labor market and that more highly motivated individuals tend to score higher on a standardized test, then failure to control for motivation will lead to overestimates of the causal effect of cognitive skills on earnings. In the data I used for this study, all dropouts indicated a desire to obtain a GED and all had sat through the seven-plus hours of testing. It is thus likely that selection into the data set itself controls for some level of motivation. I used other variables in the data to control for other potential confounding factors as well. Finally, it is likely that earnings information taken from state administrative records, as were used in this study, are a more accurate measure of earnings than self-reported earnings. In summary, while not as good as true experimental data, the data on GED candidates offer several distinct advantages over typical survey data in answering this research question.

Using these data I found that young dropouts do experience a nontrivial economic return on basic cognitive skills in their first jobs in the labor market. Based on earnings in the first three years after taking the GED exams, dropouts who score a standard deviation higher on the GED math exams can expect earnings 6.5 percent higher than those with lower scores. (A standard deviation is a measure of how much spreadb variationb there is in the data. We normally think of education interventions that can move test scores by a quarter of a standard deviation as fairly big effects.) This is the best evidence yet that basic cognitive skills, at least as represented by scores on a math exam, do matter in the types of entry level jobs that young dropouts first hold.

The implication of this finding is that public policies supporting skill-enhancing programs could have a positive impact on the economic outcomes of low-skilled individuals. One way to increase the cognitive skills of dropouts would be to keep them in school longer. However, no dropout prevention programs have, under rigorous evaluation, been proven to be able to do this consistently. The alternative is to focus on programs that could directly affect the cognitive skills of dropouts. The only program that has undergone a rigorous evaluation in this context is the federal Job Corps program. A randomized evaluation of Job Corps found that it increased the math skills of participants by a tenth of a standard deviation (Schochet et al., 2000). Since skill enhancement is only one component of Job Corps, and since the general pool of dropouts is less disadvantaged than the Job Corps-eligible pool, it is reasonable to expect that a program focused on skill-enhancement could increase basic cognitive skills of the random dropout by something more than a tenth of standard deviation.

What if we could find or develop programs that could, on average, increase the basic cognitive skills of dropouts by as much as a quarter of a standard deviation? Based on a set of reasonable assumptions concerning interest rates, inflation rates, and productivity growth in the economy, the returns to skills I measured using Florida GED candidates mean that increasing the cognitive skills of a dropout by a quarter of standard deviation would result in an increased earnings stream over a lifetime worth between $20,000 and $40,000 if paid out in a lump sum today. This calculation does not factor in the personal and societal benefits such as better parenting skills, better health, and increased civic participation that would likely result from increased cognitive skills. Ignoring these other potentially large benefits, a program that could increase the basic cognitive skills of dropouts by a quarter of a standard deviation and that costs less than $20,000 per participant would more than pay for itself from both society's and the individual's viewpoint.

Why Do Skills Matter?

In the 1980s and 1990s the college wage premiumb what college graduates earn above those with only a high school educationb grew at unprecedented rates (Murphy & Welch, 1989). By the end of the 1990s it was more important than ever to enter the labor market armed with a college degree. What caused this explosion in the importance of a college education? Most analysts now agree that changes in the structure of the US economy led to a demand for more highly skilled workers that outstripped the ever-increasing supply of college graduates (Katz & Murphy, 1992). Changes in the goods and services we tended to produce, the design and structure of the workplace, and the tools used on the job were all geared to the abilities of more, rather than less, highly skilled individuals. Economists call this type of transformation "skill-biased technological change," that is, technological change that favored particular skill groups, in this case those with higher skills.

There is a convincing argument that the driving force behind the declining relative (and absolute) earnings of lower-skilled individuals comes from the same process: a workplace that on average requires higher skills. This interpretation suggests to some that increased public support for programs that would raise the cognitive skill levels of the least educated individuals, particularly school dropouts, would be an effective way to improve their economic outcomes. Policymakers and the public could be surprised, however, and actual benefits of such programs could fall substantially below the expected benefits. This would happen if shifts in the production technology used in low-skilled jobs have "deskilled" those jobs, unlike what has happened for more highly skilled jobs in the economy.

As a simple example, consider how technological advancements have altered the job requirements for a typical entry-level type job: check-out clerk. The adoption of optical recognition technology and computerized cash registers has meant that the ability to know basic math in order to calculate change is no longer required for counter clerks. Technological innovations may mean that the ability to smile while working on your feet all day is more important for many low-skill entry-level jobs than knowledge of basic math. If this is an accurate depiction of the kind of entry-level jobs open to dropouts, then there could be an overemphasis on cognitive skill development as a means of improving the economic conditions of low-educated individuals. My research, however, indicates that this is not the caseb in the types of entry-level jobs that first employ young dropouts, basic cognitive skills matter.

In Conclusion
Skills matter more in today's labor market than they ever have. But the ramifications of this have primarily been seen in terms of relatively highly skilled individuals. As my research shows, basic cognitive skills are also important for the least skilled in the labor market: young dropouts with low levels of education and little to no work experience. The message for students, schools, and adult education programs is clear. Schools should pay attention to skill formation for all their students, including those who seem destined to drop out before earning their diploma. Adult education programs should not sacrifice skill formation at the expense of strategies aimed more toward GED test-taking skills. Students should work hard while they are in school or in GED preparation programs to acquire the types of basic cognitive skills required for them to function fully in a modern democracy and economy. Individuals drop out of school for all kinds of reasons. It is inescapable that the accumulated set of cognitive skills they possess as they step into the labor market play a major role in determining their economic future.

Katz, L., & Murphy, K. (1992). "Changes in relative wages, 1963-1987: Supply and demand factors." The Quarterly Journal of Economics, 107(1), pp. 33-78.

Murphy, K., & Welch, F. (1989). "Wage premiums for college graduates: Recent growth and possible explanations." Educational Researcher, 18(4), pp. 17-26.

Schochet, P., Burghardt, J., & Glazerman, S. (2000). National Job Corps Study: The Short-Term Impacts of Job Corps on Participants' Employment and Related Outcomes, Final Report. Princeton, NJ: Mathematica Policy Research, Inc.

The study described in this article will be published in the Economics of Education Review and will eventually be a NCSALL Research Brief.

About the Author
John Tyler is an Assistant Professor of Education, Economics, and Public Policy at Brown University in Providence, RI, a faculty research associate at the National Bureau of Economic Research, and a NCSALL researcher. His work examines the economic returns to a GED, the importance of cognitive skills in the labor market, and the impact of working while in high school on academic achievement.

Read another Focus on Basics article by John Tyler.

Updated 7/27/07 :: Copyright © 2005 NCSALL