| Herbert Turner
University of Pennsylvania
Are advanced high school mathematics courses predictors of earnings in the labor market?
FINAL REPORT
Since 1984 a major goal of education reformers has been improving math achievement. Much attention has been given to formulating policies, raising standards, strengthening curricula, and allocating additional resources to encourage all K-12 pupils in the United States, not just those who pursue post-secondary education, to complete mathematics courses that are advanced.
Although one study on direct effects of math coursework on earnings found a positive effect for female college graduates, a general conclusion reached based on past research evidence is that completing more advanced high school mathematics courses has no effect on subsequent earnings. The evidence is not conclusive, however. This dissertation addresses a number of the limitations of earlier evidence, and provides better evidence on the important question: What is the relationship between high school mathematics coursework and earnings in the labor market?
The data for this study were drawn from the National Longitudinal Survey of Youth 1979 (NLSY79), High School and Beyond (HSB), and the Occupational Network for Employment and Training (ONET). The last was used to augment the NLSY79 with such occupational characteristics as the level of math knowledge required to perform the job well. This study's target population included U.S. high school graduates with the following educational characteristics: 1) A high school diploma only; 2) some college (one, two, or three years); and 3) a college degree (four years or more). To estimate the parameters in the models proposed, ordinary least squares (OLS) and generalized least squares (GLS) were used. The HSB, NLSY79, and ONET, repeated contrast coding, and multiple imputation (MI) were used to address some of the methodological limitations of previous research.
This study found that completing Algebra I and Geometry in high school were associated with positive changes in earnings. This association, was robust to statistical controls and the use of multiple imputation which retained respondents in the analysis sample which would have been excluded due to listwise deletion. Moreover, this association applied to HSB respondents and varied by highest education attainment and gender. This finding was not discernible for NLSY79 respondents.
The study includes three caveats to be factored into the conclusions. First, earnings can also affect the completion of math courses; but determining this effect was beyond the scope of this study. Second, findings are generalized only to the target population. Third, the mathematics courses for which effects were found were taught prior to today's Standards reform movement, which emphasizes group work, applied problem-solving, and higher-order thinking.
Back to Funded Dissertation Grants Page |