Does School Quality Matter? Returns to Education and the Characteristics of Schools in South Africa

Preprint
Publisher Website
Google Scholar
    • preprint
      • Published in RePEc
Abstract
The effect of school inputs on labor market outcomes is an important and controversial topic, both in the United States and in developing countries. A large literature about American schools has not settled debate on the issue. Card and Krueger (1992) estimate the effect of pupil/teacher ratios and teachers? salaries on the rate of return to education for men born between 1920 and 1950, observed in the 1980 census. Controlling for state of birth effects, state of residence effects, and differences in returns to education between regional labor markets, they find a large, negative and significant effect of pupil/teacher ratios on the rate of return to education. Hanushek et al. (1996) and Heckman et al. (1996) challenge aspects of the Card and Krueger analysis. Hanushek et al. notes that the level at which school characteristics are aggregated affects the estimation results, and claims that aggregation biases upward estimated school quality effects. Heckman et al., replicating the Card and Krueger results in the 1970, 1980, and 1990 censuses, makes clear the importance of allowing for non-linearities in the returns to education when estimating the impact of school resources, and of allowing for differences in the impact of school quality across labor markets. They find school quality effects are weak for those with exactly 12 years of schooling, and strong only for those who attend college. Card and Krueger (1996) present a thorough review and discussion of this literature, but debate on how to interpret the literature on American schools is far from settled.

This publication has 0 references indexed in Scilit: