Much attention has been paid to the issue of possible nonlinearities in the relationship between log wages and schooling in the literature on both the United States and developing countries. The author uses data from a recent household survey for the Philippines, the 1998 Annual Poverty Indicator Survey, to test the fit of the log-linear specification for Filipino men. He presents results based on various estimation strategies, including spline regressions and semi-parametric regressions with a large number of dummy variables for years of schooling and experience. He concludes that: 1) There appear to be large differences between rates of return to education across different levels in the Philippines. The wage premia for both primary and secondary education are significantly smaller than those for tertiary education. 2) Within each level - primary, secondary, and university - the last year of schooling is disproportionately rewarded in higher wages. That is, there appear to be clear sheepskin effects associated with graduation.
(With permission from the World Bank Group)
