College Wage Premium: A Meta-Analysis

This meta-analysis is focused on the examination of college wage premium - the wage differential between college graduates and individuals without a college degree, using the most recent and state-of-art methods. To conduct the analy- sis we reviewed 41 studies, from which we collected 567 estimates. The findings show presence of a mild publication bias and p-hacking. The college wage pre- mium corrected for publication bias differ from 0.293 to 0.386, but in most of the test the estimates are below mean (0.349) showing mild positive publication bias. The findings indicate that the individuals with a college education have wages that are 29% to 39% higher than those without a college degree. By employing Bayesian model averaging, together with Frequentist check and Fre- quentist model averaging, we investigate the heterogeneity between estimates. Out results suggest that controlling for gender, unemployment rate, and study major are few of the significant variables affecting the college wage premium. While the first two mentioned have negative relationship on the college wage premium, controlling for study major increases college wage premium