Stephen B. Jarrell

Gender Wage Discrimination Bias? A Meta-Regression Analysis

This study provides a quantitative review of the empirical literature on gender wage discrimination. Although there is considerable agreement that gender wage discrimination exists, estimates of its magnitude vary widely. Our meta-regression analysis (MRA) reveals that the estimated gender gap has been steadily declining and the wage rate calculation to be crucial. Large biases are likely when researchers omit experience or fail to correct for selection bias. Finally, there appears to be significant gender bias in gender research. However, it is a virtuous variety where researchers tend to compensate for potential bias implicit in their gender membership.

Declining Bias and Gender Wage Discrimination? A Meta-Regression Analysis

This paper extends, tests, and revises a previous meta-regression analysis of the gender wage gap (Stanley and Jarrell 1998). We find that there remains a strong, though dampened, tendency for discrimination estimates to fall, and male researchers still report significantly larger amounts of wage discrimination against women. This extensive research base, containing 104 estimates, suggests that there is less need to correct for selection bias—an indirect sign of lessened discrimination. There is evidence that gender research is changing and improving. Although gender wage discrimination has lessened, the research base still finds a significant gender wage inequality.

Could It Be Better to Discard 90% of the Data? A Statistical Paradox

Conventional practice is to draw inferences from all available data and research results. When a scientific literature is plagued by publication selection bias, a simple discarding of the vast majority of empirical results can actually improve statistical inference and estimation. Simulations demonstrate that, if statistical significance is used as a criterion for reporting or publishing estimates, discarding 90% of the published findings greatly reduces publication selection bias and is often more efficient than conventional summary statistics. Improving statistical estimation and inference through removing so much data goes against statistical theory and practice; hence, it is paradoxical. We investigate a very simple method to reduce the effects of publication bias and to improve the efficiency of summary estimates of accumulated empirical research results that averages the most precise 10% of the reported estimates (i.e., ‘Top10’). In the process, the critical importance of precision (the inverse of an estimate’s standard error) as a measure of a study’s quality is brought to light. Reviewers and journal editors should use precision, when possible, as one objective measure of a study’s quality.