Sources of Bias in Women's Wage Equations: Results Using Sibling Data
利用姐妹数据同时处理女性工资方程估计中的异质性偏差和内生性偏差，发现OLS估计中教育回报存在向上偏差，且白人女性的婚姻溢价为正。
We use data on sisters to jointly address heterogeneity bias and endogeneity bias in estimates of wage equations for women. This analysis yields evidence of biases in OLS estimates of wage equations for white and black women, some of which are detected only when these two sources of bias are addressed simultaneously. For both white and black women there is evidence of upward bias in the estimated returns to schooling. Bias-corrected estimates of the effect of marriage on wages, for white women, suggest a positive marriage premium. We also use the sibling data to identify our models, and test a number of other commonly used identifying assumptions as overidentifying restrictions.