The Usefulness of Predicted Values for Prior Work Experience in Analyzing Labor Market Outcomes for Women
提出一种改进方法，在缺乏实际工作经验数据时，利用其他数据集按职业预测工作经验，用于估计女性工资方程，并证明该方法优于使用潜在经验或忽略职业的预测值。
This paper proposes an improved way of treating experience in estimating wage equations for women when measures of actual experience are lacking. It shows that using a predicted value for experience from occupation-specific equations estimated on another data set containing actual experience is preferable to using either potential experience (time since school leaving) or predicted experience without taking account of the woman's occupation. Results also show that the use of potential experience may bias the estimated impact offactors such as race and schooling on women's wages.