Hourly Earnings in the United States: Another Look at Unionization, Schooling, Sickness, and Unemployment Using PSID Data
利用面板数据估计工会化、教育、疾病和失业对个人时薪的影响，发现工会效应约15%-20%，疾病长期影响约8%，失业效应短期显著但衰减快，教育效应因工具变量不足难以准确估计。
This paper makes use of panel data to consider the effects of unionization, schooling, sickness, and unemployment on individual earnings. Our results may be summarized as follows. The unionization effect is around 15%-20%. This measure is consistently estimated controlling for both fixed individual attributes and serially uncorrelated measurement error. Sickness has little immediate impact, but there is some indication that after 4 years earnings have fallen permanently by around 8% consequent on a 10-week spell of sickness. A spell of unemployment, on the other hand, leads to a sharp fall in earnings, but this effect decays rapidly leaving the possibility of a small permanent impact. Finally, controlling for all fixed individual attributes, we are unable to pin down the schooling effect with any accuracy owing to the absence of instruments that are both valid and effective within the data set.