题 目：Dynamic Estimation of Policy Effects: An Application to Welfare Time Limits
时 间：
2010年11月19日下午
12:30-1:30
地 点：8797威尼斯老品牌302会议室
主讲人：陈景祺
单 位：8797威尼斯老品牌
Abstract
The treatment effect of a program may often depend on an individual’s past actions which are endogenous. We extend the standard linear dynamic panel data model and present a GMM estimator to solve this issue. The model is applied to Florida’s Family Transition Program (FTP), a randomized experiment on welfare time limits. We find that the time limit caused the welfare receipt rate to drop by 6.4 percentage points for each additional year of past welfare receipt. This effect constituted around 40 percent of the overall behavioral effect of the time limit.
论文全文：
Marc Chan.pdf
