Summary: | Fr lich, who is not further identified, summarizes the estimation methods for dichotomous selectivity models he finds most promising for structural analysis and evaluation studies in econometrics. Their purpose is to reduce selection bias, which results from the fact that which people choose to do one thing or another (undergo treatment or not, for example) is not random, but related to other factors that may or may not be observable. The core idea of all the approaches is forecasting counterfactual outcomes. Annotation c. Book News, Inc., Portland, OR (booknews.com)
|