Who Matches? Propensity Scores and Bias in the Causal Effects of Education on Participation

View Researcher's Other Codes

Disclaimer: The provided code links for this paper are external links. Science Nest has no responsibility for the accuracy, legality or content of these links. Also, by downloading this code(s), you agree to comply with the terms of use as set out by the author(s) of the code(s).

Authors John A. Henderson, Sara Chatfield
Journal/Conference Name THE JOURNAL OF POLITICS
Paper Category
Paper Abstract Matching is increasingly being used in political science to reduce selection bias in casual estimation. In a recent study, Cindy Kam and Carl Palmer employ propensity score matching to assess the decades-long consensus that college attendance increases political participation. The authors find no effect, upending a major pillar in social science. While we agree that selection is a serious concern, we argue it is so substantial in the choice to attend college that good matches may be very difficult to obtain. In this situation, propensity score matching may actually increase bias. We match on nearly 800,000 propensity scores and use genetic matching to recover better matches with lower covariate "imbalances". We find even modest improvements reverses the null finding, however, no matching approach yields unbiased estimates. Ultimately, we show that 'balance' in the covariates and robustness to sensitivity diagnostics should guide the matching enterprise.
Date of publication 2011
Code Programming Language R

Copyright Researcher 2022