CitationDiaz, Juan Jose & Handa, Sudhanshu (2006). An Assessment of Propensity Score Matching as a Nonexperimental Impact Estimator: Evidence from Mexico’s PROGRESA Program. Journal of Human Resources, 41(2), 319-345.
AbstractNot all policy questions can be addressed by social experiments. Non-experimental evaluation methods provide an alternative to the experimental design, but their results depend on non-testable assumptions and therefore are less clear. This paper presents evidence on the reliability of propensity score matching (PSM), which estimates treatment effects under the assumption of selection on observables, using a social experiment that was launched to evaluate the PROGRESA program in Mexico. We test the ability of PSM to select a comparison group that is identical to the control group in the randomized experiment, a direct measure of bias in the PSM technique. We find that PSM performs well for outcomes that are measured comparably across survey instruments and when a rich set of control variables is available. However, even small differences in the way outcomes are measured can lead to bias in the technique.
Reference TypeJournal Article
Journal TitleJournal of Human Resources
Author(s)Diaz, Juan Jose