Application of Propensity Score Theory to the Measurement of the Effects ofMilitary Service

摘要

This report focuses on the suitability or propensity score theory to determinethe effects of military service on the later life of a Service participant. The limitations inherent in the non-experimental determination of the effects of service have previously precluded strong determination of cause, in part because of the bias introduced by self-selection into military service. Those who serve differ from those who do not serve in at least two ways: they have served, and they have chosen to serve. To attribute any differences later in life to the first of those variables while ignoring the second is not defensible. The present selection presents a method of simulating the phenomena so modeled, then illustrates the simulation with a sample execution. The output of the simulation is examined to determine whether plausible values of the effect of service in the output variables might reasonably be expected to be detected. The differences built into the simulation were recovered, but were not statistically significant.