Unconditional Effects of General Policy Interventions

TL;DR

This paper develops a semiparametric framework to evaluate the unconditional effects of broad policy interventions, including location-scale and simultaneous shifts, with applications to education and health outcomes.

Contribution

It introduces a novel semiparametric estimation method for unconditional policy effects accounting for complex shifts in covariates.

Findings

Estimates the effects of education on wages using the proposed method.

Analyzes the impact of smoking during pregnancy on birth weight.

Demonstrates the method's finite sample performance through simulations.

Abstract

This paper studies the unconditional effects of a general policy intervention, which includes location-scale shifts and simultaneous shifts as special cases. The location-scale shift is intended to study a counterfactual policy aimed at changing not only the mean or location of a covariate but also its dispersion or scale. The simultaneous shift refers to the situation where shifts in two or more covariates take place simultaneously. For example, a shift in one covariate is compensated at a certain rate by a shift in another covariate. Not accounting for these possible scale or simultaneous shifts will result in an incorrect assessment of the potential policy effects on an outcome variable of interest. The unconditional policy parameters are estimated with simple semiparametric estimators, for which asymptotic properties are studied. Monte Carlo simulations are implemented to study their finite sample performances. The proposed approach is applied to a Mincer equation to study the effects of changing years of education on wages and to study the effect of smoking during pregnancy on birth weight.