Online Program Home
My Program

Abstract Details

Activity Number: 83 - Applications in Surveys and Social Science
Type: Contributed
Date/Time: Sunday, July 28, 2019 : 4:00 PM to 5:50 PM
Sponsor: Government Statistics Section
Abstract #306692
Title: Causal Inference for Policy Analysis: When Programs for Some Affect Outcomes for Others
Author(s): Daniel Wilmoth*
Companies: U.S. Small Business Administration
Keywords: economics; policy analysis; causal inference; SUTVA; experiment; market

Economists and policy analysts increasingly use experimental and quasi-experimental research methods to predict the effects of policy decisions. The identification of causal relationships through such methods typically relies on an assumption that treatment affects only the treated, sometimes called the Stable Unit Treatment Value Assumption (SUTVA). This essay argues that the assumption is rarely satisfied in policy contexts, where populations are richly linked through markets and other mechanisms. The redirection of resources to implement a treatment may violate the assumption, as when an experimental school competes with other schools for teachers. The performance of the treated relative to the untreated may influence results for both groups, as when college admission criteria incorporate class rankings. Economic outcomes like income reflect wages and other prices, which are shared by all participants in a market and jointly determined by their decisions. This essay presents an analysis of causal relationships when the SUTVA is violated, describing conditions and methods for identifying relationships that can be used to predict the effects of policy decisions.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program