The Use of Propensity Score Matching in Evaluation

Duration:

October 2001 to June 2002

PSI researchers:


Background, summary and aims

Background

This project discusses the value of propensity score matching in evaluating policy interventions, particularly labour market programmes. The technique is used to make causal inferences about a treatment in the absence of random assignment. The paper discusses the theory underpinning the approach, and the issues that arise in its empirical application. The paper is written for an audience including policy advisers, government analysts and academics.

Study Design

The study is based on a review of the theoretical and empirical literatures, and draws on recent evaluations to illustrate points of interest. The work is joint with Susan Purdon at the National Centre for Social Research.

Importance of Research

The research is primarily aimed at government analysts and researchers who commission research and interpret its findings. It is hoped that the paper will contribute to their understanding of this technique, which is increasingly popular among evaluation researchers trying to isolate the effect of programmes on outcomes such as employment, wages and benefit receipt.

Publications

Bryson, A., Dorsett, R. and Purdon, S. (2002) The use of propensity score matching in the evaluation of active labour market policies, Department for Work and Pensions Working Paper No.4. Available online at http://www.dwp.gov.uk/asd/asd5/wp-index.html.