Propensity Score Matching and Policy Impact Analysis: A Demonstration in Eviews

59 Pages Posted: 6 Oct 2006

Date Written: April 2006

Abstract

Effective development policymaking creates a need for reliable methods of assessing effectiveness. There should be, therefore, an intimate relationship between effective policymaking and impact analysis. The goal of a development intervention defines the metric by which to assess its impact, while impact evaluation can produce reliable information on which policymakers may base decisions to modify or cancel ineffective programs and thus make the most of limited resources. This paper reviews the logic of propensity score matching (PSM) and, using data on the National Support Work Demonstration, compares that approach with other evaluation methods such as double difference, instrumental variable, and Heckman's method of selection bias correction. In addition, it demonstrates how to implement nearest-neighbor and kernel-based methods, and plot program incidence curves in EViews. In the end, the plausibility of an evaluation method hinges critically on the correctness of the socioeconomic model underlying program design and implementation, and on the quality and quantity of available data. In any case, PSM can act as an effective adjuvant to other methods.

Keywords: EViews, Double Difference, Impact Analysis, Instrumental Variables, Kernel

JEL Classification: C14, C81, C87, I38

Suggested Citation

Essama-Nssah, B., Propensity Score Matching and Policy Impact Analysis: A Demonstration in Eviews (April 2006). World Bank Policy Research Working Paper No. 3877, Available at SSRN: https://ssrn.com/abstract=935403

B. Essama-Nssah (Contact Author)

World Bank ( email )

Washington, DC 20433
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
803
Abstract Views
2,776
Rank
56,802
PlumX Metrics