Background
Propensity Score Matching (PSM) is a statistical technique employed in observational studies to estimate the causal effect of a treatment, intervention, or policy. The main goal of PSM is to mimic the conditions of a randomized controlled trial (RCT) by comparing treated units with a control group that is statistically similar in terms of observed covariates.
Historical Context
PSM has its intellectual roots in the work on causal inference by Donald Rubin and Paul Rosenbaum in the 1980s. Their pioneering research laid the groundwork for modern uses of the method, giving researchers a powerful tool to approximate randomized experiments in non-experimental settings.
Definitions and Concepts
Propensity Score: The probability of receiving the treatment given a set of observed covariates.
Matching: The process of pairing treated units with control units that have similar propensity scores.
Causal Effect: The difference in outcomes between the treated and matched control units, used to estimate the treatment’s impact.
Major Analytical Frameworks
Classical Economics
Classical economic models seldom deal directly with non-experimental data, focusing instead on theoretical foundations and market mechanisms.
Neoclassical Economics
Neoclassical models often assume perfect information and sometimes employ observational data; still, they generally depend less on techniques like PSM compared to more empirical fields.
Keynesian Economics
Keynesian Economics typically focuses on macroeconomic policies where PSM might be relevant in evaluating the effects of fiscal or monetary interventions in non-experimental data.
Marxian Economics
This school emphasizes the structural causes of economic phenomena, where PSM can be used to assess the impacts of policy on class structures and labor relations.
Institutional Economics
Institutional economists may employ PSM to study the impacts of policies aimed at altering institutional arrangements or behavioral norms.
Behavioral Economics
Behavioral economists use PSM to estimate the causal effects of interventions meant to change individual or group behaviors.
Post-Keynesian Economics
This branch may use PSM for macro policy evaluations, especially where randomized controls are not feasible.
Austrian Economics
Austrian economics relies less on empirical methods including PSM, often favoring qualitative assessments.
Development Economics
This field extensively uses PSM to evaluate the impacts of policy interventions in developing countries, helping to identify effective strategies for economic development.
Monetarism
Monetarists might employ PSM to assess the effects of monetary policy changes in non-experimental settings.
Comparative Analysis
Traditional econometric techniques—like regression analysis—are often complemented by PSM, especially when addressing issues of selection bias in observational studies.
Case Studies
- Evaluating the impact of vocational training programs on employment rates.
- Assessing the outcomes of health interventions in non-experimental settings.
- Measuring the effects of educational policies on student achievement.
Suggested Books for Further Studies
- “Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction” by Guido W. Imbens and Donald B. Rubin
- “Matching, Regression Discontinuity, Difference in Differences, and Beyond” by Matias D. Cattaneo and Sebastián Calónico
Related Terms with Definitions
- Randomized Controlled Trial (RCT): An experimental design for testing the efficacy of interventions.
- Covariates: The observable characteristics that are used to create matched pairs in PSM.
- Selection Bias: The bias introduced when the treatment group and control group differ in ways that affect the outcome.
- Causal Inference: Methods used to infer the causal effects of treatments or policy interventions.