Statistical Tools for Program Evaluation

Jean-Michel JOSSELIN is professor of economics at the University of Rennes 1 (France). His research and teaching interests include law and economics, public policy evaluation, with a focus on public health (cost effectiveness analysis, decision analytic modelling in oncology and chronic diseases). He is researcher at the French National Center for Scientific Research (CREM-CNRS and Condorcet Center for Political Economy), affiliated with the health research chair Hospinnomics (AP-HP and Paris School of Economics), member since 2012 of the Health Economics and Public Health Committee (CEESP) of the French National Authority for Health (HAS).Benoît Le Maux is associate professor of economics at the University of Rennes 1 (France). His research and teaching relate to public policy evaluation and democracy, with a special emphasis on public expenditures (microeconomic analysis of collective decision-making, econometrics, quasi-experimental techniques). He is researcher at the French National Center for Scientific Research (CREM-CNRS and Condorcet Center for Political Economy) and currently in charge of coordinating a master program in public finance and management at the Faculty of Economics of the University of Rennes 1.

• Statistical Tools for Program Evaluation: introduction and overview.- Part I: Identifying the context of the program. Description of problems and needs, assessment of relationships among variables.- Sampling and construction of variables.- Descriptive statistics and interval estimation.- Measuring and visualizing associations.- Econometric analysis.- Estimation of welfare changes.- Part II: Ex ante evaluation. Assessment of different policy strategies that might solve the problems.- Financial appraisal.- Budget impact analysis.- Cost benefit analysis.- Cost effectiveness analysis.- Multi-criteria decision analysis.- Part III: Ex post evaluation. Assessment of the effects of a strategy after its implementation.- Follow-up by benchmarking.- Randomized controlled experiments.- Quasi-experiments.