Beyond the average effect of the innovation subsidies: Using case selection via matching to break impasse in delivering useful advice to policy makers

Abstract

Experts and stakeholders promote the combined use of counterfactual and theory-based approaches in program evaluation. We illustrated combined application of these two approaches in a single evaluation study of innovation subsidies, using "case selection via matching" and follow-up in-depth interviews. We conducted interviews in contrasting pairs of companies-one successful and one unsuccessful-which were otherwise similar on a defined set of covariates. Our procedure helped to reveal factors, which might be overlooked or simply not available in data at hand and hence not accounted for in analyses of the intervention effects. As such it extends beyond the average effect estimate to highlight causes of an intervention success or failure.

Keywords: Case selection via matching; Case study; Counterfactual impact evaluation; Innovation support; Theory-based evaluation.