There is a growing interest, notably in development economics, in extending project evaluation methods to the evaluation of multiple interventions (“programs”). In program evaluations one is interested in the aggregate impact of a program rather than the effect on individual beneficiaries. In many situations randomized controlled trials cannot identify this impact. We propose a measure of program impact, the total program effect (TPE), which is a generalization of the treatment effect on the treated (ATET). We show how the TPE can be estimated.
program evaluation randomized controlled trials policy evaluation treatment heterogeneity budget support sector-wide programs aid effectiveness