Attribution, in impact evaluation, is the task of establishing how much of an observed change in outcomes can be credited to a particular programme, rather than to other influences such as economic growth, weather or other projects. It requires an estimate of the counterfactual: what would have happened to the same people without the programme.
The strongest method is the randomised controlled trial (RCT), which randomly assigns eligible people to a treatment group and a control group and compares them over time. Where randomisation is not possible, evaluators use matched comparison groups, difference-in-differences or regression methods, starting from a baseline. In microfinance, a group of RCTs published in the 2010s found more modest effects on household income than earlier claims, largely because they could attribute changes more rigorously. Attribution is often contrasted with contribution analysis, which asks whether a programme plausibly helped bring about a change among several causes, without isolating its share. Both are part of impact assessment.
Ref 1270000015