Counting participation is important. Understanding what changed because of a program is harder.
Program dashboards often report how many people participated and what services they received. These measures are essential for operations, but they do not show whether outcomes changed because of the program.
Causal evaluation begins with a credible comparison. Depending on the context, difference-in-differences or another quasi-experimental approach may help distinguish program effects from broader trends. These methods depend on assumptions that must be examined, not merely named.
Where credible identification is not possible, researchers should say so. Thoughtful descriptive tracking, beneficiary interviews, and a clearly specified theory of change may still produce valuable learning.
Good impact measurement is proportionate to the decision. It defines what is knowable now, what evidence must be collected next, and how findings will influence program design.