# Why is this level of evaluation important? Interventions in the development sector aim to improve the quality of people’s lives. Impact evaluations (IEs) measure the effects of the intervention on outcomes such as mortality, learning outcomes, and earnings. The main issue these evaluations face is that the world is a messy place: as an intervention is being implemented, many other things are happening that would make a simple before-and-after comparison a bad way to judge program effectiveness. To address this, we consider the counterfactual: what would have happened to the same people in the absence of the intervention. Because we cannot observe both realities at once (the same people with and without the intervention), we estimate the counterfactual using a comparison group that is as similar as possible to the group that received the intervention. It represents what would have happened without the program. Comparing outcomes across these groups helps us isolate the intervention’s impact. There are a number of ways to estimate or measure the counterfactual. The most straightforward approach is usually a randomized controlled trial (RCT). In an RCT, participants are randomly assigned to one or more treatment groups that receive an intervention (or variants of it) and a comparison group that does not. Researchers then measure outcomes across groups. Well-designed randomized evaluations enable credible, and are less-prone-to-bias estimates of causal impact—that is, which changes in participants’ lives can be attributed to the program. Other techniques for the counterfactual construction include propensity score matching, difference in differences, and regression discontinuity designs. These are discussed further below, but in general require more technical econometric expertise and contextual knowledge in order to execute well.
***
💬 Want to suggest edits or provide feedback? {% embed url="" %}