What works when it comes to preventing eviction? There are many different policies, programs, and technology interventions. On this page, we highlight the growing number of studies that evaluate different interventions.
- How can we evaluate eviction prevention efforts?
- What do we know about what works to prevent evictions?
How can we evaluate eviction prevention efforts?
Your evaluation plan will depend on the type of program, as well as the context that you’re in. We have outlined some of the most common study designs here, with links to more resources on these program evaluations.
Random Controlled Trials (RCTs)
Randomized Controlled Trials are often considered the gold standard of evaluating policy impact. See below for many examples of RCTs used to study whether an intervention helps stop eviction judgments, produces better outcomes for low-income tenants, and other key outcomes.
- With RCTs, researchers identify what intervention is supposed to lead to what outcome. They also identify what population is being targeted.
- That target population is then randomly assigned to either recieve the intervention ‘treatment’ or be in a control group.
- Then the resarchers study the outcomes for the traeatment group and the control group. Are there significant differences in outcomes among these 2 groups?
- If there are signficiant differences, then this can be attributed to the intervention. This kind of analysis can provide reliable evidence about whether a treatment does produce its intended outcomes. RCTs account for other time trends, and address the counterfactual (what would have happened to a person if they didnt’ recevive the treatment?).
See more at the World Bank’s RCT study design page.
Difference-in-Differences evaluation can be a useful alternative, especially when it’s not feasible or ethical to do an RCT.
- Like with an RCT, the research team identifies the intervention, the intended outcome, and the target population. They must also identify another group that is similar to the treatment group (who gets the program) to be the control group (who does not receive the program).
- The team will track data on these two groups before the intervention and after it.
- They will compare what the difference is for before the intervention and after it for the control group. Then they’ll do the same calculation for the treatment group.
- What’s the difference between these two differences? Has there been a bigger change for the treatment group, then there has been for the control group? This will indicate whether the intervention has had the expected impact.
See more at the World Bank’s data study resources.
Surveys and qualitative interviews can help you understand changes that are not necessarily recorded in data sets. They can help uncover important behaviors, events, experiences, and dynamics.
- The research team should identify what outcomes or scenarios they want to study. It could be about the experience of certain events, sense of procedural justice, the occurrences in the court procedure, etc.
- They should identify what existing instruments they can use to capture and measure these outcomes. It is much easier than creating a brand new survey instrument.
- The survey could be given before the intervention and after. It could be done as an exit interveiew immediately after receiving a service , or it could be a few months after a crisis to learn more about what occurred.
See more at the World Bank’s page on survey design and pilots.