The methodological core of our research program is a focus on systematic impact evaluations that leverage experimental and quasi-experimental methods with common study protocols to quantify the social and economic returns to integration policies across Europe and the United States. This approach is driven by a number of methodological hurdles that prior integration policy evaluations have failed to adequately address.
The Challenge
Why is it so challenging to estimate the impact of integration policy? It is easiest to think about the methodological issues by working through an example. So, let’s think about the impact of citizenship. Many people believe that citizenship generates lots of positive outcomes for immigrants and their communities; more and better paying jobs, and deeper participation as community members. To come to this conclusion people compare immigrants who have become citizens and immigrants who have not, and observe that naturalized citizens enjoy better outcomes (social, political, economic). The significant problem with this conclusion is that immigrants decide whether or not to apply for citizenship, and then the government either accepts or rejects the application (see the illustrative figure below). The immigrants that end up becoming citizens could differ in a number of ways from those immigrants who do not apply and/or have their applications rejected. Up until now researchers have not adequately assessed the effect of citizenship because of this double selection problem (see Hainmueller, Hangartner and Pietrantuono (2014) for an in-depth explanation).
The Solution
To address the aforementioned methodological issue and similar issues involved in evaluating other integration policies (asylum, integration contracts, treatment of undocumented immigrants, etc.), we use experimental and quasi-experimental evaluation techniques. For instance, to test the effect of citizenship, we will work directly with organizations who provide citizenship education workshops. Our partner organizations will send out information to a random sample of citizenship eligible immigrants. By surveying both immigrants who receive and do not receive this encouragement, we reduce the impact of selection bias and can provide a more concrete understanding of citizenship’s impact on social, political and economic outcomes. Additionally, we can use the timing and introduction of new policies like integration contracts—immigrants sign a document to commit to rapidly integrate into the host society by acquiring knowledge of the local language and traditions and making efforts to find a job and attain financial self-sufficiency—to understand their impact on immigrant integration. Using a quasi-experimental approach(regression discontinuity design paired with a differences-in-differences), we can minimize concerns about any observed effects associated with time. The end result of the experimental and quasi-experimental evaluation techniques will be rigorous evaluation of integration policies that can help balance the debate around immigration, and provide citizens and policy makers with knowledge to adapt and develop policies that work; for immigrants and their host societies.
