The Causal Effects of Citizenship: Evidence from a Regression Discontinuity Design
Research project funded by the Swiss National Science Foundation (CHF/USD 181,000). PI (with Jens Hainmueller).
Despite heated debates about citizenship policy, there exists almost no rigorous evidence on how naturalization affects the lives of immigrants, their behavior, and outcomes for the communities in which they live. How valuable is Swiss citizenship for immigrants? To what extent does citizenship improve immigrants’ economic outcomes? How does it affect their level of social and political integration into their communities and into Swiss society in general? Answers to these questions are critical to inform ongoing policy debates in Switzerland and other countries, but research to date has failed to provide the necessary empirical evidence. A few existing studies examine the impact of citizenship by comparing the wages and employment of naturalized with non-naturalized immigrants, but these studies are plagued by strong selection bias, because immigrants selectively apply for citizenship based on differences in unobserved background characteristics (such as motivation, etc.) that severely confound the comparisons. Moreover, existing studies only focus on short term effects on wages and employment, but do not address long term effects and other important outcomes such as immigrants’ educational attainment, income, social and political integration, political behavior, political sentiments and perceptions. In other research we extensively study the drivers of anti-immigrant sentiments in Switzerland by examining how naturalization rates vary at the local level (SNF grant no. 100017_132004). Here we request support to carry out a new research project that for the first time isolates the causal effects of Swiss citizenship on a wide range of economic, social, and political outcomes. In order to overcome the selection bias, our proposed study relies on a so-called regression discontinuity design to identify the causal effect of naturalization. Prior to 2003, citizens voted on individual naturalization applications in many Swiss municipalities using secret ballot referenda (see Hainmueller and Hangartner 2013). We collected a database of all immigrants that applied for citizenship in these communities between 1970 and 2003 and surveying them now allows us to compare the outcomes of lucky and unlucky applicants that barely won or lost a majority of votes in their naturalization referenda. Balance checks suggest that for such referenda which are decided by a thin margin, the naturalization decision is essential as good as random so that narrowly rejected and narrowly approved applicants are very similar on all confounding characteristics. This allows us to remove the selection effects and obtain unbiased estimates of the long term impacts of citizenship that are about as credible as those obtained from a randomized experiment. While we have collected a comprehensive database with immigrant applicant characteristics (including contact information), in order to conduct this study we need to collect outcome data from the approved and rejected immigrants. We plan to collect this data with a newly developed survey instrument that measures a wide array of social, economic, and political outcomes including measures of social inclusion, economic success, and political participation and perceptions. A pilot test in which we surveyed 160 immigrant applicants revealed that our research design is feasible and the preliminary results look promising. Our regression discontinuity design estimates find that citizenship dramatically increases turnout, trust in institutions, and subjective well-being among the naturalized immigrants. Given the advantages in research design and the amount of new evidence collected, we are confident that this study will make an important contribution to our understanding of the value of citizenship and inform ongoing policy debates with the necessary empirical evidence.
Project period 2012-2015