In the United States and Europe, illegal immigrants cannot work or start a new economic activity, at least officially. Such immigrants can work only in the informal economy, and thus receive lower earnings than legal immigrants (Kossoudji and Cobb-Clark, 2002, Kaushal, 2006, Amuedo-Dorantes et al., 2007, Accetturo and Infante, 2010, and Lozano and Sorensen, 2011). According to the Becker-Ehrlich model of crime, a lower income from legitimate activities means a higher propensity to participate in illicit activities.
The presence of large illegal populations therefore raises crime concerns in destination countries. According to an annual survey conducted in North American and European countries, approximately two-thirds of the people interviewed are concerned that illegal immigrants increase crime; only half express the same concern about legal immigrants (Transatlantic Trends, 2009). Moving from perceptions to statistics, illegals constitute 20–30 percent of all immigrants in Italy but represent 80 percent of those arrested for serious crimes (Italian Ministry of Interior, 2007).
The higher criminal propensity of illegals, however, may reflect differences between legal and illegal residents, as opposed to the (causal) effect of legal status. In particular, illegal immigrants are typically young, single, male, and less educated than legal immigrants (Cohn and Passel, 2009; Mastrobuoni and Pinotti, 2014; and Caponi and Plesca, 2013). More generally, the two groups could differ along other dimensions relevant to criminal behavior. For instance, individuals who are less risk averse or have a higher propensity to violate laws would be more likely to reside illegally in a country and to commit crime. It is thus difficult to identify the causal effect of legal status on crime committed by immigrants.
The institutional framework in Italy, however, allows estimation of the causal impact of legal status on criminal behavior. The primary path to legal status in Italy is a work-related residence permit. Fixed quotas of permits are available each year, and applications must be submitted online by prospective employers, starting on a given “Click Day.” Applications are processed on a first-come, firstserved basis until the quotas are exhausted. The rationing of quotas, and the frequency of applications during the first hours of Click Days, are such that several thousand applicants are denied legal status every year simply because their employer applied a few minutes or seconds after the cutoff time.
My research matches data on applicants for 2007 with restricted-use data on all foreigners prosecuted for a serious crime in Italy in the subsequent year. I then compare the number of crimes per applicant between those who applied shortly before versus shortly after the cutoff. For most categories of immigrants, this cutoff occurred fewer than 30 minutes after the start of the Click Day. Most important, the exact timing of the cutoff was unknown ex ante, as it depended on the timing of all applications and on how many were rejected for being inaccurate, false, or incomplete.
These complexities provide a compelling argument that immigrants applying just before versus just after the cutoff differ only due to randomness, not any inherent characteristics. This means comparison of these two groups tells us about the impact of legal status per se.
My empirical results suggest that legal status significantly reduces the number of serious crimes committed by immigrants in Italy, from 2.9 to 1.2 per 100 applicants in the first year after Click Day. When I distinguish between different types of crimes and different categories of applicants, the effect is greater for economically motivated crimes, applicants experiencing better economic opportunities after obtaining legal status, and applicants subject to a higher risk of expulsion as illegal immigrants. These findings contribute to the growing body of evidence regarding the effect of immigration on crime. Earlier work by Butcher and Piehl (1998) shows no evidence that immigration affected crime across U.S. counties during the 1980s, whereas Spenkuch (2014) reaches a different conclusion for subsequent periods.
Borjas et al. (2010) find that immigration increases crime, although only indirectly (by raising the crime rate of native black males).
For other countries, Bianchi et al. (2012) conclude that the causal effect of immigration on crime across Italian provinces is not significantly different from zero, while Alonso-Borrego et al. (2012) obtain an opposite result across Spanish provinces. Finally, Bell et al. (2013) distinguish between two large immigrant waves in the UK, namely, asylum seekers and the post-2004 inflow from EU accession countries. Only the former group, which was characterized by limited economic opportunities in official labor markets, caused a significant increase in (property) crimes.
This last result suggests that average estimates across geographical areas may mask significant heterogeneity in the criminality of different immigrant groups. Indeed, the propensity to commit crimes should differ across immigrants depending on individual characteristics, such as age, gender, and education, in the same manner as for natives. However, immigrants differ among themselves with respect to another important factor: legal status.
Previous evidence on the effects of legal status relies primarily on the 1986 Immigration Reform and Control Act (IRCA), which granted legal resident status to long-time illegal immigrants in the United States. Baker (2013) shows that legalization reduced crime across counties. At the same time, the IRCA enforced stronger control over hiring illegal immigrants, creating obstacles to employment for those not legalized. Freedman et al. (2013) analyze these provisions and document increased felony charges filed against Hispanic residents of San Antonio, Texas, after the expiration of the IRCA amnesty deadline.
For other countries, Mastrobuoni and Pinotti (2014) exploit variation in legal status among pardoned prison inmates in Italy after the EU expansion of January 2007. They find that after the EU accession, recidivism declined markedly among inmates from new EU member countries, whereas no change occurred in a control group of inmates from EU-candidate member countries.
These results can inform the current debate on immigration policy. In the context of increasing pressures at the border as a result of economic and political turmoil in several areas of the world, immigration reform is near the top of the agenda in the United States and other countries. Switzerland, for instance, passed a referendum in February, 2014, to reduce immigration quotas and limit the free mobility of European Union citizens.
My research suggests that the effect of such restrictions will depend crucially on the effective degree of enforcement. If enforcement is low, implying that substantial numbers of illegal immigrants enter the country, then the ultimate effect of the policy is to exclude these immigrants from the official labor market and thus raise the risk that they become involved in criminal activity.