Abstract
This paper aims to improve our understanding of emerging patterns of interior immigration control in the United States by examining local law enforcement cooperation with federal immigration authorities under the 287(g) Program. While several recent studies have drawn attention to the shifting terrain of immigration enforcement away from borders into the interior, few have attempted to systematically explain reasons for this shift. Using a county-level dataset of all counties in the United States, this study finds that, despite the purported links between the 287(g) Program and public safety, county crime rates are not significantly related to local cooperation with federal immigration authorities. Rather, the results indicate that local decisions to engage in immigration control efforts are driven by political factors, particularly a county's partisan composition, and demographic pressures related to increases in the Hispanic/Latino population. Demography is not, however, destiny, as new immigration in localities with established foreign-born populations does not significantly increase the likelihood of cooperation under 287(g). This suggests that the partisan composition of the places where immigrants settle mediates the extent to which the demand for tighter immigration enforcement exceeds the political influence immigrant communities have in their efforts to contest restrictive immigration control policies.
Acknowledgements
My thanks to two anonymous JEMS referees, whose insightful comments have greatly strengthened the article.
Notes
1. Examples of restrictive anti-immigrant ordinances include efforts to compel landlords to verify the immigration status of prospective tenants, denying business licences or city contracts to those who hire undocumented immigrants, using local police to facilitate deportations in conjunction with Immigration and Customs Enforcement (ICE), and to establish English as the official language of the city. It is also important to note the competing measures that cities have passed that explicitly protect the rights of undocumented immigrants, including so-called ‘sanctuary ordinances’ and other measures that prohibit local cooperation with ICE or the use of public funds to enforce federal immigration laws.
2. Macro-economic trends are often viewed as the most reliable predictor of control. As the economy grows, control policies become less restrictive as the demand for labour increases; as economies contract, policies become more restrictive. In other words, political economy accounts view immigration policy as oscillating in tandem with the booms and busts of the economy. The empirical record suggests that this relationship is, however, by no means absolute. As Money (Citation1997) notes, countries have responded differently to immigration despite similar macro-economic characteristics since at least the 1950s.
3. Moreover, the IACP report notes that state and local police do not have the training or expertise to enforce immigration laws. Given shrinking local budgets, many executives feel they do not have the resources to tackle the immigration issue.
4. Past research of local government policies towards immigrant integration indicates that the ideological and partisan leanings of the electorate and of governing institutions play an important role. For instance, in a 2003 survey of more than 300 Californian cities, Lewis and Ramakrishnan (Citation2007) found that municipal governments with Republican-leaning electorates and conservative city councils were less likely to provide translation of City Hall documents and interpretation services in public meetings.
5. Census data are primarily from 2000. While more recent data are available from the Census American Communities Survey (ACS), the coverage of these data are limited to large counties only. The 2005–07 ACS provides data on the population characteristics of all geographical areas in the United States with populations of 20,000 or more.
6. There is some ‘noise’ with county crime-rate data because it is unclear to what extent immigrants are responsible for these crimes. However, as other analysts have commented, the argument does not necessarily depend on whether immigrants are in fact committing crimes, but on whether the public believes they are.
7. The complete table of correlations is available on request from the author.
8. Since only a fraction of the counties in the United States have applied to become 287(g) partners, the results were checked using a rare events logistic regression model. The results obtained were qualitatively similar.
9. Given data limitation issues, we cannot rule out that more recent immigration growth from 2000 onward may in fact matter for county decisions to pursue formal cooperation with ICE on immigration enforcement.