The effects of ‘sanctuary city’ policies on the local economy: empirical evidence from US counties

ABSTRACT

This study derives the economic development effects of sanctuary city (SC) policies at the US county level using data from close to 3,000 counties between 2013 and 2017. A county is taken to provide sanctuary to undocumented immigrants if county jails do not accept ICE detainer requests, i.e. do not hold detainees beyond their original release dates. Endogeneity problems and fixed effects in the panel data are fully accounted for by combining coarsened exact matching with the fixed effects instrumental variables method. We can infer that SC policy probably is a significant driver of economic development. The results show that implementing an SC policy for 2.8–3.4 years lowers the unemployment rate by 0.18 percentage points, reduces wages by 1.6%, and increases housing prices by 5.5%, suggesting that these policies have positive effects on the local economy and can contribute to reversing economic declines in some jurisdictions.

Highlights

• The economic development effects of sanctuary city (SC) policies are derived using data from about 3,000 counties between 2013 and 2017.

• Endogeneity problems and fixed effects are addressed by combining coarsened exact matching with the fixed effects instrumental variables method.

• Implementing an SC policy for 2.8–3.4 years lowers the unemployment rate by 0.18 percentage points, reduces wages by 1.6%, and increases housing prices by 5.5%.

• Immigrant-inclusive policies have positive effects on the local economy and can contribute to reversing economic decline.

KEYWORDS:

• Sanctuary cities
• immigration
• economic effects
• panel data
• immigrant-inclusive policy

Acknowledgements

Data on sanctuary status come from the Immigrant Legal Resource Center (ILRC) through a Freedom of Information Act (FOIA) request to ICE.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Our data also show that other types of sanctuary policies have been adopted by far fewer counties, and most of them are included in the group rejecting ICE detainers. In 2017, for example, 171 counties (6%) restricted notification to ICE on release dates or other information, with 167 counties rejecting ICE detainers and only four counties accepting them.

2 Because local labour markets tend to determine wages, the unemployment rate, and housing prices simultaneously, the unemployment rate is suspected of endogeneity as well. The results of an endogeneity test, however, show that only the housing price is endogenous; the unemployment rate is not part of the wage model.

3 For a recent review and discussion of the economic effects of urbanization, see Gross and Ouyang (2020).