Sanctuary Interrupted: Analyzing Federal Funding, Sanctuary Policies, and Immigration Detention from 2010 to 2018

Background/Purpose:

While calls to defund sanctuary jurisdictions were popularized in 2016, politicians—led largely by the Trump Administration—renewed their demands to withhold federal grants from sanctuary jurisdictions that limit local law enforcement agencies’ (LEAs) collaboration with Immigration and Customs Enforcement (ICE) for immigration detention. Politicians in sanctuary jurisdictions have challenged these attempts to withhold federal LEA grants, arguing the grants are unrelated to federal-local immigration enforcement collaboration. However, cross-sectional research has suggested that federal LEA grants—such as the State Criminal Alien Assistance Program (SCAAP)—incentivize and subsidize LEAs’ engagement in racialized policing practices that target Latinx immigrants, even in sanctuary jurisdictions. Thus, focusing on 2010 to 2018— a period of heightened political discourse surrounding sanctuary policies, federal grants, and immigration detention—I use mixed-effects negative binomial regression models (NBRM) to test whether SCAAP predicts the number of migrants held for ICE and whether sanctuary policies moderate this relationship.

Methods:

This county-level panel dataset (N=21,797) draws on numerous secondary data sources, including Vera Institute of Justice, ICE, and Bureau of Justice Statistics. The dependent variable is count of migrants held in county jails for ICE. Independent variables are county receipt of SCAAP the previous year and presence of an active sanctuary policy. Covariates include urbanicity, percent-Latinx residents, active 287(g) agreement, excess jail bedspace, jail capacity, unemployment, political majority (presidential), time, and presence of a detention agreement. Overdispersion and intra-class correlation tests indicated that mixed-effects NBRM (with state-level random effects) was appropriate.

Results:

Overall, 23.68% (n=246) of counties with sanctuary policies and 40.55% (n=2,480) of counties that received SCAAP the previous year held at least one immigrant for ICE. Receiving SCAAP increased the expected count of immigrants held for ICE by 118.15% (B=.78, SE=.06, p<.001) while county sanctuary policies reduced the expected count by 69.58% (B=-1.19, SE=.13, p<.001). However, sanctuary policies did not significantly moderate the relationship between SCAAP and the count of immigrants held for ICE (B=.16, SE=.14, p=.247). Receiving SCAAP increased the expected count of immigrants held regardless of whether a county adopted a sanctuary policy. In fact, sanctuary counties that received SCAAP held nearly double the number of immigrants than those that did not receive SCAAP (ME=2.02, SE=.28).

Conclusions/Implications:

Recent Trump Administration efforts to intensify immigration arrests and removals illustrate the urgent need for research uncovering human cost of federal LEA grants. My findings support emerging research suggesting that federal LEA grants incentivize federal-local immigration enforcement collaboration and that traditional sanctuary measures offer immigrants minimal protection from this violent system. Although sanctuary policies decreased the number of immigrants LEAs held for ICE, they failed to disrupt SCAAP’s effects or preclude LEAs’ participation in immigration detention, likely because sanctuary policies contain carve-outs for immigrants with criminal convictions. Social work research, advocacy, and education must destabilize hegemonic political discourses—and the structures supporting them—that legitimate criminalized immigrants’ detention and deportation. Advancing justice and safety for immigrants requires advocacy efforts that center their constitutional and human rights and target the funding structures subsidizing LEAs’ engagement in racialized policing practices.