Legislative policies that criminalize immigrants have a "chilling effect" on public program participation among eligible immigrants. However, little is known about the effect of local enforcement actions by Immigration and Customs Enforcement (ICE). In this study, we linked county-level data on the number of detainer requests (or immigration holds) issued by ICE to individual-level data from the 2011, 2016, and 2019 American Community Surveys. We fit adjusted logistic regression models to assess the association between detainer requests and enrollment in Medicaid and the Supplemental Nutrition Assistance Program (SNAP) among those likely eligible for each program in US-born versus immigrant households. A higher volume of detainer requests was associated with lower enrollment in both Medicaid and SNAP, particularly among adults in households with at least one immigrant relative to US-born households. We observed the most pronounced effects in 2011 and 2019.

Introduction: Low-income immigrants who are eligible to participate in the Supplemental Nutrition Assistance Program (SNAP) participate at lower rates compared to non-immigrants. Immigrant households may be more likely to participate in SNAP if they live in areas with policies that integrate them into society and protect them from deportation. Methods: Data on low-income immigrant households came from the 2019 American Community Survey (N = 87,678). The outcome was whether any household member received SNAP in the previous 12 months. Immigrant policy exposures came from two sources: the State Immigration Policy Resource, which includes 18 immigrant criminalizing and integrating policies, and a database that identified 'sanctuary policies' (SP), which we summarized at the county level. Multivariable logistic regression adjusted for person/household-level and area-level confounders. Results: Living in a jurisdiction with a SP was associated with 21% higher odds of enrolling in SNAP compared to living in a jurisdiction without a SP (adjusted odds ratio [aOR] 1.21, 95% CI=1.11,1.31). Relative to the least immigrant friendly states, living in the most immigrant-friendly states was associated with 16% higher odds of SNAP enrollment (aOR=1.16, 95%CI=1.06-1.28). When SP and state-level immigrant friendly policy environment were cross-classified, SNAP participation was 23% and 26% higher for those living in jurisdictions with one- and both- exposures, respectively, relative to those with neither (aOR 1.23; CI 1.12,1.36; aOR 1.26; CI 1.15,1.37). Conclusions: Many at high risk of food insecurity - including immigrants and citizens in households with immigrants - are eligible for SNAP but under-enroll. Policies that welcome and safeguard immigrants could reduce under enrollment.

The purpose of this study was to use participatory systems thinking to develop a dynamic conceptual framework of racial/ethnic and other intersecting disparities (e.g., income) in food access and diet in Philadelphia and to identify policy levers to address these disparities. We conducted three group model building workshops, each consisting of a series of scripted activities. Key artifacts or outputs included qualitative system maps, or causal loop diagrams, identifying the variables, relationships, and feedback loops that drive diet disparities in Philadelphia, Pennsylvania. We used semi-structured methods informed by inductive thematic analysis and network measures to synthesize findings into a single causal loop diagram. There were twenty-nine participants with differing vantages and expertise in Philadelphia’s food system, broadly representing the policy, community, and research domains. In the synthesis model, participants identified 14 reinforcing feedback loops and one balancing feedback loop that drive diet and food access disparities in Philadelphia. The most highly connected variables were upstream factors, including those related to racism (e.g., residential segregation) and community power (e.g., community land control). Consistent with existing frameworks, addressing disparities will require a focus on upstream social determinants. However, existing frameworks should be adapted to emphasize and disrupt the interdependent, reinforcing feedback loops that maintain and exacerbate disparities in fundamental social causes. Our findings suggest that promising policies include those that empower minoritized communities, address socioeconomic inequities, improve community land control, and increase access to affordable, healthy, and culturally meaningful foods.

U.S. Immigration and Customs Enforcement (ICE) issues detainer requests to local law enforcement agencies to hold detainees suspected of being undocumented immigrants until they can be transferred into ICE custody. We examined the association between area-level detainer requests and self-rated health among Latine adults. We linked health data from Latine adults included in the 2017–2020 Behavioral Risk Factor Surveillance System (n = 69,386) to detainer requests per 1,000 non-citizens in core-based statistical areas, (n = 152 across 49 states). We fit logistic regression models of self-rated fair/poor health on detainer requests, adjusted for individual- and area-level confounders. In adjusted analyses, we found that Latine adults living in areas with the highest quartile of requests had 24

The purpose of this study was to investigate the association between county-and state-level immigrant criminalizing and integrating policies and Latino household participation in the largest safety net program against food insecurity in the U.S., the Supplemental Nutrition Assistance Program (SNAP). Our outcome, county-level proportion of SNAP-participating Latino households, and county-level covariates were obtained from the American Community Survey 1-year county files (N = 675 counties) for 13 years (2007-2019). Our exposures were county-level presence of sanctuary policies and a state-level immigrant friendliness score, created based on 19 immigrant criminalizing and integrating state-level policies obtained from the Urban Institute's State Immigration Policies Resource. We classified every county in the sample as 1) sanctuary policy + immigrant friendly state, 2) sanctuary policy + immigrant unfriendly state, 3) no sanctuary policy + immigrant friendly state, and 4) no sanctuary policy + immigrant unfriendly state. Using multivariable generalized linear models that adjusted for poverty levels and other social composition characteristics of counties, we found that county-level SNAP participation among Latino households was 1.1 percentage-point higher in counties with sanctuary policies (B = 1.12, 95%CI = 0.26-1.98), compared to counties with no sanctuary policies, and 1.6 percentage-point higher in counties with sanctuary policies in immigrant friendly states (B = 1.59, 95%CI = 0.33-2.84), compared to counties with no sanctuary policy in immigrant unfriendly states. Local and state immigration policy, even when unrelated to SNAP eligibility, may influence SNAP participation among Latino households. Jurisdictions which lack sanctuary policies or have more criminalizing and less integrating policies should consider adopting targeted outreach strategies to increase SNAP enrollment among Latino households.