Introduction

Food insecurity, limited access to adequate food, is associated with poor nutrition and numerous health, developmental, and psychosocial problems [1, 2]. In 2017, 11.8% of US households and 15.7% of those with children experienced food insecurity [3]. Children’s Supplemental Nutrition Assistance Program (SNAP) benefits decrease the impact of food insecurity on child health [4].

Unauthorized immigrants are ineligible for SNAP. However, US-born children of unauthorized immigrants are eligible if the household meets eligibility criteria [5]. When this study was conducted, SNAP receipt was not considered during immigration or naturalization processes, including in the public charge determination of likelihood of future dependence on public resources for income or long-term institutionalization [6]. A rule that was due to take effect October 15, 2019, Inadmissibility on Public Charge Grounds, includes SNAP and other non-cash benefits in the determination [7]. The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), a program for pregnant, postpartum, or breastfeeding women and children up to age 5 who meet income-eligibility, neither requires citizenship nor immigration authorization and is not included in the new rule [8].

Welfare Reform had an enduring effect on SNAP enrollment among eligible US-born children of non-citizens. Enrollment fell from 39% in 1994 to 19% in 2000, which, along with reductions in benefit amounts, resulted in high food insecurity levels [9]. In 2012, 42.4% and 51.4% of eligible children of non-citizens and citizens, respectively, received SNAP [10].

This study examined community-identified barriers to nutrition benefits enrollment among children of Latin American-born immigrants and Latin American-born pregnant, postpartum, and breast-feeding women in the face of increased anti-immigrant rhetoric.

Methods

This institutional review board-exempt study, conducted from February to September 2016 with Spanish-speaking adults at a NYC-based organization serving large numbers of Latinx clients, utilized an orally administered Spanish-language needs assessment survey. Informed consent was obtained orally. The survey explored household demographics, participants’ knowledge and experiences of SNAP and WIC and enrollment barriers, including reasons those appearing eligible did not enroll or disenrolled, and SNAP/WIC rumors. Afterwards, a statement was read explaining rumor inaccuracies, and offering information and referrals to safety net services.

Analyses

Sample frequencies and percentages of sociodemographic descriptors were calculated. SNAP and WIC eligibility were estimated based on state program criteria. Frequencies of non-participation reasons among eligible non-participants were calculated. For those in eligible households, descriptive statistics were stratified by benefit participation and tested for differences. Descriptive statistics and analyses were calculated using SPSS v22 and R v3.2.2.

Two coders independently performed thematic analyses of string variable data consisting of non-participation reasons and rumor content. Differences were resolved by consensus.

To understand how 6 rumors of repercussions affected enrollment, we regression modelled to estimate the enrollment probability by rumor heard, controlling for age, gender, income, and education level.

Results

Sample Enrollment and Sociodemographic Descriptors

Of 128 people invited to participate, 28 declined for reasons including time constraints and disinterest, 13 terminated when called for their service appointment, and 87 completed the interview. Non-responses were considered randomly missing.

Table 1 shows sample eligibility, enrollment, sociodemographic variables, and enrollment predictors among eligible children and women. Based on assessed SNAP eligibility, 51 (58.6%) had at least 1 eligible child. Of these 51 households, 25 (49%) had no participating child. Non-participation reasons included lack of necessary financial documents [7 (28% of non-participants)]; not knowing about SNAP [4 (16%)], how to apply [4 (16%)], whether it would help [2 (8%)], and where to apply [1 (4%)]; denial of application [2 (8%)]; or another reason [20(80.0%)] (not shown).

Table 1 SNAP and WIC household eligibility, enrollment, sociodemographic variables, predictors of enrollment and odds ratios among eligible children of Latin American-born parents and pregnant or breastfeeding Latin American-women, community servicing organization in NYC, 2016 (N = 100)
Full size table

Based on assessed WIC eligibility, twelve women and children in 33 households appeared eligible. Among these, 4 (33.3%) women and children in 7 (21.2%) separate households were not enrolled. Two (18.1%) participants did not know where to apply, 2 (18.1%) were unaware of WIC, and 7 (63.6%) provided another non-participation reason (not shown).

Thematic analysis of other reasons provided for SNAP/WIC non-enrollment despite eligibility identified themes of fear of repercussions, misinformation/lack of information, logistical barriers, lack of perceived need, not enough information to determine the reason, and participant planned to apply (not shown).

Seventy-one (81.6%) of the total participants had heard at least 1 of 6 rumors of SNAP/WIC usage repercussions (not shown): 36 (50.7%) had heard children could be removed; 48 (67.6%) heard children could later be conscripted; 47 (66.2%) heard the benefit would have to be repaid; 19 (26.8%) heard unauthorized relatives could be reported; 13 (18.3%) heard relatives could be deported; and 28 (39.4%) heard SNAP/WIC receipt would preclude relatives’ status adjustment. Ten (14.1%) had heard of other repercussions, e.g., college aid ineligibility.

Regression Models

Comparing the odds ratios for nutrition program enrollment among participants who heard misinformation and those who did not, the only significant association was with “If someone uses food stamps or WIC, the people in the family who don’t have documents could be reported to the government.” Among these households, it was 85% less likely that eligible persons were enrolled (OR 0.15, CI 0.03, 0.94). Three other rumors, regarding conscription, deportation, and relatives’ immigration applications, were negatively associated with enrollment. Those who heard the military service rumor were 75% less likely to enroll in nutrition benefits. Our model of 6 rumors estimated an average decrease of 27% in enrollment probability per rumor heard (P = .10).

Discussion

This study elucidated barriers to SNAP/WIC enrollment in Latinx immigrant households, with potential consequences for child health and development, during the final year of a presidential campaign featuring deprecatory rhetoric about immigrants and proposals to deport all unauthorized immigrants. SNAP participation has consistently been disproportionately low among eligible children of immigrants, a disparity that worsened under Welfare Reform and the accompanying heightened scrutiny of immigrant communities [9,10,11]. Participation among eligible persons will certainly worsen with expansion of public charge criteria to include SNAP, announced in August, 2019 [7]. A larger study, examining the use of permitted nutrition and other services and interventions to mitigate disproportionately low participation, should take place.

Nearly a third of the sample (32.7%), whose mean household income was 83% of the poverty threshold, had not enrolled eligible persons in SNAP/WIC. Our findings confirmed several barriers previously described [12], but additional concerns were identified, including that SNAP/WIC receipt precludes college aid, US-born children could lose their citizenship, US-born children could become vulnerable to removal, and that migrants could be extradited to repay benefits.

Our small sample size forced the use of a regression model with limited confounders to maintain power to examine associations with enrollment, potentially biasing estimates. Larger studies could account for additional confounders, such as length of stay, age of oldest child, and acculturation. Longitudinal data could test our hypothesis that misinformation mediates the relationship of income with SNAP/WIC enrollment. Including dietary intake and food insecurity measures could quantify the impact of SNAP/WIC at various household income levels.