Association of Maternal Eligibility for the Deferred Action for Childhood Arrivals Program With Citizen Children’s Participation in the Women, Infants, and Children Program

Abstract

This study examines the association of maternal eligibility for Deferred Action for Childhood Arrivals (DACA) immigration program with participation of citizen children in the Women, Infants, and Children (WIC) health care and supplemental nutrition program.

Nearly 7% of children living in the United States, the vast majority of whom are US citizens, have at least 1 undocumented immigrant parent.¹ These children face several disadvantages, culminating in reduced lifetime socioeconomic mobility and reduced well-being.¹ One mechanism underlying these adverse consequences could be failure to receive critical public benefits despite meeting eligibility criteria because undocumented parents may be less likely to apply for these services on their child’s behalf if they fear being discovered by immigration authorities.^2,3

Policies that bring undocumented parents “out of the shadows,” such as the 2012 Deferred Action for Childhood Arrivals (DACA) program, may have positive spillover effects for their children by improving uptake of public benefits. We examined the association of parental DACA eligibility with children’s participation in the Women, Infants, and Children (WIC) program, a benefit that has been shown to improve child health and socioeconomic outcomes.⁴

Methods

We used data from the 2010-2015 National Health Interview Surveys (NHIS). Our sample consisted of US citizen children who were 5 years of age or younger (reflecting WIC age eligibility criteria) and whose mothers were Hispanic and not US citizens. The latter criterion follows earlier work that noted that a large percentage (>60%) of self-reported noncitizens are undocumented.⁵ In addition, we further restricted the sample to children whose mothers had lived in the United States for at least 5 years. We also restricted our sample to children of mothers who were 19 years of age or older and had received at least a high school diploma or General Educational Development certificate, to hold fixed 2 key DACA eligibility criteria.⁵ This study was acknowledged as exempt, non-human subjects research by the Johns Hopkins School of Medicine Institutional Review Board.

Our main outcome was whether the child was enrolled in WIC in the previous calendar year (the period queried by the NHIS). Our main exposure—whether the mother met DACA eligibility criteria—was defined on the basis of the mother’s age at immigration (≤16 years) and age at DACA implementation (≤31 years at policy implementation).⁵

We estimated a difference-in-difference model that compared changes in WIC enrollment among children whose mothers met the 2 DACA age eligibility criteria before (survey years, 2010-2012) vs after (survey years, 2014-2015) policy introduction with changes among children whose mothers did not meet these criteria. (Survey year 2013 was excluded because it corresponded to WIC participation in the year DACA was implemented). We adjusted for sociodemographic characteristics of both the child and mother.⁶ All descriptive statistics and analyses were performed using NHIS sample weights.

Results

Our final sample consisted of 1911 children 5 years or younger, of whom 33.8% had a mother who likely met DACA eligibility criteria (Table 1). Overall, 43.1% children participated in the WIC program during the study period.

Table 1. Selected Descriptive Statistics of Citizen Children and Noncitizen Hispanic Mothers by Likely DACA Eligibility^a.

Variable	Likely DACA-Ineligible Mothers and Childrenb	Likely DACA-Eligible Mothers and Childrenc
Sample, No. (%)	66.2	33.8
Child
Age, mean (SD), y	2.65 (1.96)	2.37 (1.95)
Sex, %
Male	52.5	51.9
Female	47.5	48.1
Mother
Age, mean (SD), y	34.0 (5.90)	25.7 (4.08)
Educational level, %
High schoold	59.4	70.1
Some college	40.6	29.9
Comfortable with English language, %e	46.3	59.8
Married, %	76.9	61.9
No. of years in US, %
5-9	41.3	11.1
10-14	33.5	27.9
≥15	25.2	61

Abbreviation: DACA, Deferred Action for Childhood Arrivals.

^aAll descriptive statistics are weighted by National Health Interview Surveys sample weights. The primary analytic sample consists of children age 5 years and younger who are US citizens and who have noncitizen Hispanic mothers with a high school diploma or GED certificate who have lived in the United States for at least 5 years. Mothers likely eligible for DACA are defined as those who, as of June 2012, were 31 years or younger and had lived in the United States for at least 5 years.

^bConstitutes 1258 unweighted or 2 777 930 weighted individuals.

^cConstitutes 653 unweighted or 1 418 199 weighted individuals.

^dIncludes receipt of general educational development (GED) certificate.

^eConsidered “comfortable” if the interview was administered in English only.

A mother’s DACA eligibility was associated with a 12.3% (95% CI, 0.7%-23.9%) higher likelihood that her child participated in WIC (Table 2). Among children whose families met broad WIC income eligibility criteria (family income <185% of the federal poverty level or Medicaid receipt), we found similar estimates (β coefficient, 13.5%; 95% CI, 0.81%-26.3%). We did not find an association among children with noncitizen mothers who likely would not have met DACA eligibility criteria on the basis of educational attainment (falsification test).⁵

Table 2. Association of Maternal DACA Eligibility With Children’s Participation in WIC^a.

Variable	Children Aged ≤5 y (n = 1911)	Children Aged ≤5 y Likely Meeting WIC Income Eligibility Criteria (n = 1572)b	Falsification Test (n = 2986)c
Difference-in-difference estimate, % (95% CI)	12.3 (0.7 to 23.9)	13.5 (0.81 to 26.3)	3.7 (−7.2 to 14.6)
P value	.04	.04	.50

Abbreviations: DACA, Deferred Action for Childhood Arrivals; WIC, Women, Infants, and Children program.

^aLinear probability (ordinary least squares) difference-in-difference estimates. Models were weighted to reflect National Health Interview Surveys sampling weights. Confidence intervals were corrected for clustering at the household level. All models were adjusted for child and mother characteristics in Table 1. Because WIC participation was defined in the previous calendar year and DACA was implemented in June 2012, we chose to exclude 2013 data and define the prepolicy period as 2010 to 2012 and the postpolicy period as 2014 to 2015.

^bWIC eligibility was defined as family income <185% of the federal poverty level or child receiving Medicaid benefits.

^cChildren with mothers without a high school diploma or General Educational Development certificate.

Discussion

Maternal eligibility for the DACA program was associated with increased participation in WIC by their citizen children. These results highlight the potential for multigenerational spillover effects of immigration policy.

A limitation of our analysis is that we were unable to explicitly identify undocumented parents or make strict determinations of DACA eligibility. However, these limitations are true of all nationally representative data sets and likely will result in underestimates of program effects.⁵ A second limitation is that data were self-reported. Third, even with the quasi-experimental research design, it is possible that the findings were biased by unmeasured confounders.

Our findings—along with growing evidence of the direct health consequences of DACA for both beneficiaries and their children^5,7—should be considered in ongoing debates around immigration policy.