Parental Deportation, Safe-Zone Schools, and the Socio-Emotional and Behavioral Health of Children Left Behind

Over 3.6 million immigrants were deported from the U.S. between 2008 and 2018 (U.S. Immigration and Customs Enforcement 2015, ²⁰¹⁸) and approximately 5.1 million children live in households with at least one unauthorized immigrant parent; most of them—over 4 million—are U.S. citizens (Capps, Fix and Zong, 2016). Qualitative research has documented the numerous risks facing these children and those in families fearing deportation, which include family fragmentation, economic hardship, food insecurity, and housing instability (Brabeck and Xu 2010; Chaudry et al. 2010). Specifically, children of parents with a history or at risk of deportation suffer from an increased prevalence of post-traumatic stress disorder (PTSD) symptoms (Rojas-Flores et al. 2017), as well as higher levels of internalizing (e.g., anxiety and depression) and externalizing problems (e.g., aggression) (Allen, Cisneros, and Tellez 2015). Even children who have not experienced parental deportation but are at risk may suffer negative consequences in the form of deteriorated parent-child relationships, negative emotions, and diminished parental financial ability (Brabeck and Xu 2010). These risk factors likely explain the adverse effects that immigration enforcement has been found to have on the educational outcomes of children left behind (Amuedo-Dorantes and Lopez, 2017). As a result, many school districts across the country are implementing sanctuary or ‘safe-zone’ policies to support the well-being and academic progression of children living in migrant households.

Although there is no standard model, safe-zone schools commonly limit their cooperation with immigration authorities, restrict agents’ access to campuses, and provide resources for students and their immigrant families. Safe-zone initiatives may affect children by creating more inclusive and welcoming environments, providing a physical space where students and families are less likely to interact with immigration authorities, and allocating more resources to staff training and student support. In prior work, Amuedo-Dorantes, Bucheli, and Martinez-Donate (2020) document how safe-zone policies aid the academic performance of children in mixed-status households directly impacted or threatened by immigration enforcement. Yet, to date, not much is known about these policies’ impact on youth’s socio-emotional and behavioral outcomes.

We address this gap assessing how school districts’ adoption of safe-zone policies correlates with prosocial behavioral scales, externalizing and internalizing behaviors, and an overall socio-emotional and behavioral difficulties index. We rely on data from two rounds of Between the Lines (BTL), a binational study conducted from 2019–2020 that collected detailed information on the health and well-being of children from mixed-status households who had experienced or were at risk for parental deportation. We combine this information with data on safe-zone policies covering the school districts where children attended school during the study period. Using a quasi-natural experimental approach that exploits the geographic and temporal variation in safe-zone policies over the period the collected information refers to, we document positive and significant correlations between safe-zone policies and children’s prosocial behavior, externalizing, internalizing, and total difficulties scores.

I. Data

The main data source for this study is Between the Lines (BTL)—a longitudinal binational survey completed from 2019–2020 that collected detailed qualitative and quantitative information on U.S.-citizen children of Mexican immigrant parents who had experienced or were at risk for deportation. Among the baseline sample of 100 U.S.-born adolescents with undocumented Mexican immigrant parents, 51 were recruited during the deportation of one of their parents.¹ Community-based organizations and exposed families referred an additional 49 comparable individuals with a Mexican parent at risk of deportation. During the first survey round, individuals and their adult caregivers completed a phone survey with current and retrospective information on children’s health and well-being, education outcomes, and socio-ecological health determinants at the time of parental deportation or upon enrollment for comparison families. Six months later, the same individuals completed a follow-up survey. Our final sample includes 75 children and their adult caregivers who completed both survey rounds and provided information for three points in time: a pre-recruitment snapshot collected retrospectively, a baseline, and a six-month follow-up.

To assess the role of safe-zone policies on adolescents’ socio-emotional and behavioral outcomes, we merged data from the BTL study with school district safe-zone policies to which children in the BTL sample were exposed. We first identified school districts where our survey participants reported attending school in the most recent academic year. Then, we examined each school district’s board of education resolutions to identify whether they had instituted safe-zone policies. We classified a school district as having a safe-zone policy if its governing board of education had approved and adopted resolutions that prohibited ICE activities in school facilities, banned collecting and sharing information with immigration authorities, or allocated resources to staff training initiatives and counseling services. Our sample includes 26 children attending school in a safe-zone school district and 49 children attending school in a district without a safe-zone policy.

In addition, we collect and merge information on county-level interior immigration enforcement policies. To capture the local immigration enforcement climate, we construct a composite index that captures the enactment of various immigration policies at the county level.² The index, which ranges from 0 (no policies in place) to 4 (all policies), provides a way to proxy for an overall harsher anti-immigrant climate generated by the enactment of multiple interconnected policies. We match the immigration enforcement index to our individual respondents using the county where their schools are located.

II. Methodology

To learn about the impact that safe-zone policies might have had on socio-emotional and behavioral outcomes of children of mixed-status families, we estimate the following model:

$$Y_{idm}=\alpha +{\beta }_1{SZ}_{dm}+{\beta }_2{IE}_{cm}+{\gamma }^{\mathrm{\text{'}}}{X}_{im}+{\delta }_s+{\delta }_m+{\epsilon }_{idm}$$ (1)

where the dependent variable ($Y_{idm}$) captures socio-emotional and behavioral outcomes for child i attending school in district d when interviewed in survey round m. The outcomes include four scores from the Strengths and Difficulties Questionnaire (SDQ): i) prosocial behavior, ii) externalizing, iii) internalizing, and iv) total difficulties.³ Table 1 details the BTL survey items and subscales used to estimate each score.

Table 1:

Items used to Construct the Socio-emotional and Behavioral Indicators

Scale/Score	Subscales/Items
Prosocial Scale	1. Is considerate of other people’s feelings  2. Shares readily with other youth, for example, books, games, food  3. Is helpful if someone is hurt, upset or feeling ill  4. Is kind to younger children  5. Offers to help others (parents, teachers, children)
Externalizing	Conduct problems  6. Loses temper  7. Is well behaved, does what adults request  8. Fights with other youth or bullies them  9. Lies or cheats  10. Steals from home, school or elsewhere Hyperactivity  11. Is restless, overactive, cannot stay still for long  12. Is fidgeting or squirming  13. Is easily distracted, concentration wanders  14. Thinks things out before acting  15. Has good attention span, sees work through to the end
Internalizing	Emotional Problems  16. Complains of headaches, stomachaches or sickness  17. Has many worries or seems worried  18. Is unhappy, depressed or tearful  19. Is nervous in new situations, easily loses confidence  20. Has many fears, easily scared Peer Problems  21. Prefers to be alone than with other youth  22. Has at least one good friend  23. Is liked by other youth  24. Is picked on or bullied by other youth  25. Gets along better with adults than with other youth
Total Difficulties	Average of the externalizing and internalizing scores

Note: The four scales and 25 items come from the Strengths and Difficulties Questionnaire (SDQ) used to screen child mental health and behavior problems (Goodman, Lamping, and Ploubidis 2010). Caregivers in our sample provide information for each item using a three-point scale, where 0 means “never” or “almost never,” 1 indicates “once in a while” or “sometimes,” and 2 means “regularly.” Items 7, 14, 15, 22 and 23 are reverse-coded to preserve consistency across the items. Each scale is calculated as the average score of all items within the scale.

Because the impact of school districts’ safe zones on children’s schooling outcomes might significantly differ across counties, we not only account for the child’s exposure to a safe zone (${SZ}_{dm}$) based on their school district, but also their household’s exposure to interior immigration enforcement in county c (${IE}_{cm}$). In addition, we account for basic child traits included in vector $X_{im}$, such as the age, gender, high school enrollment, presence of older siblings, number of cities the child has lived in since age 6 to capture school changes that can impact academic achievement, and an indicator of whether the child has suffered the separation from a parent due to deportation. The vector $X_{im}$ also includes information on the child’s caregiver –namely, whether they completed high school and if they are employed. Finally, the model includes rounds fixed effects (${\delta }_m$) to address any variations in immigration policies in each interview round and geographic fixed effects to control for time-invariant characteristics. We cannot include school district or county fixed effects due to our small sample size. Instead, we include state fixed effects (${\delta }_s$).

We estimate equation (1) as a random-effects model given that there is only a six-month gap between the survey rounds and most individual and household-level covariates remain unchanged throughout the study period. We first experiment with a simple model specification that only includes information on whether a safe-zone policy is in place, along with basic geographic and temporal fixed effects. Subsequently, we repeat the estimation including information on the extent of interior immigration enforcement in place and the demographic and household traits in vector X. Because of the small sample size, standard errors are clustered at the state level to ensure multiple observations per cluster.

III. Safe-Zones and Children’s Socio-Emotional and Behavioral Outcomes

Panel A in Table 2 displays the results from estimating equation (1) using two model specifications that differ in their controls to address concerns regarding the impact of potentially endogenous regressors. For brevity, we focus on the estimated impact of safe-zone policies. In the absence of any immigration enforcement and youth-related controls, safe-zone policies appear to be significantly correlated to positive socio-emotional and behavioral outcomes, raising the prosocial scale by 7 percent, lowering the externalizing and internalizing scores by 10 and 15%, respectively, and generally reducing the score of overall difficulties by 13%. Some of these impacts vary as we account for additional controls. Specifically, the correlation between school districts’ safe-zone policies and children’s externalizing scores strengthens (now dropping by 21%), whereas the link between safe-zone policies and children’s internalizing scores weakens and becomes non-statistically different from zero. In contrast, the correlation between safe-zone policies and children’s prosocial and overall difficulty scores proves resilient to the inclusion of controls for the immigration enforcement climate and children’s traits, underscoring the credible relevance of these school districts’ policies in improving children’s well-being.

Table 2:

The Impact of Exposure to Safe-Zone Policies on Children’s Behavioral Outcomes

Panel A: Impact by Current Exposure
	Prosocial Scale		Externalizing Score		Internalizing Score		Total Difficulties Score
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
School district SZ policy	0.193***	0.187***	−0.064**	−0.140***	−0.138***	−0.058	−0.099***	−0.098**
	(0.034)	(0.065)	(0.028)	(0.030)	(0.024)	(0.074)	(0.002)	(0.048)
Individual-level controls	N	Y	N	Y	N	Y	N	Y
Immigration enforcement	N	Y	N	Y	N	Y	N	Y
Survey round FE	Y	Y	Y	Y	Y	Y	Y	Y
State FE	Y	Y	Y	Y	Y	Y	Y	Y
Dependent variable mean	2.65	2.65	0.66	0.66	0.90	0.90	0.78	0.78
Observations	225	225	225	225	225	225	225	225
R-squared	0.112	0.173	0.108	0.210	0.115	0.199	0.086	0.167
Panel B: Impact by Age at First Exposure and by Length of Exposure
	Age at First Exposure				Length of Exposure
	Prosocial Scale	Externalizing Score	Internalizing Score	Total Diff. Score	Prosocial Scale	Externalizing Score	Internalizing Score	Total Diff. Score
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
SZ exposure before age 12	0.336***	−0.251***	−0.048	−0.151***	-	-	-	-
	(0.069)	(0.028)	(0.083)	(0.054)
SZ exposure at age 12+	0.078	−0.059	−0.066	−0.059	-	-	-	-
	(0.057)	(0.037)	(0.073)	(0.052)
Years exposed to SZ	-	-	-	-	0.079***	−0.041***	−0.027	−0.034**
					(0.016)	(0.012)	(0.019)	(0.015)
Individual-level controls	Y	Y	Y	Y	Y	Y	Y	Y
Immigration Enforcement	Y	Y	Y	Y	Y	Y	Y	Y
Survey round FE	Y	Y	Y	Y	Y	Y	Y	Y
State FE	Y	Y	Y	Y	Y	Y	Y	Y
Dependent variable mean	2.65	0.66	0.90	0.78	2.65	0.66	0.90	0.78
Observations	275	275	275	275	275	275	275	275
R-squared	0.187	0.223	0.199	0.171	0.183	0.209	0.201	0.169

Note: The most complete specifications in Panel A and all specifications in Panel B include a constant term and the following controls: age, gender, high school status, number of older siblings, cities of residence since age 6, whether the caregiver completed high school, the caregiver’s employment status, and whether a child’s parent has been deported. SZ = safe zone. Standard errors clustered at state level in parentheses.

^***p<0.01,

^**p<0.05,

^*p<0.1.

From a policy perspective, it is critical to learn about the role of safe-zone policies in safeguarding youth’s mental health based on their age at first exposure and the length of time they benefited from the policy. According to the estimates in Panel B of Table 2, the beneficial impact of safe zones appears to be concentrated among youth who were exposed to the policy pre-adolescence. In addition, duration of policy exposure is key, with each additional year of policy exposure being associated with prosocial scores that are 3% higher, externalizing scores that are 6% lower, and an overall total difficulty score that is 4% lower.

IV. Summary and Conclusions

We examine how safe-zone policies adopted by school districts appear to impact prosocial behavioral scales, externalizing and internalizing behaviors, and overall socio-emotional and behavioral difficulties of children in mixed-status households threatened by or impacted by parental deportation. The results reveal non-negligible gains to children’s socio-emotional and behavioral health that are worth considering by policymakers and educators to incentivize their adoption to partially offset the damaging impacts of intensified immigration enforcement. Despite their preliminary nature, given the small sample of youth we work with, as well as the possibility for the found correlations to not be causal, our findings warrant further consideration in light of the low cost of these policies and their positive impacts on children’s health.