Abstract
The population of young adults who are undocumented immigrants is growing in the United States, and little is known about the prevalence and experiences of undocumented transition-age youth in foster care. This policy brief uses administrative child protective case records to provide a foundational measure of youth immigration status documentation in California leading up to the enactment of Assembly Bill 829, which aimed to enhance coordination between child welfare and legal services through timely attorney notification. In addition, we examine relationships between immigration status and Transitional Independent Living Plan (a semi-annual document describing youths’ goals and the resources needed to achieve them; TILP) outcomes. Missingness in immigration status was common and increased over time. U.S. citizens and undocumented youth were more likely to have TILPs than youth who were legal residents or had a missing immigration status. Findings highlight opportunities to both invest in services for undocumented young people exiting care and explore barriers to immigration status documentation in administrative records.
Introduction
An estimated 1.57 million young people ages 16 to 24 are living in the United States as undocumented immigrants (Migration Policy Institute, 2019). Despite well-documented material, educational, and service needs within the broader undocumented immigrant population (Acevedo-Garcia et al., 2021; Ayon, 2014; Kreisberg & Hsin, 2021), little is known about the prevalence and experiences of undocumented children and young adults who come to the attention of the public child welfare system. This gap is largely driven by the logistic and ethical challenges of documenting youth immigration status. Studies examining nativity—where youth are born, but not their current legal status—suggest that between 8% and 10% of children and youth below age 18 with child welfare experience are immigrants, which is double the rate of foreign-born children in the general U.S. population (Dettlaff et al., 2009; Finno-Velasquez, 2013; Kids Count, n.d.). In 2011, the Shattered Families National Study estimated that at least 5,100 children and youth below age 18 were in foster care due to parental detention or deportation (Wessler, 2011). The Shattered Families Study also projected this number would grow to more than 10,000 by 2016, although a follow-up study that corroborates these estimates has yet to be published.
Research examining the needs and experiences of undocumented transition-age youth (youth ages 16 to 24 with foster care experience or “TAY”) is especially scant. Preliminary evidence suggests immigrant youth who have contact with the child welfare system may experience distinct barriers to permanency (e.g., parental deportation, limited kin networks in the United States) that increase their likelihood of aging out of care (Finno-Velasquez & Dettlaff, 2018). As they enter early adulthood, undocumented young adults may struggle to acquire the employment or postsecondary training opportunities necessary to achieve self-sufficiency. Consequently, there is a pressing need for additional evidence that can help caseworkers and other service providers adapt independent living plans to the unique needs of undocumented TAY.
Recognizing this need, California passed Assembly Bill 829 (AB 829) in 2022, which requires child welfare agencies to notify children’s and young adults’ attorneys of their legal status. In subsequent guidance detailing the implementation efforts for AB 829, the California Department of Social Services (CDSS) encouraged caseworkers to document youth legal status in administrative records to track notification efforts and ensure subsequent services reflected young people’s immigration status. In a state where almost half of all children come from immigrant families (Cuellar Mejia et al., 2024), AB 829 has the potential to build a valuable data source that can inform research and service development for undocumented young people transitioning from foster care. To inform these efforts, the current policy brief uses California’s child protective case records to provide a baseline measurement of TAY’s legal status documentation in administrative records before the implementation of AB 829. In doing so, this brief assesses the utility of administrative records to learn about and develop policy recommendations for undocumented young people transitioning out of foster care. In addition, we examine how legal status is associated with various outcomes related to youth Transitional Independent Living Plans (TILPs), documents co-developed by youth and their caseworkers, outlining young people’s goals and service needs for achieving self-sufficiency in early adulthood. Per state policy, all young people in care on or after their 16th birthday should have at least one TILP. Following their 16th birthday, TILPs are updated every 6 months until young people leave care to ensure services are reflective of their evolving needs and goals.
Method
Sample
The current brief relied on data from statewide administrative child protective case records stored in California’s Child Welfare Services/Case Management System (CWS/CMS). These data were accessed through a long-standing research partnership between the CDSS and the California Child Welfare Indicators Project at UC Berkeley. The use of these records for research was approved by the California Committee for the Protection of Human Subjects (Protocol 12-10-0800) and the Office for Protection of Human Subjects at UC Berkeley (Protocol ID: 2010-01-592). Youth in the current sample were in child welfare-supervised foster care for at least 12 months on or after their 16th birthday between 2015 and 2022. To capture complete records of lifetime foster care involvement, we restricted analyses to youth who turned 21 by December 31, 2022 (following California’s implementation of extended foster care that allows youth to remain in care until age 21). Youth predominantly absent from placement or in legal guardianships were excluded from analyses. The analytic sample comprised 19,610 youth.
Measures
Dependent Variables
Analyses examined factors associated with three related measures of TILP completion. The first measure indicated whether youth had completed at least one TILP during their time in care (dichotomous yes/no). The second measure summed the number of TILPs that youth had before their final exit from foster care. Our final measure quantified the discrepancy between the number of TILPs youth had documented in their case file and the number they should have had per state regulations (at least one by their 16th birthday and one for every additional 6 months in care following their 16th birthday). To calculate the number of required TILPs, we summed the number of days youth spent in care after their 16th birthday and divided by 180 (approximately 6 months). We added 1 to represent the TILP that should be completed by their 16th birthday. Then, we subtracted the total number of completed TILPs from this number to obtain the final TILP discrepancy score.
Independent Variable
Youth immigration status was derived from three variables in child welfare records: immigration status, special immigrant juvenile status (SIJS), and country of origin. Immigration status was recoded into four categories: U.S. citizen, legal resident (also known as lawful permanent residents or “green card” holders), undocumented, and missing. Youth who were missing their immigration status but were born in the United States or had received their legal residency card through a SIJS petition were recoded as U.S. citizens and legal residents, respectively. Even when pooled together, few youth had another immigration status (e.g., unaccompanied youth, asylees, refugees), particularly when spread across the 8-year observation period. Consequently, these youth were excluded from analyses to ensure compliance with data deidentification requirements.
Covariates
We used youth- and county-level variables as covariates in our analyses. Youth’s assigned birth sex (female or male), race and ethnicity (Latine, missing, Black, white, Asian, and Native American), and primary language (English, Spanish, or other) were derived from demographic fields. Age at first entry (0–5, 6–11, or 12–17), age at final exit (before or after their 18th birthday), and time in care after their 16th birthday (more or less than 2 years) were derived from birth and out-of-home placement dates. The final exit year was derived from out-of-home placement end dates for the youth’s final placement. Predominant placement before age 18 was derived by calculating the percentage of placement days spent as a minor in the following settings: foster family home, congregate care, foster family agency, relative, mixed or other, and transitional housing program. Finally, we measured the urbanicity (i.e., population size) of the last county to supervise the youth’s out-of-home placements using four categories: rural, urban, large urban, and Los Angeles County only.
Data Analysis
Descriptive statistics examined the sociodemographic and foster care characteristics of youth in our sample. Counties with high (≥20%) and low (<20%) missingness in immigration status were used to document differences (chi-square and t-tests) across TILP completion variables. The percentage of any TILP completion and immigration status missingness over time was examined using a chi-square test of trends. Bivariate associations between immigration status and TILP outcomes were tested using chi-square and one-way analyses of variance with the Bonferroni correction. Finally, multivariate logistic and negative binomial regression were used to examine the association between immigration status and TILP completion outcomes (any TILP, number of TILPs, and TILP discrepancy scores), adjusting for covariates. These regression models were fit using the entire sample, youth supervised by low-missingness counties, and youth supervised by high-missingness counties. County urbanicity was omitted from models of low and high missingness due to collinearity assumption violations. Given the elevated chance of Type I errors due to the large sample size and number of comparisons conducted, we interpreted statistical significance at p < .001. We also include three measures of effect size (Cramer’s V, Cohen’s d, and χ2 for chi-square, t-tests, and one-way analyses of variance, respectively) to contextualize the practical significance of our findings. Given the high degree of missingness in youth immigration status (20.6%), we conducted three sensitivity analyses to determine whether missingness may have biased our results. In the first analysis, young people with a missing immigration status were recoded as U.S. citizens, whereas in the second, young people were recoded as undocumented. The final sensitivity analysis omitted youth who were missing their immigration status. Each version of the recoded immigration status variable was used to predict TILP completion outcomes in multivariate models. All analyses were conducted using Stata MP, 17th edition.
Results
The sociodemographic characteristics of the study sample are reported in Supplemental Appendix A (Appendix A: Table 1). More than half of the sample was assigned female at birth, and 46.2% were Latine. Although most youth spoke English as their primary language (88.2%), the proportion of youth who spoke Spanish as their primary language was significantly higher in counties with higher rates of immigration status missingness, χ2(2, N=19,610) = 160.4, p < .001. Three quarters of the sample were U.S. citizens. There was a significant difference in the proportion of youth who were U.S. citizens between low- and high-missingness counties (89.4% and 63.4%, respectively; χ2(3, N=19,610) = 2,002.2, p < .001). Across the sample, one fifth had missing immigration status, with this rate increasing to nearly one third in high-missingness counties.
Chi-square tests of trends showed the proportion of youth with at least one TILP increased significantly between 2015 and 2022, χ2(1, N=19,610) = 167.7, p < .001, V = .10; Supplemental Appendix A: Table 2. Similarly, the proportion of youth with a missing immigration status also increased during the observation period, from 19.8% to 25.1%, χ2(1, N=19,610) = 37.9, p < .001, V = .05. All TILP completion outcomes varied significantly by immigration status (Supplemental Appendix A: Table 3). A greater proportion of U.S. citizens and undocumented youth had at least one TILP compared with youth who were legal residents or missing their immigration status, χ2(3, N=19,610) = 166.0, p < .001, V = .09. Youth with a missing immigration status also had fewer TILPs than youth who were U.S. citizens, legal residents, or undocumented, F(3, 19,606) = 107.5, p < .001, h2 = .02. Youth with missing immigration status had higher TILP discrepancies than U.S. citizens and undocumented youth (p < .001), n2 = .02. Youth who were legal residents also had significantly higher discrepancies than U.S. citizens and undocumented youth, F(3, 19,606) = 75.6, p < .001, h2 = .02.
Multivariate models using the entire sample found that youth with missing immigration status had lower odds of having any TILPs (odds ratio [OR] = 0.57, p < .001), fewer TILPs (incidence rate ratio [IRR] = 0.65, p < .001), and higher TILP discrepancies (IRR = 1.2, p < .001; Supplemental Appendix A: Table 4) than youth who were U.S. citizens (referent). This difference in the number of TILPs and TILP discrepancies remained significant in counties with low-, but not high-, missingness (Supplemental Appendix A: Tables 5 and 6). Legal residents in low-missingness counties had significantly more TILPs (IRR = 1.36. p < .001) and lower TILP discrepancies (IRR = 0.78. p < .001) than U.S. citizens. Among the entire sample, youth who spoke Spanish as their primary language had lower odds of having at least one TILP (OR = 0.80, p < .001). This pattern was replicated in high- but not low-missingness counties.
Sensitivity analyses showed that immigration status missingness considerably biased results. In the model where youth with missing immigration status were recoded as U.S. citizens, legal residents had significantly greater odds of having a TILP (OR = 1.40, p < .001) and more TILPs (IRR = 1.30, p < .001) compared with U.S. citizens (Supplemental Appendix A: Table 7). Nevertheless, when youth with missing immigration status were recoded as undocumented, models show undocumented youth had significantly lower odds of having a TILP (OR = 0.59, p < .001), fewer TILPs (IRR = 0.68, p < .001), and higher TILP discrepancy (IRR = 1.23, p < .001; Supplemental Appendix A: Table 8). Our final sensitivity analysis showed that after omitting youth with a missing immigration status from our sample, there were no significant differences in TILP completion outcomes among U.S. citizens, legal residents, or youth who were undocumented. Legal residents were marginally more likely to have more TILPs overall, although this was not significant at our adjusted alpha level (p = .001; Supplemental Appendix A: Table 9).
Discussion
Our policy brief provides a foundational understanding of California’s documentation of TAY immigration status preceding the implementation of AB 829 and its relationships with TILP outcomes. We found that missingness in immigration status was common, varied by county, and increased in the years leading to the implementation of AB 829. At the same time, the proportion of young people with at least one completed TILP increased significantly over time. Unadjusted analyses showed that TILP completion varied significantly by immigration status, with U.S. citizens and undocumented youth more likely to have TILPs compared with legal residents or those with missing immigration status. Multivariate models showed that young people with a missing immigration status were less likely to have any TILPs, had fewer TILPs, and had higher TILP discrepancies than U.S. citizens. Sensitivity analyses suggested that immigration status missingness considerably biased results, highlighting the limitations of administrative records when studying a population of young people whose legal status may not be fully known or documented in administrative systems.
Limitations
This study has limitations. Data are derived from administrative case records not intended for research purposes. Administrative records are not psychometrically tested, and data entry practices vary by county and over time. For instance, administrative records do not allow us to discern whether immigration status missingness is legitimate or selectively not entered. Second, our data only follow young people until their 21st birthday due to extended foster care eligibility criteria. Future research that follows youth throughout early adulthood may yield additional insights. Due to data de-identification requirements, our analyses omitted young people who were asylees, refugees, or had other immigration statuses that occurred at low frequencies in our sample. Future research is needed to understand how young people with other immigration statuses navigate child welfare and immigration services. Given the results of our sensitivity analyses, future studies using complete immigration status data are needed to build a comprehensive understanding of the relationship between immigration status and TAY outcomes. Finally, our results may not be generalizable to states with different immigration policies and legal service landscapes. California is home to several sanctuary cities that may have local policies that encourage discretion when documenting legal status in government databases.
Implications for Practice
Our findings highlight opportunities to align social work practice and policy with the service needs of undocumented young people in foster care by (a) providing caseworkers with adequate training and resources, (b) creating additional legal pathways to residency and citizenship, and (c) funding research and evaluations of services for this unique population.
Recommendation 1: Provide Caseworkers With Training and Resources Pertaining to Undocumented Youth in Foster Care
The high degree of missing data in our analyses is likely linked to several factors (e.g., caseworker discretion, administrative burden). However, young people’s hesitation to disclose their legal status cannot be understated. Although child welfare agencies in California do not report youth or families to Immigration and Customs Enforcement (ICE), our observation period included the end of the Obama administration and the entirety of the first Trump administration. While President Obama implemented the Deferred Action for Childhood Arrivals (DACA) and Deferred Action for Parents of Americans and Lawful Permanent Residents (DAPA), previous policies, including the 1996 Illegal Immigration Reform and Immigration Responsibility Act (IIRIRA) and the 1996 Anti-Terrorism and Effective Death Penalty Act (AEDPA) led to the rise of mass deportation during the Obama administration (Martínez et al., 2018). Trump’s first administration ushered in an era of anti-immigrant sentiment and highly publicized policies that further punished and criminalized immigrants (e.g., the public charge rule; the separation of families at the U.S.–Mexico border; Perreira et al., 2018). Consequently, young people who came of age during a time when immigration was criminalized and immigrants faced significant penalties may mistrust governmental institutions (Chavez et al., 2019; Wray-Lake et al., 2018).
As Trump’s second administration introduces a new wave of anti-immigrant policies, it is crucial that caseworkers have the training, resources, and flexibility to assess and address the needs of undocumented young people transitioning from foster care. Allowing caseworkers to exercise discretion when documenting legal status in administrative systems, as is the case with AB 829, can help foster trust with young people and their families. In addition, establishing SIJS units, child welfare units that specialize in working with undocumented youth and their families, can ensure young people are served by caseworkers and professionals who are well-acquainted with local immigration services and can provide support around their immigration needs (Lovato et al., 2024). Although several organizations have developed best practices when working with immigrant families (Migration Policy Institute, 2021; Prandini et al., 2019), few acknowledge the unique challenges facing undocumented young adults. Collaborations between SIJS and transition-age youth units in child welfare agencies may provide a natural context for developing practices pertaining to this population. Ensuring future services and resources are provided in youths’ preferred language may also address the observed disparities in the TILP-related outcomes of youth who predominantly spoke Spanish.
Recommendation 2: Create Additional Avenues for Residency and Citizenship
Increasing access to residency and citizenship is a necessary part of ensuring the safety and well-being of undocumented young people aging out of care. SIJS offers a path to legal residency for immigrant youth who are dependents of the juvenile court and cannot safely reunify with their parents (Junck, 2012). However, the criteria to qualify for SIJS are somewhat strict, with some courts requiring petitions to be filed before age 18. Additional pathways to legal residency and citizenship are needed to accommodate youth whose situations do not align neatly with SIJS criteria but still face significant vulnerabilities. The ongoing implementation of the Family First Prevention Services Act (FFPSA) can potentially address this service need. Signed into law in 2017, the FFPSA allows child welfare agencies to draw upon Title IV-E funding—federal dollars previously reserved for children and youth in out-of-home care—to fund services that can prevent out-of-home care entry (Testa & Kelly, 2020). As prevention service networks emerge, integrating legal services for undocumented and mixed-status families can provide additional reprieve that may stave off foster care entry. Considering local agencies have flexibility in building their prevention networks, incorporating such services into prevention networks can serve as a stop-gap solution while federal and state policymakers pursue longer-term changes (e.g., increasing the flexibility of SIJS criteria).
Recommendation 3: Funding Research and Evaluation of Best Practices and Young Adults’ Experiences
Our analyses underscore the limitations of using administrative records to examine a population of young people whose legal status may not be fully known or documented in government databases. Although data in CWS/CMS is not shared with immigration enforcement, additional efforts are needed to help young people and their caseworkers feel safe enough to document immigration status in administrative records before these data can be reliably used in research. In the meantime, other data can offer crucial insights into the experiences of young people. Researchers can loan their expertise to evaluations of emerging best practices, focusing on processes and secondhand data from caseworkers and other professionals to protect young people’s identities. Ethical uses of anonymous data sources, such as deidentified social media posts, can also shed light on young people’s experiences without compromising their safety (Fowler et al., 2022, 2023).
As of this writing, it is uncertain whether researchers will be able to draw upon federal funding sources to study immigrant and other marginalized populations in the years ahead (Marvis, 2025). Researchers will increasingly depend on funding from state and local policymakers and philanthropic organizations. Such entities looking to improve outcomes for undocumented young people exiting care can provide funding opportunities and, in the case of policymakers, use this research to make data-driven decisions on policies affecting child welfare policy and practice.
Conclusion
The current study leverages population-level data to provide a foundational measurement of TAY’s legal status documentation in child protective case records. Our results underscore both the limitations of administrative case records in studying this population and the importance of investing in services and research tailored to the unique circumstances of undocumented young adults transitioning from foster care. As the current sociopolitical climate attempts to constrain opportunities to support immigrant communities, collaborations between specialized child welfare units, emerging prevention service networks, and researchers offer the possibility to build service infrastructures that can promote the safety and well-being of undocumented young people with foster care experience.
Supplemental Material
sj-pdf-1-fis-10.1177_10443894251340097 – Supplemental material for An Examination of Immigration Status and Its Implications for Transition-Age Youth in the Child Welfare System
Supplemental material, sj-pdf-1-fis-10.1177_10443894251340097 for An Examination of Immigration Status and Its Implications for Transition-Age Youth in the Child Welfare System by Anthony Gómez, Kristina K. Lovato, Andrea Lane Eastman and Mark E. Courtney in Families in Society
Footnotes
Acknowledgements
The Transition-Age Youth Research & Evaluation Hub (TAY-Hub) seeks to improve policies and practices affecting TAY by monitoring outcomes and through applied research that is grounded in engagement with members of the child welfare services community, including those with lived experience of foster care.
Disposition editor: Cristina Mogro-Wilson
Disclaimer
This study was performed with the permission of the California Department of Social Services; however, the opinions and conclusions are solely those of the authors and do not represent the policy or opinions of the collaborating agencies or any California department.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The TAY-Hub is deeply grateful for the generosity of our funders: the Conrad N. Hilton Foundation, Tipping Point Community, Walter S. Johnson Foundation, and California College Pathways Funders Alliance.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
