Abstract
Between October 1, 2014, and March 1, 2023, the US Office of Refugee Resettlement (ORR) took custody of 568,890 unaccompanied migrant children. Drawing on a unique dataset that ORR produced in response to Freedom of Information Act requests and litigation, we provide the first comprehensive, long-term demographic study of the population of unaccompanied migrant children while in ORR custody. Our analysis reveals the children's differential treatment and experiences across time and demographic characteristics. We contextualize ORR's treatment of the children in its custody by identifying legal standards governing their care. We also examine how specific instances of legal, political, and social change in the United States correlate with fluctuations in the origin-based demographics of unaccompanied migrant children seeking refuge in the United States. Results highlight the differential treatment and intense uncertainty that unaccompanied migrant children face in the United States based on their age, gender, country and region of origin, ORR program placements, and discharge types.
Introduction
Between October 1, 2014, and March 1, 2023, the United States Department of Health and Human Services’ Office of Refugee Resettlement (ORR) took custody of nearly 570,000 unaccompanied migrant children. Administrative records on unaccompanied migrant children taken into ORR custody prior to October 1, 2014, do not exist. Thus, our study begins at the earliest time the data is available. These children represent a significant –– and extremely vulnerable –– subset of migrants who enter US territory every year without lawful immigration status. Recent research on unaccompanied migrant children in the United States highlights the dangers they flee in their own countries, as well as the anxiety and precarity they face as they seek asylum (Galli 2023). Many border-crossing children experience stress and trauma before, during, and after their migration journeys (Berger Cardoso et al. 2019), even as they also show high levels of resiliency (Olwig and Gulløv 2003).
Recent research on unaccompanied children's treatment while in US government custody highlights disparate experiences across government agencies, time, and social and political contexts (see, e.g., Heidbrink 2014; Pedroza 2022; Schrag 2020; Terrio 2015). While unaccompanied migrant children have gained more legal protections in recent decades, current systems are still set up to prioritize enforcement, rather than immigration relief (Heidbrink 2014; Terrio 2015). However, in-depth, demographic research on the experiences of unaccompanied children in ORR custody remains relatively sparse.
Despite the size and vulnerability of the unaccompanied migrant child population in the United States, no comprehensive study exists regarding these children or how demographic characteristics may affect their experiences and treatment while in ORR custody. Utilizing a unique dataset obtained through Freedom of Information Act (FOIA) requests and litigation, we provide an analysis that reveals the children's differential treatment and experiences across time and based on certain demographic characteristics. Our research builds on Emily Ryo and Reed Humphrey's “Children in Custody” as we present a comprehensive, detailed analysis across time and demographic characteristics that their data did not permit (Ryo and Humphrey 2021). Our study also heeds UNICEF's renewed call for improved data and analysis regarding migrant and displaced children (UNICEF 2023).
We contextualize ORR's treatment of the children in its custody by identifying legal standards governing their care. We also examine how specific instances of legal, political, and social change in the United States correlate with fluctuations in the origin-based demographics of unaccompanied migrant children seeking refuge in the United States. Results highlight the intense uncertainty that unaccompanied migrant children face in the United States. Moreover, their experiences in ORR custody have long-term implications for their health and well-being throughout the life course.
Definition of and Legal Framework for Unaccompanied Children
Before the late 1990s, no legal category differentiated unaccompanied children from other migrants arriving at the border. In 1997, the Flores Settlement Agreement, a court-supervised consent decree between plaintiffs and the US government, first intimated such a category as it required the segregation of migrant children from unrelated adults (Flores 1997). Five years later, the United States Congress created a strict legal definition of unaccompanied border-crossing children in the Homeland Security Act of 2002 (6 U.S.C. § 279(g)(2)).
Congress reiterated that legal definition in the William Wilberforce Trafficking Victims Protection Reauthorization Act of 2008 (TVPRA). The TVPRA defines an “unaccompanied alien child” as a person who is (1) under 18 years of age, (2) without lawful immigration status in the United States, and (3) for whom no parent or legal guardian is in the United States or no parent or legal guardian in the United States is “available to provide care and physical custody” (6 USC § 279(g)(2), 8 USC § 1232(g)).
When federal Department of Homeland Security (DHS) officials detain an irregular migrant — someone they suspect of entering the country without legal authorization — they make an initial decision as to whether the person is an unaccompanied child. With respect to the third prong of the test, DHS does not distinguish between children it has separated from family members and those who arrive truly unaccompanied. If DHS determines that an irregular migrant meets the definition of an unaccompanied child, it transfers custody of the child to ORR. In turn, ORR places the children with contracted private entities in locations across the United States. Children remain in ORR custody and in varying placements until they no longer meet all aspects of the legal definition of an “unaccompanied alien child.”
We note that the TVPRA treats unaccompanied children from countries contiguous with the United States (i.e., Mexico and Canada) differently than children from all other countries (8 USC § 1232 (a)(2)(A)). If, in the detaining DHS officials’ judgment, an unaccompanied child from Mexico or Canada does not meet very limited criteria regarding human trafficking and credible fear of persecution and is able to make independent decisions, the detaining officer can immediately return the child to Mexico or Canada rather than referring the child to ORR (8 USC § 1232 (a)(2)(A)&(B)). Thus, the number of unaccompanied children admitted into ORR custody does not represent all the unaccompanied children from Canada and Mexico who arrive in the United States.
As a federal statute, the TVPRA creates a mandatory structure and timeline for detained children's custody and release. For example, once ORR has received custody of a child, it must “promptly place[]” the child “in the least restrictive setting that is in the best interest of the child” (8 USC § 1232(c)(2)). The statute also limits when a child can be placed in more restrictive facilities (8 USC § 1232(c)(2)). Additionally, the Flores Settlement Agreement sets a detailed national standard for the care of detained border-crossing children (Flores 1997). Like the TVPRA, Flores requires that, with limited exceptions, detained migrant children be placed “in the least restrictive setting appropriate to the minor's age and special needs” (Flores 1997, 4–5, 7, 12). Flores adds that those placements must be in State-licensed programs in “safe and sanitary” conditions where the detained children receive medical, dental, and psychological care as well as education and recreational activities (Flores 1997, Exhibits 1 & 2). Flores obligates the US government to release detained children from its custody “without unnecessary delay” (Flores 1997, Exhibit 2). Despite various legal and administrative challenges to it, the Flores Settlement Agreement was in effect during the time of our study and remains so today.
US Migration Policy Context: October 2014–March 2023
Below, we highlight select legal, political, and social events that took place in the United States during the study period. These events correlate with the number of unaccompanied migrant children referred and admitted to ORR custody, as well as the amount of time children spent in custody. As we note in the Results and Discussion sections below, conditions in sending countries also influence changes in admissions into ORR custody, but given their world-wide scope, our select events timeline does not address them. We focus on internal US policy here. We nonetheless acknowledge other factors outside the scope of our detailed consideration that may correlate with changes in migration flows such as US foreign policy, policies and practices in countries of origin, and events at the regional and world-wide level.
Drastic Increase in Unaccompanied Minors and Families Arriving at the Border
In spring and summer 2014, immediately prior to the beginning of our study period, the United States experienced at its border with Mexico a sudden increase in unaccompanied minors and families (parents traveling with minor children) from El Salvador, Guatemala, and Honduras (Lind 2014). Many of these children and families fled to the United States to escape violence, poverty, exploitation, and coercion from increasingly powerful and violent gangs in their countries of origin (Ataiants et al. 2018; Rose 2014; UN High Commissioner for Refugees 2020). According to US Customs and Border Protection (CBP), arrivals of unaccompanied minors from these three countries in Fiscal Year (FY) 1 2014 increased two to three times over arrivals from those same countries in FY 2013 and anywhere from eight to 18 times over arrivals from those countries in FYs 2009, 2010, and 2011 (US CBP 2015). Prior to this surge, ORR had only maintained 6,000 to 8,000 beds for unaccompanied minors; the arrival of more than 10,000 minors per month during the height of the surge and over 68,000 unaccompanied children during FY 2014 overwhelmed ORR and left it scrambling to provide sufficient housing and services to meet the legal standard of care for children in its custody (Lind 2014; US CBP 2015, 2019). This significant shift in 2014 proved a watershed in US migration trends among unaccompanied minors and families from Central America (and beyond), as they have continued to arrive at the US border at much higher levels than before 2014 (Chisti, Pierce, and Telus 2019; US CBP 2023). Before, during, and after the surge, the Obama Administration maintained stringent immigration enforcement policies that resulted in the removal of “more than 3.4 million noncitizens from the United States” through fall 2016, more than any prior presidential administration (Martínez, Slack, and Martínez-Schuldt 2018, 173).
US Presidential Election
The 2016 US presidential election, and the year of campaigning that preceded it, also shaped views on immigration and migration behavior. In a June 2015 speech launching his presidential campaign, Donald Trump asserted that, “The US has become a dumping ground for everybody else's problems,” a claim he supported with assertions of the poor quality of migrants arriving to the United States from Mexico, “South and Latin America,” and the Middle East (Phillips 2017). Immigration restriction, punishment for undocumented immigrants, and plans to “build a wall” became regular talking points in Trump's speeches and a key focus of his campaign (Qiu 2016). Trump's inflammatory rhetoric around migrants and migration contributed to further polarization and politicization (Hoekstra and Orozco-Aleman 2021). Trump's electoral win in November 2016 changed the United States’ immigration policy and enforcement landscape. A temporary but significant drop in unauthorized entry attempts at the US border with Mexico accompanied his presidential inauguration in January 2017 (Hoekstra and Orozco-Aleman 2021). During his first week as president, Trump signed three executive orders that collectively authorized: (1) expanding migrant detention; (2) increasing enforcement in the US interior and at the US-Mexico border; (3) limiting access to asylum; (4) suspending the refugee resettlement program; (5) banning legal entry for all foreign nationals from seven predominantly Muslim countries; and (6) initiating necessary steps to build a 2,000 mile-long wall along the US-Mexico border (Center for Migration Studies 2017).
“Zero Tolerance” and Family Separation Policy
Within days of Trump's inauguration, officials in his administration began discussing the unprecedented possibility of “separating families crossing the border together” (American Oversight 2023; Dickerson 2022; Southern Poverty Law Center 2022). As a policy option, separating families would enable the government to detain and criminally charge unauthorized adult border crossers to punish them and, some hoped, to deter future unauthorized migration to the United States (American Oversight 2023; Dickerson 2022; Southern Poverty Law Center 2022). Sometime between early March and July 2017, the El Paso sector of Customs and Border Patrol began de facto family separation by suspending its “family unit policy,” which had mandated that border-crossing parents and their children be kept together, whether detained in DHS facilities for families or released into the United States to await their immigration case proceedings (American Oversight 2023). By the end of 2017, officials working in many sectors along the US-Mexico border were separating parents and their children, including families arriving at official ports of entry (Southern Poverty Law Center 2022).
On May 7, 2018, the US Department of Justice announced the official implementation of a “‘zero tolerance’ policy for illegal entry on [the US] Southwest border” requiring the criminal prosecution of all unauthorized border crossers –– including those seeking asylum (Sessions 2018; Southern Poverty Law Center 2022). Zero tolerance required that all adults seeking to enter the United States without authorization be separated from any children, including their own, traveling with them; adults were transferred to immigrant detention centers and children were relocated into ORR care (Román and Sagás 2021). Mass outrage followed. The publicization and ensuing media coverage prompted Trump to sign an executive order on June 20, 2018, mandating that DHS cease separating parents and children except in rare cases (Southern Poverty Law Center 2022; Trump 2018). While the official implementation of family separation ended in June 2018, the consequences of the program lingered far beyond. The American Civil Liberties Union estimated that the Trump Administration separated over 5,400 children from their parents at the border under the policy (Spagat 2019). Over two years later, in late 2020, at least 545 separated children had not yet been reunited with their parents (Dickerson 2020).
In April 2018, the Trump Administration implemented additional surveillance of the sponsors to whom it would release unaccompanied children, including information sharing among federal agencies and expanded fingerprinting requirements for all members of a potential sponsor's household (US ORR 2018; US ORR et al. 2018). Both the information sharing and the fingerprinting requirement deterred potential sponsors, particularly those without a legal immigration status, and increased the time unaccompanied children spent in ORR custody (Women’s Refugee Commission 2019). ORR rescinded significant portions of the surveillance policies in December 2018 and March 2019 (US ORR 2019).
Migrant Protection Protocols/Remain in Mexico Policy
In December 2018, the Trump Administration's Department of Homeland Security implemented a new policy along the US-Mexico border. Incongruently named the Migrant Protection Protocols and commonly called “Remain in Mexico,” the policy required most border crossers — including asylum seekers — to wait in Mexico, rather than in the United States, for the months or years it could take for resolution of their US immigration cases (Nielsen 2019, 2018; TRAC Immigration 2018). From its inception, the Migrant Protection Protocols exempted unaccompanied minors from its strictures (US DHS 2019).
COVID-19 and Title 42 Border Closure
The onset of the COVID-19 pandemic and the subsequent closure of the US-Mexico border brought with it a significant decrease in the number of unaccompanied children admitted into ORR custody. On March 20, 2020, the Centers for Disease Control and Prevention (CDC) relied on two sections of the Public Health Service Act of 1944, colloquially known as Title 42, to close the border to irregular migrants who arrived in the United States by land from Canada or Mexico, regardless of their country of origin, and to suspend their right under both US and international law to claim asylum in the United States (American Immigration Council 2021; US CDC 2020d).
The CDC's order applied only to migrants who, because of their irregular status, would have been placed “into a congregate setting” for processing in Customs and Border Patrol facilities at or near the border (US CDC 2020d). The CDC justified its order as “necessary to protect the public health” on the grounds that irregular migrants arriving by land posed “a serious danger” of introducing and spreading COVID-19 while in US immigration processing (US CDC 2020d). The CDC repeatedly extended the substantive provisions of the order (US CDC 2020b, 2020a, 2020c, 2021c).
The CDC's order allowed the expulsion of unaccompanied children until November 18, 2020, when a federal district court prohibited the practice (P.J.E.S. v. Wolf 2020). The CDC published notice that it would exempt unaccompanied children from its Title 42 expulsion order in February 2021 (US CDC 2021a, 2021b).
US Presidential Election
Joe Biden was elected president of the United States on November 3, 2020, and inaugurated on January 20, 2021 (National Archives 2021). Biden campaigned on promises to reverse Trump Administration policies and to reform immigration to support families and asylum seekers, among others (Rose 2014). However, to date, Biden's mixed record and lack of success on immigration matters has drawn strong criticism from the left and the right (Shear 2023). Moreover, through the end date of this study, federal court litigation stymied the Biden Administration’s attempts to formally terminate both the Remain in Mexico policy and the Title 42 border closure (Chisti and Bush-Joseph 2023). In February 2021, the Biden Administration formally announced its policy to exempt unaccompanied children from expulsion under the Title 42 border closure, as a federal judge had required in November 2020. Media and human rights reports suggest that some families still subject to Title 42 opted to send their children alone once this exemption was implemented (see Montoya-Galvez 2022; Pillai and Artiga 2022). In FY 2022, the last complete year for which records are available, the Biden Administration engaged in nearly 2.4 million migrant enforcement encounters (i.e., apprehensions, processing of noncitizens, etc.) along the border with Mexico, the highest number ever recorded (Ruiz Soto 2022).
Current Study
Even with increasing political and media attention to the numbers and fates of unaccompanied migrant children in US government custody, no comprehensive study provides a demographic profile of these children or accounts for their experiences in ORR custody. In this study, we generate population descriptive statistics of unaccompanied border-crossing children in ORR custody from October 1, 2014, to March 1, 2023.
Method
Data
We draw on information that ORR provided in response to Freedom of Information Act (FOIA) requests and litigation to produce a detailed demographic dataset of all children who entered ORR custody between October 1, 2014, and March 1, 2023. 2 During that period, ORR took custody of 568,890 children.
ORR produced the data in two batches (October 1, 2014–July 31, 2020, and August 1, 2020–March 1, 2023). For each of the 568,890 children, the dataset provides date of entry into ORR custody; gender; country of origin; program placements by name, type, and date; and discharge categories and dates. ORR provided year of birth for children in Batch 1 and age for children in Batch 2. FOIA officers blinded the batches by removing each child's Alien Registration Number and ORR-assigned identification number.
The Department of Health and Human Services makes some aggregate data regarding unaccompanied children available on its website beginning with FY 2018 (US HHS 2024). ORR makes some aggregate data available beginning with FY 2012 (US ORR 2024). That publicly available data, however, does not permit detailed analysis of children's differential experiences in ORR custody both because it is aggregate and because it is incomplete. In its aggregated form, the data does not allow cross-tabulation over children's multiple demographic characteristics. The publicly available data is incomplete, in part, because it does not include all placement types for all children nor does it include all discharge types. The granularity and completeness of the data ORR produced for us permits our detailed analysis of children's differential experiences in ORR custody across multiple demographic characteristics. As our Results and Discussion sections present, our study reveals important aspects of children's differential experiences in ORR custody that the publicly available data has obscured.
Measures
Individual sociodemographic information. Available demographic characteristics of the unaccompanied migrant children (hereafter referred to as “children”) include gender, country of birth, year of birth (for Batch 1), and age at entry into ORR custody (for Batch 2). Gender is provided as male or female. We grouped the 125 listed countries of birth into nine regions for ease of visualization and interpretation (see Table 1). 3
Demographic Characteristics of Unaccompanied Migrant Children in ORR Custody Between October 1, 2014, and March 1, 2023.
Note: ORR = Office of Refugee Resettlement.
Data from October 1, 2014, to March 1, 2023.
For age, we used the year of birth and year of ORR acceptance to calculate age at entry into ORR custody for Batch 1, to match the available age data provided for Batch 2. We then bundled these ages into five groups: 0–5, 6–12, 13–15, 16–18, and 19 +. Although 18 years old are adults not eligible for ORR custody, we include them in the age group “16–18” because of data limitations. Our age calculation strategy for children in Batch 1 assigns 18 to any child who would turn 18 during the calendar year in which they were admitted into ORR custody. Although most children were still 17 at initial ORR intake, our data do not allow for a more accurate calculation, thus the delimitation of this group to include 18 years old for analysis. We also bundled a small group of individuals 19 years and older, all ages not eligible for ORR custody. Ninety-nine percent of those 19 and older were individuals whom ORR considered children on their entry into custody but later determined to be adults. Data production from ORR included the redetermined birth year or age rather than at admission.
ORR-specific information. “ORR Acceptance Date” is the date on or after October 1, 2014, and on or before February 28, 2023, that ORR took custody of a child. “Initial Placement” is the date on which ORR first placed a child in a facility. “Placement type,” described in Table 2, is one of 12 placements available for unaccompanied children including: Shelter, Emergency Intake Site, Transitional Foster Care, Influx Care Facility, Long Term Foster Care, Staff Secure, Virtual, Secure, Residential Treatment Center, Therapeutic Group Home, Therapeutic Staff Secure, and Group Home (see Table 2 for full definitions). The dataset allows longitudinal observations of a given migrant child from the date of their acceptance through final discharge from ORR custody.
Placement Types and Average Time to Final Discharge.
Note: Data from October 1, 2014, to March 1, 2023. ORR = Office of Refugee Resettlement.
Source: US ORR (2023b, 2023c.)
Standards vary. Compliance with Flores is required only after an EIS has been in operation for six months (US ORR 2022).
The dataset categorizes both transfers among ORR placements and a child's final release from ORR custody as “Discharges” (see Table 3 for full definitions). We define both internal transfers, such as departure from one facility and placement at another, and release from ORR custody as discharges. Subsequently, “Discharge Date” is the date on which ORR transferred a child from one placement to another or the date ORR released a child from its custody. ORR further classifies a child's final discharge from its custody as one of 16 types (see Table 3): Reunified (Individual Sponsor), Transfer, 4 Age Out, Age Redetermination, Voluntary Departure, Other, Reunified (Program/Facility), Ordered Removed, Ran Away from Facility, Ran Away on Field Trip, DHS Family Shelter, Local Law Enforcement, Immigration Relief Granted, Marshal's Service, Referral Canceled by Referring Agency, and Referral Canceled-OCONUS Age Out (US ORR 2023b, sec. 2.2.1).
Discharge Types and Average Time to Final Discharge.
Note: Data from October 1, 2014, to March 1, 2023. Sorted from largest to smallest number of children who experienced each discharge type as their first discharge. Cases without information on first discharge (n = 4,875; 0.86%) are not included in the table.
“Total number of discharges” is a summed measure that includes the number of internal transfers and final discharge. We identified initial discharge type versus final discharge type, which are often the same as 89 percent of children are discharged only once. Lastly, we used the date of ORR acceptance and date of final discharge to calculate the total “time in ORR custody,” measured in days.
In order to protect the identities and private information of unaccompanied migrant children, we aggregate data wherever possible even as we provide a detailed demographic profile of them. In cases where a specific category (i.e., region of birth, age group, etc.) includes fewer than 100 children, we have replaced the actual numerical count with “< 100” and instead provide general frequencies.
Results
Admissions Into ORR Custody Over Time: A General Timeline
Figure 1 documents the number of unaccompanied migrant children admitted into ORR custody from October 1, 2014, through February 28, 2023, with select political, social, and legal events noted. We do not include admissions trends after February 28, 2023, that ORR has made publicly available because the end date of our data does not permit their detailed analysis. Admissions are depicted at the month-year level, where each datapoint captures the number of new admissions into ORR custody, rather than a current census of all children in custody at the time. Four hundred fifteen admissions on March 1, 2023, are excluded from this and all subsequent figures, as we do not have complete information for the month of March 2023; this data is retained in all aggregate estimates in all tables. We highlight a few important trends here.
Number of Admissions Into Office of Refugee Resettlement (ORR) Custody Over Time.
During the Obama Administration, from October 2014 to shortly before Donald Trump took office in January 2017, the number of children admitted into ORR custody rose relatively steadily, with a dip in January and February of 2016 during the Obama Administration. A precipitous decline began in late 2016 and extended through the first four months of the Trump Administration.
Other key events merit attention. The Trump Administration's family separation policy was formally short-lived, officially lasting only about six weeks; during that time, the number of children admitted to ORR custody does not appear to fluctuate. However, as we note in greater detail in Figure 3, the time children spent in ORR custody increased significantly during this period.
While the number of admissions into ORR custody overall was lower during the Trump Administration than during the Biden Administration, we saw a short-lived increase in the number of admissions shortly after the Remain in Mexico policy took effect in December 2018. By mid-2019, the number of children admitted into ORR custody declined to the running average. When the Title 42 border closure took effect in March 2020, the number of new children admitted to ORR custody per month dropped to below 200 per month from April through July 2020, during the initial peak of the COVID-19 pandemic.
A gradual rebound in child admissions began in August 2020 and then skyrocketed in 2021, likely related to President Biden's inauguration and changing expectations regarding immigration policy. In February 2021, the Biden Administration formally announced its policy to exempt unaccompanied children from expulsion under the Title 42 border closure, as a federal judge had required in November 2020. Under the Biden Administration, the number of new unaccompanied migrant child admissions peaked at 19,131 in April 2021. From April 2021 through February 2023, the number of children admitted into ORR custody has fluctuated from month to month with even the lows consistently higher than peak rates during the Trump and Obama Administrations.
Demographic Characteristics of Unaccompanied Children and Average Days in ORR Custody
To better understand the population, we identify demographic characteristics of unaccompanied migrant children in ORR custody between October 1, 2014, and March 1, 2023, in Table 1. Here, we include key descriptive statistics across demographic characteristics for all children during the study period. All tables are presented at the person-level. Overall, 66 percent of the children are male, 5 with little variation in average age at ORR acceptance (14 for females, 15 for males), average days from acceptance to first placement (1 for females, 0.98 for males), and average days in ORR custody by gender (38 for females, 41 for males). Roughly 60 percent of children are aged 16–18 at ORR acceptance, while 3 percent are ages 0–5, 13 percent are ages 6–12, 24 percent are ages 13–15, and 0.5 percent are ages 19 or older. The percentage of female unaccompanied migrant children decreases as age approaches 18. Gender parity is almost reached among children ages 12 and younger; conversely, female migrants make up only 36 percent of children aged 13–15, and roughly 30 percent of children 16 and older. Average days to first placement are similar across all age groups. 6 Average days in ORR custody range from a high of 43 days for children aged 0–5 and a low of 27 days for those 19 or older. 7
ORR took custody of 48 percent of the nearly 570,000 children in the dataset during the span of just two years: 2021 (26%) and 2022 (22%). In contrast, of the full years for which we have data (2015–2022), the year 2020 accounts for < 3 percent of the admissions, due in large part to the COVID-19 pandemic-related restrictions on migrant admissions. Female unaccompanied children make up roughly one-third of the admissions with a high of 37 percent in both 2022 and the first two months of 2023. Similarly, the mean age of children admitted into ORR custody hovers around 15 for all years with similar average days from acceptance to first placement across time. Lastly, average days in ORR custody reached peaks in 2017 (61 days) and 2018 (78 days) and declined to a low of 17 days in the first two months of 2023.
Children from the Northern Triangle countries of El Salvador, Guatemala, and Honduras make up roughly 91 percent of the children admitted to ORR custody between October 1, 2014, and February 28, 2023. 8 Guatemala alone accounted for 46 percent of those children. Just over 39 percent of the children from El Salvador were female, compared to 29 percent from Guatemala, and 37 percent from Honduras. The Northern Triangle children were roughly 15 years old, had on average one day between ORR acceptance and first placement, and averaged from 33 to 43 days in ORR custody.
Our analysis of data by region also reveals an important difference over time. Immigration enforcement data show that in FY 2022, for the first time ever, Customs and Border Patrol encountered more Venezuelans, Cubans, and Nicaraguans than individuals from the Northern Triangle (Ruiz Soto 2022). Data for unaccompanied children admitted to ORR custody, however, did not reflect a similar demographic shift. Rather, in FY 2022 and throughout the timeframe of our study, the three Northern Triangle countries remained constant as the primary countries of origin for children admitted to ORR custody, even as admissions from other countries fluctuated.
Children from North America — Mexico, Canada, and the United States — form the next largest birth region after the Northern Triangle. ORR admitted to its custody almost 16,000 children from Mexico, over 1,000 identified as born in the United States, and fewer than 100 children from Canada. The average age of children from North America ranged from 13 for Mexico-born children to only 0.3 — about four months — for children born in the United States (In a separate paper, we address the anomaly of children born in the United States, and thus US citizens, in ORR custody. Precisely, 97 percent of the US-citizen children in our study were infants born to girls also in ORR custody.).
South America made up roughly 2.5 percent of the children in the dataset with Colombia and Ecuador sending the most children from the region. Central America, excluding the Northern Triangle countries, made up almost 2 percent, Asia and the Pacific just over 1 percent, and the West Indies and the Caribbean 0.6 percent. The regions of Africa and Europe each contributed just over 0.1 percent. A small number of children were identified as being stateless or with an unknown country of birth.
Important variations occur by region. For example, on average, unaccompanied children from Africa spent the longest time in ORR custody by far, almost 125 days compared to 40 days for children from the Northern Triangle and 26 days for children from South America. Further, only 10 percent of children from Asia and the Pacific were female. For important variations by country of birth rather than region, see Supplemental Table A1 for differences in age, gender, days to first placement, and days in ORR custody.
Figure 2 shows admission into ORR custody by age group over time. Children ages 16–18 consistently made up the largest number of admissions across all years, followed by ages 13–15, 6–12, 0–5, and then 19 +. This ordering of prevalence persists across all years and generally follows the spikes and dips seen in all admissions. However, children ages 16–18 see the most dramatic fluctuations across time.
Admissions Into Office of Refugee Resettlement (ORR) Custody by Age Group Over Time.
Figure 3 shows the average days in ORR custody over time for all unaccompanied children admitted between October 1, 2014, and February 28, 2023. Average days in ORR custody are determined by taking the date at final discharge and subtracting the ORR admission date. Average days in ORR custody hovered between 40 and 60 days until roughly November 2016 when it began to rise and then spiked at just above 100 days in mid-2018 following the Trump Administration's family separation policy and increased surveillance of potential sponsors. Following the spike in 2018, average days in ORR custody swiftly declined to a low of just under 20 days in 2020 followed by an increase back to 40 days and then a steady decline to an average of 17 days in early 2023.
Average Days in Office of Refugee Resettlement (ORR) Custody Over Time.
Figure 4 represents the average days in ORR custody by region of birth. As noted above, children from Africa consistently have the highest average days in ORR custody across time with occasional spikes for those from other regions (In a separate paper, we address the custody, treatment, and discharge of children from African countries, which significantly differ when compared to other children across numerous measures.).
Average Days in Office of Refugee Resettlement (ORR) Custody by Region of Birth.
Placement and Discharge Types
To better understand the experiences of unaccompanied migrant children in ORR custody, we provide descriptive statistics and definitions for all placement types (Table 2) and all discharge types (Table 3).
Table 2 lists each placement type as published ORR documents define it, with the exception of “Virtual” (US ORR 2022, 2023b, sec. “Terms”). Because ORR has not published a definition of a “Virtual” placement, we rely on an informal description ORR provided orally. 9 For each placement type, we include the total number of children in the placement, the frequency (%) of each placement, the average days to final discharge by placement type (including timing of placement), and the average age of children by placement type. 10 We do not disaggregate frequencies for placement types by gender, as gender breakdowns are consistent with the overall gender distribution of children in ORR custody. Placement types are sorted by the frequency of a child ever experiencing the placement, from most common to least common. Shelters are both the most common first placement (70%) and the most common placement overall (73%); over 400,000 unaccompanied children spent at least some time in a shelter while they were in ORR custody. Children who were ever placed in shelters spent an average of 45 days in ORR custody. The next most common placement types are emergency intake sites, transitional foster care, and influx care facilities.
Emergency Intake Sites (EIS) and Influx Care Facilities (ICF) deserve particular consideration. Our data show that ORR first began using both in mid-March 2021. Both were also designed to provide temporary housing during specifically defined emergencies but without the full panoply of Flores Settlement protections (US ORR 2021, 2023b). A total of 85,530 children, about 15 percent of children during the study period, were ever placed in an EIS. Children ever placed in an EIS spent, on average, 31 days in ORR custody while children ever placed in an ICF averaged 25 days. About 6 percent (n = 36,531) of the children spent at least some time in an ICF.
Although not common, virtual placements also warrant scrutiny because they too are an ad hoc crisis response. Like EIS and ICF, ORR first used “Virtual” as a program type in March 2021. ORR has not, however, published a definition of a virtual placement nor when ORR may use it. In oral communication with ORR, 11 a “Virtual” placement was described as an ad hoc means ORR created to assess unaccompanied children who may be in transit to the United States prior to their physical arrival. This placement type was used primarily, although not exclusively, for children from Afghanistan after the fall of Kabul in August 2021. Our data show that 1,258 children were ever placed in a virtual program, 95 percent of whom were from Afghanistan. Children who were ever placed in a virtual program averaged about 8 days in ORR custody, the shortest average time in custody across all placement types. We return to the issue of EIS, ICF, and Virtual placements in the “Discussion” section, below.
Placement types more restrictive than a shelter — residential treatment centers, staff secure facilities, and secure facilities — were less likely as a child's first placement type, but some children were sent to such facilities during their time in ORR custody. More restrictive placements have stricter security measures, lower children-to-staff ratios, and staff trained to address complicated and severe behavioral issues. ORR placed 1,303 children in staff-secure facilities as their first placement, while 2,864 experienced staff-secure facilities at some point.
ORR infrequently sent children first to longer-term placement types. For example, ORR placed fewer than 100 children in long-term foster care initially but transferred about 3,400 children to it at some later point. The average time in ORR custody for children who were ever placed in long-term foster care was about 425 days, the longest average time in custody across all placement types. That longest average time in custody is of particular concern for several reasons. First, younger children were more likely to ever experience long-term foster care, along with its inherent impermanence. A child's age at exposure to adverse experiences like insecurity, disruption, and trauma, as well as their multiplicity, can affect well-being (Ertanir et al. 2023; Barajas-Gonzalez et al. 2021; Kuhlman et al. 2015). Second, ORR long-term foster care allows a child's placement with a family or in a group home, although the data does not differentiate between the two (US ORR 2023b, secs. 1.2.6, 3.6). Residential settings like group homes rather than family-based care pose risks to children's development (Groza and Bunkers 2017).
We also note that the average age of children by placement type is generally consistent with the average age of children in ORR custody across the full sample, with a few notable exceptions. Younger children were more likely than older children to experience transitional foster care, long-term foster care as a first placement, and therapeutic group homes. See Table 2 for additional information on placement types and average time in ORR custody by all placement types. 12
Table 3 follows a similar format to Table 2 and includes a list of each discharge type, as well as accompanying definitions and descriptive statistics for each type. We sorted discharge types by the number of children who experienced each type first, from most common to least common. 13 Here, we emphasize key findings.
First, for 86 percent of children, “Reunified (Individual Sponsor)” was both their first and their final discharge. Another 9 percent experienced “Reunified (Individual Sponsor)” as their final discharge after at least one internal transfer. That is, ORR released 95 percent of children from its custody through reunification with an individual sponsor. Children released to an individual sponsor spent an average of 33 days in ORR custody, although there were significant variations within the years of study. ORR prefers parents or legal guardians, close family, and extended family as sponsors for children; however, if none are available, ORR may also release a child to “distant relatives and unrelated adult individuals” (US ORR 2023b, sec. 2.2.1).
ORR's release of a child from its custody to an individual sponsor does not resolve the child's unauthorized presence in the United States. In fact, a mandatory Sponsor Care Agreement requires the sponsor to agree to “ensure the unaccompanied child's presence at all future [immigration] proceedings” and to “ensure the unaccompanied child reports to ICE for removal from the United States if an immigration judge issues a removal order or voluntary departure order” (US ORR 2023b, sec. 2.8.1). Children released from ORR custody to an individual sponsor are still subject to physical removal from the United States, depending on the outcome of their immigration cases.
After reunification with an individual sponsor, the next most common discharge type was “Transfer.” Discharge through a transfer involves the movement of children between ORR-contracted care providers such as from a shelter to long-term foster care or back. ORR also counts a change in the contract holder for a specific facility as a transfer. Transfers made up 10 percent of first discharges and about 10 percent of all discharges. Children with more than one discharge (n = 56,518) were transferred at least once, prolonging their time in ORR custody to an average of 95 days across time.
Discharges through “Age Out,” and “Age Redetermination” are the next most frequent types. An age out discharge occurs when a child turns 18 while in ORR custody and is no longer a minor. Age outs account for about 1.6 percent of all first discharges and about 2 percent of all final discharges. Children who aged out spent an average of 83 days in ORR custody.
An age redetermination discharge occurs when ORR reevaluates the individual's age and decides whether they are an adult. Age redetermination discharges made up a little over 1 percent of all first discharges and 1.2 percent of all final discharges in our study. While the percentage of individuals experiencing an age redetermination is small, the absolute number is still considerable: 6,203. Not surprisingly, the average ages for children whose final discharges were aging out or age redetermination were the highest compared to all other final discharges.
We note important gender differences by final discharge type. Male unaccompanied minors experienced harsher final discharges at higher rates. For example, only 1.7 percent of unaccompanied children who were referred to local law enforcement were female. Similarly, only 6.7 percent of minors who were ordered removed from the United States were female. Additional information on gender breakdowns for each final discharge are available in Table 3.
Lastly, Figure 5 shows the most common final discharge types by year. 14 Reunification with an individual sponsor is more common across all years, peaking in 2021, consistent with a peak in overall admissions into ORR custody in the same year (Panel A). Panel B provides other common final discharge types. Frequencies for all final discharge types also tend to peak in 2021, with the exception of the “Other” category (highest in 2018), age redeterminations (highest in 2016), and voluntary departures (highest in 2018).
Common Final Discharge Types by Year.
Discussion
This Dispatch provides the first comprehensive, long-term demographic study of the population of unaccompanied migrant children in ORR custody. The time period of our study — the eight years and five months from October 1, 2014, to March 1, 2023 — spans three different US presidential administrations; historic social, political, and legal changes; and unparalleled fluctuations in the numbers of unaccompanied border-crossing children arriving to the United States. Prior academic research has done vital work on discrete aspects of unaccompanied migrant children's experiences in ORR custody: with COVID-19 (Foppiano Palacios, Tucker, and Travassos 2023), in long-term foster care (Crea et al. 2022; Evans et al. 2024; Ramos 2022), and as particular demographic subgroups (Carlson, Cacciatore, and Klimek 2012; Crea et al. 2018; Evans, Culley, and Crea 2022). Our study adds value to the scholarly discourse through its thematic breadth, its region-level analysis for children from all sending countries, and the recent end date of the study period.
Over the study period, ORR took custody of more than half a million children from 125 countries, 91 percent originating in the Northern Triangle of El Salvador, Guatemala, and Honduras. The majority of children (60%) in ORR custody during this time were aged 16–18. Almost twice as many boys as girls — 66 percent versus 34 percent, respectively — entered ORR custody during this time period. Admissions for all children reached unprecedented highs in calendar years 2021 and 2022. Youths spent an average of 40 days in ORR custody from their date of admission to the date of their final discharge, with a peak of over 100 days during the immediate aftermath of the formal family separation policy and heightened scrutiny of potential sponsors in 2018. In the time frame of our study, we observed a lower number of children in ORR custody during the Obama Administration (October 1, 2014, to January 20, 2017) and the Trump Administration (January 20, 2017, to January 20, 2021), even as the average time in custody during these administrations was longer than during the Biden Administration (January 21, 2021, to March 1, 2023).
High rates of new admissions of unaccompanied children to ORR custody should raise concern. Despite the Flores Settlement, the TVPRA, and the ORR Program Policy Guide, children in ORR custody continue to face significant risks to their health and well-being. In 2021, the US Senate Finance Committee issued a scathing report based on data from 2014 to 2020, charging that through “years of mismanagement and lack of oversight,” ORR had allowed “abuse and substandard care” for the children in its custody (US Senate Committee on Finance 2021). The report noted systemic problems created by “ORR's lack of a data-driven oversight program,” including risks of sexual abuse (US Senate Committee on Finance 2021). Individual children bear the harm (e.g., violence, lack of mental health services, extended separation from family, etc.) of ORR's systemic failures (Bogado et al. 2018; Bogado and Lewis 2020a, 2020b; Bogado and Morel 2021; Bogado 2022).
We also examined placement and discharge types as proxies for children's experiences. Significant variation existed in the types of placements that children experienced. A total of 36,531 children spent at least some time in influx care facilities, 85,530 spent time in emergency intake sites, and 1,258 children received virtual program placements. ORR first used all three crisis-management placement types in mid-March 2021 when it was unable to take custody of and care for unaccompanied border-crossing children as the Flores Settlement Agreement required. Nearly 60,000 children were transferred across at least two placements.
ICFs, EISs, and virtual placements all highlight challenges ORR has faced in response to surges in unaccompanied children. ORR has incorporated guidance for ICFs in its documents while acknowledging that they may not be licensed and thus not fully compliant with Flores (Montoya-Galvez 2023; US ORR 2023b, sec. “Terms,” 2023c). ORR offered a post-hoc justification for EISs in April 2021 (US ORR 2021, 2022). Still, numerous reports from government officials, advocates, and unaccompanied children themselves recount unsafe and distressing conditions in EISs in violation of Flores (Alvarez 2021a, 2021b; Drake and Cruz 2021; Ordoñez 2021; Taxin 2021). ORR has yet to define publicly exactly what a virtual program is or when it may invoke one. Its Unaccompanied Afghan Minor Processing field guidance provides only an implicit description and only for certain Afghan children arriving after the fall of Kabul in August 2021 (US ORR 2023a).
ORR eventually discharged the majority of children (95%) during the study period through reunification with an individual sponsor. The Flores Settlement Agreement and the ORR Program Policy Guide prioritize the discharge of unaccompanied migrant children to family members. Yet, the release of a child to individual sponsors — even family members — can pose risks. Prior to a child's release, US law charges ORR with vetting potential sponsors’ “ability to provide for the child's physical and mental well-being,” as well as risks that the sponsor might “engage the child in criminal, harmful or exploitative activity” (US ORR 2023b, sec. 2.1). Post-release, ORR provides affirmative oversight only through a single “Safety and Well Being Follow Up Call” made after thirty days (US ORR 2023b, sec. 2.8.4).
The adequacy of ORR's vetting of individual sponsors, the speed with which it has released children to individual sponsors, and its limited post-release follow-up came under criticism just as the study period ended. A New York Times investigation reported widespread labor trafficking of children released to individual sponsors, as well as repeated government failure to act when alerted to the abuse (Dreier 2023a, 2023b). The New York Times also reported that, in its post-release follow-up calls, ORR had failed to connect with 85,000 children released to individual sponsors over the past two years (Dreier 2023a). In April 2023 congressional hearings, ORR director Robin Dunn Marcos defended ORR's vetting procedures and noted that ORR did not have legal responsibility for the children or the authority to track them after their release from ORR custody (Marcos 2023; US House Committee on Oversight and Accountability 2023). While both Democrats and Republicans initially expressed outrage at the labor trafficking of border-crossing children and their post-release vulnerability, to date, the political quagmire has stymied change (Demirjian and Dreier 2023).
Roughly 15,000 children aged out of unaccompanied child status or had their age redetermined to be adults. ORR releases unaccompanied children who age-out to Immigration and Customs Enforcement (ICE). Federal law directs ICE to “consider placement in the least restrictive setting” and makes individuals who age out eligible for alternatives to adult immigration detention (8 USC § 1232(c)(2)(B)). The law does not, however, require the least restrictive setting as the TVPRA and Flores Settlement do for children. Thus, individuals who age out of ORR custody lose some of the protections mandated for children. ORR likewise discharges to ICE individuals determined to be adults. If the individual was never a child in ORR custody, ICE treats them as an adult without the protections considered for those who age out and are mandated for children.
Our analysis provides an up-to-date profile of the scope and experiences of unaccompanied migrant children in the United States over almost a decade. For children in ORR custody, we document disparate experiences by key individual sociodemographic characteristics including age, sex, and region and country of origin. Prior research has highlighted the long-term consequences of migration for adults, children, and families (Garip 2008; Lu and Li 2020; Song and Glick, 2022; Van Hook and Glick 2020); even so, relatively little is known about the long-term impacts of migration for unaccompanied children (Maioli et al. 2021; Patel et al. 2021). Collectively, our results underscore the magnitude of unaccompanied child migration to the United States and the disparate experiences children face in ORR custody.
Our study is not without limitations. First, we rely on administrative data provided by the US government. While these data give us a window into the experiences of unaccompanied migrant children, administrative data are certainly not free from error or bias. Recent research highlights potential technical and ethical issues with using administrative data, including misclassification of important identities (e.g., race) and the fact that research participants cannot opt out of data collection (Penner and Dodge 2019; Laniyonu and Donahue 2023). Even so, this unique analysis highlights important trends and patterns in the admission of children into US government custody and their experiences while in custody. Researchers should continue to study the experiences of unaccompanied migrant children in diverse ways, ensuring their safety and privacy. Second, the data do not allow us to track children before or after ORR custody. We do not capture children's experiences in the custody of Customs and Border Patrol prior to their referral and admission into ORR custody. Thus, the relatively quick time our data shows between admission to ORR custody and first placement (one day, on average) undoubtedly masks the overall time children spend in US government custody (Flagg and Preston 2022; Ordoñez and Farrington 2021). Likewise, we cannot gauge how ORR custody may affect children later in life as the dataset ends on their discharge. Third, our data does not include children's reasons for migrating and we can only correlate events that may have shaped migration flows during the study period. Lastly, our quantitative approach reveals broad patterns, but does not capture children's individual voices, the complexity of their lives, and their treatment within facilities and during transfers and placements, even as we can observe their movement, one by one, through ORR custody. Qualitative research is particularly important for documenting the lived experiences of unaccompanied border-crossing children and their families (Schmidt 2017; Barros-Lane, Brabeck, and Berger Cardoso 2022; Galli 2023; Rafieifar et al. 2023).
Despite these limitations, the current study marks significant progress in better understanding the sizable and growing population of unaccompanied border-crossing children in the United States. Now and in the future, it is vitally important that researchers, policymakers, advocates, and other key stakeholders work together and with unaccompanied border-crossing children and their families, to understand and improve children's experiences in US government custody.
Supplemental Material
sj-docx-1-mrx-10.1177_01979183241252034 - Supplemental material for Unaccompanied Migrant Children in US Government Custody: 2014–2023
Supplemental material, sj-docx-1-mrx-10.1177_01979183241252034 for Unaccompanied Migrant Children in US Government Custody: 2014–2023 by Melissa Alcaraz, Hayley Pierce, Jane Lilly López and Kif Augustine-Adams in International Migration Review
Footnotes
Acknowledgments
A previous version of this paper was presented at the 2023 Annual Meeting of the Law & Society Association. Kif Augustine-Adams was a visiting scholar at the Center for Advanced Migration Studies (AMIS) at Copenhagen University during the preparation of this article.
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) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Notes
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.
