The United States government, operating through the United States Marshals Service (USMS) and the Bureau of Immigration and Customs Enforcement (ICE), is responsible for housing more than 90,000 persons pending the outcome of criminal proceedings in the federal courts (56,000) or immigration related removal proceedings (35,000). While a portion of these prisoners and detainees are housed in facilities operated by federal agencies, more than three-quarters are housed in facilities operated by non-federal entities, including 40 percent housed in facilities operated by State or local jurisdictions located throughout the United States. Collectively, these facilities house nearly 40,000 federal prisoners and detainees. More than $4 billion is appropriated annually to the USMS and ICE to support federal detention operations. Most of this funding is expended to reimburse State and local jurisdictions for the use of their detention bedspace by USMS and ICE.
Historic Trends In Federal Detention Populations
Over the past 14 years, the number of federal prisoners or detainees in custody has ranged from approximately 80,000 to more than 112,000. The variation in the detention population is primarily attributable to two factors: (1) the number of persons received into the custody of the agencies and (2) the amount of time those persons are held in custody.
Between 2011 and 2017, the USMS detention population declined from 61,701 to 50,232. This decrease was primarily attributable to a change in the prosecutorial priorities of the United States Attorneys. However, following a change in Presidential administration in 2017 and the implementation of Zero-Tolerance immigration policies along the southwest border, the number of prisoners received into USMS custody substantially increased, resulting in a concomitant increase in the USMS detention population.
With the onset of the COVID-19 pandemic in March-2020, federal law enforcement activity was substantially curtailed resulting in a decrease in the number of prisoners received into the custody of the USMS, with only the most serious offenders being received into USMS custody. However, because of the slowdown in court case processing, a backlog criminal cases began to accumulate in the courts resulting in an increase in the USMS detention population. Further, because of social distancing restrictions in the federal prisons, a backlog of sentenced prisoners awaiting commitment to the Bureau of Prisons also began to accumulate resulting in further increases in the USMS detention population.
Accordingly, despite a more than 50 percent decrease in the number of prisoners received into custody, the USMS detention population increased from approximately 60,000 at the start of the COVID-19 pandemic to a peak of approximately 65,000 by June 2021. The increase in the detention population was attributable to a more than 90 percent increase in time-in-detention between 2019 and 2021, which increased from 132 days to 254 days. As time-in-detention has started to decrease (229 days during fiscal year 2023), the USMS detention population has also begun to decrease (59,000).
Similarly, the ICE detention population increased from 33,330 to 38,106 between fiscal year 2014 and 2017. The increase in the ICE detention population, however, was attributable to an increase in time-in-detention rather than an increase in the number of aliens received into custody. During this period the number of aliens received into custody decreased from 210,822 to 187,356 while time-in-detention increased from approximately 29 days to 34 days.
By contrast to the USMS, following the onset of the COVID-19 pandemic, the ICE detention population decreased by more than 50%. While time-in-detention did substantially increase for those aliens in detention, because the international borders were effectively closed and most illegal border crossers were allowed to voluntarily depart without being taken into custody, 80 percent fewer aliens were received into ICE custody. Accordingly, the ICE detention population decreased from a peak of 50,922 to 19,068.
Following the end of the COVID-19 pandemic and its restrictions – to include the end of the Title 42 limitations on migration to the United States – ICE enforcement activity resumed. Consequently, the number of aliens received into ICE custody increased from 109,776 during 2020 to 358,625 during 2023; time-in-detention decreased from 63.5 days to 37.5 days.
Projecting Future Federal Detention Populations
At the State or local level, detention population projections are generally impacted by macro-societal factors such as changing demographics, economic conditions and poverty rates, employment participation rates, illicit drug use, and policing practices.
By contrast, at the federal level, detention population projections often vary from presidential administration-to-administration reflecting the new and/or changing priorities of the nation’s U.S. Attorneys, who are appointed by the President, and the administration’s law enforcement priorities. Similarly, ICE detention populations can vary based on the administration’s response to unforeseen political events abroad, dire economic conditions in neighboring countries, or natural disasters outside the United States that cause mass migration and/or federal policy decisions to detain, release, or remove unlawful entrants.
The federal detention population can also be directly impacted by newly enacted legislation and/or agency appropriations. Specifically, Congressional funding for federal law enforcement, border enforcement, and/or prosecutorial efforts can have a direct impact on the detention population. For example, the creation of the Department of Homeland Security (DHS) in 2003 likely had one of the largest single measurable impacts observed on the detention population. Following its creation, Congress substantially increased funding for Customs and Border Patrol (CBP) agents. Between fiscal year 2003 and 2008, the number of CBP agents increased by approximately 33 percent, increasing from approximately 27,000 on-board during fiscal year 2003 to 37,000 during 2008. As a result, CBP was able to stem the number of illegal entries along the border from an estimated 1.2 million annually to approximately 700,000.
While CBP apprehended few unlawful entrants, they referred a greater number of those apprehended to the US Attorneys for criminal prosecution. Between fiscal year 2003 and 2008, the number of criminal referrals increased three-fold, from approximately 25,000 to 78,400. As a result, the average daily population of immigration offenders held in USMS custody increased by more than 4,000 during this period, increasing from approximately 8,000 during fiscal year 2003 to 12,500 during fiscal year 2008. And, similarly, ICE was required to detain a greater number of those apprehended by CBP pending for removal proceedings. Despite growth in the ICE detention population being constrained by funding limitations, ICE detention population increased by an estimated 5,000 prisoners, from approximately 20,000 detainees during fiscal year 2003 to 25,000 during fiscal year 2008.
Federal law also specifies classes of defendants and unlawful entrants who must be detained pending adjudication or removal from the United States. Specifically, criminal defendants lacking established ties to the community in which they are being prosecuted, those who pose a continued risk to the community or potential witnesses and/or those charged with specific violent, drug, or weapons-related offenses are presumptively required to be detained pending adjudication of their criminal case. (See, 18 USC § 3142.) Similarly, detention determinations made by ICE for immigration detention are based on the risked posed to the community and the likelihood of appearance at scheduled hearings; the likelihood of availability of relief from removal and the likelihood of relief being granted; and prior immigration violation history. (See, 8 USC § 1226.) Additionally, as practical matter, ICE considers its available bedspace capacity when making detention determinations, reserving limited detention space for whom they are unable to release pending removal.
Accordingly, as the type of offense for which defendants are charged or the nature of the immigration violation changes over time, the proportion of defendants or unlawful entrants received into custody who are detained can also increase or decrease depending on criminal histories and other individual characteristics that statutorily prescribe detention. For example, in federal judicial districts where a high proportion of defendants are prosecuted for immigration offenses, the court-ordered detention rate approaches 100 percent because those defendants have few, if any, ties to the community where they are being prosecuted and present a high risk of failing to appear for future court appearances. By contrast, in districts where there is a greater proportion of financial crimes, the court-ordered detention rate is less because defendants typically have well-established ties to the community where they are prosecuted.
Additionally, with respect to those charged with criminal offenses, as the potential sentence length increases, the time to adjudicate those cases also increases as greater effort is expended to negotiate plea agreements or to adjudicate by trial. As a result, USMS prisoners charged with more serious offenses such as violent, drug, and weapons offenses are typically detained an average of four times longer than those charged with less serious offenses such immigration or property offenses. Similarly, ICE detainees who consent to a “voluntary departure” or “expedited removal” leave the United States shortly after apprehension and therefore are held in detention for relatively short periods. By contrast those who require formal removal or deportation proceedings maybe detained several months pending the outcome of the removal proceedings.
Finally, the ebbs and flows of detainees into the criminal and immigration court systems can result in substantially case processing backlogs that further impact the size of the detention population. Court resources are limited and substantial influx of new cases can substantially increase case processing time resulting in increased detention time for those ordered detained. Alternatively, delays in the nomination and confirmation of new federal judges following changes in presidential administrations can also result in delays in court case processing.
Despite various factors that can lead to year-to-year changes in the number of defendants prosecuted in the federal courts, the US Attorneys leading the 94 federal judicial districts have a core responsibility to investigate and prosecute a wide range of criminal activities in addition to responding to the changing priorities of various Presidential administrations. Referrals to the US Attorneys representing a wide array of criminal activity are made from the federal law enforcement agencies such as the Federal Bureau of Investigation (FBI), Drug Enforcement Administration (DEA), and Bureau of Alcohol, Tobacco, and Firearms (ATF); State or local law enforcement agencies might also refer matters to the US Attorneys. Following referral, the US Attorneys evaluate and make determinations whether to adopt for prosecution. Accordingly, the US Attorneys, despite periods where certain types of cases might take higher or lower precedence for prosecution, have a core workload that typically results in many defendants being ordered detained by the courts and remanded to the USMS for detention. The USMS, therefore, core population of prisoners for whom they are responsible for securing detention bedspace in the federal judicial districts.
Finally, the statistical accuracy of detention population projections is generally dependent upon the size of the base population, the variability of the underlying data series trend, and the length forecast interval. For example, projections of larger populations tend to be more precise than smaller populations; increased variability in the underlying trend yields less precise projections; and longer forecast intervals result in greater uncertainty of future events.
Conclusion
While expectations of future federal detention population levels are highly dependent on the prosecutorial priorities of future Presidential administrations, other exogenous factors can also have a substantial impact. Past surveys of experts suggest that projection models are considered reliable if the projections are within two percent of the actual population at a one-year interval. Finally, regardless of the absolute population level, the USMS and ICE will continue to have a core requirement to house federal detainees near federal courthouses, the international borders, and ports of entry.
John Scalia is a professional statistician with expertise in detention management and detention population projection model development. As the former chief statistician for the United States Marshals Service Prisoner Operations Division, Mr. Scalia was responsible for development of the USMS’s detention population projection and cost estimation models.