3.3.1 Federal IHP

In 1989, the BOP opened a centralized IHP court site for federal prisoners at the FCI in Oakdale, Louisiana (INS et al. 1996: 2). The Oakdale prison, which opened in 1986 (Reference KahnKahn 1996: 151), was unique because it was the first federal facility operated jointly by the BOP and the INS to house both federal prisoners and immigration detainees (Reference MarcusMarcus 1986: A14; BOP, Report to the Attorney General, H. of Rep. 1989: 302; Reference DowDow 2004: 161–62). Multiple “courtrooms for immigration judges” were constructed “within the secure perimeter of the institution” (Kathleen Hawk, Director, Federal Bureau of Prisons, H. of Rep. 1994: 169). Immigration judges hired at Oakdale handled both the IHP cases of immigrants imprisoned in the BOP facility as well as the non-IHP cases of those detained in the administrative side of the facility (Reference KahnKahn 1996: 151).

Oakdale was also unique because of its huge detention capacity. As originally established, BOP prisoners were brought from facilities throughout the United States to Oakdale at the tail end of their sentences (John J. Clark, Assistant Director, Community Corrections & Detentions, H. of Rep. 1997: 55; GAO 1997: 17–18). A full 600 beds were reserved for “male, non-Cuban, non-Mexican inmates” to participate in the IHP court (Doris Meissner, INS Commissioner, U.S. Senate 1994: 46; H. of Rep. 1997: 55). At the same time as the BOP developed Oakdale for non-Cuban, non-Mexican men, the BOP opened a separate IHP location at the La Tuna FCI in Anthony, Texas, specifically to target Mexican men (INS et al. 1996: 2). Focusing on Mexican incarceration at La Tuna was consistent with the facility's historical roots: it was opened in 1932 and immediately filled with more than 90% Mexican inmates who were forced to labor on the prison's farm (Reference HernándezHernández 2017: 140).

The immigration court data we relied on contained information on national origin and citizenship (Appendix), and thus allowed us to confirm that the IHP often segregated individuals by both national origin and citizenship within IHP courtrooms and residential prison facilities. The data did not include information on the racial and ethnic identifications of IHP participants. However, given the correlation between race, ethnicity, and national origin (Reference GómezGómez 2020), patterns in the immigration court data suggest that racial and ethnic segregation also occurred in the IHP courts and facilities we studied.

As Table 3 highlights, 89% of the IHP cases of individuals held in La Tuna from 1989 to 1990 were of Mexican men and an additional 4% were Cuban. In contrast, the FCI-Oakdale prison did, as intended, almost entirely exclude Mexican and Cuban nationals. As also shown in Table 3, during this two-year period Oakdale's IHP included a concentrated population of men from Jamaica, Nigeria, and Haiti, suggesting that Black immigrants were targeted for incarceration and rapid deportation at the Oakdale IHP. Through these data, we begin to see how the highly segregated organization of the IHP court shaped residential groupings in participating federal prisons.

Table 3 Descriptive Statistics for Federal IHP, by Facility (Fiscal Years 1989–1990)

Building on the Oakdale model, in 1994 the BOP opened a second massive joint BOP–INS facility in Eloy, Arizona (Reference Loyd and MountzLoyd and Mountz 2018: 195). Concept Inc., a private prison contractor, was hired to run the Eloy prison. An attorney who represented immigrants in Eloy described the new facility as “a really wretched place to be, with nothing to do and very little in the way of programming” (Reference VolanteVolante 1994: A2). Central to the design of Eloy and Oakdale as facilities serving both BOP and INS was the construction of “courtroom facilities for hearings” and the commitment from EOIR “to provide a sufficient number of immigration judges and court personnel” (Kathleen Hawk, Director, Federal Bureau of Prisons, H. of Rep. 1994: 168; Reference CreppyCreppy 1995a). By 1998, there were fifteen federal prisons hosting IHP courts (BOP 1998: 16; GAO 1998: 4).

Relying on EOIR court data and other public records, we identified twenty-three different federal prison facilities holding noncitizens participating in the IHP in 2019. Some cities had more than one participating facility. These federal IHP prisons included sixteen operated by the BOP,Footnote ⁵ and seven operated by private for-profit contractors such as CoreCivic or the GEO Group (Reference ChacónChacón 2017).Footnote ⁶ Figure 1 marks with dots the cities with one or more BOP facilities where IHP participants were held. The crosses in Figure 1 represent cities where IHP participants were held in facilities run by private contractors.

Figure 1. Federal Prison Facilities Participating in the IHP (Fiscal Year 2019)

All seven contract IHP facilities also operated as Criminal Alien Requirement (CAR) prisons (Reference KaufmanKaufman 2019: 1401–02). These CAR prisons are run by private corporations, house only noncitizens, and provide less programming and inferior conditions to BOP-run facilities. (Reference GreeneGreene 2001; Reference Loyd and MountzLoyd and Mountz 2018: 196–97; Reference KaufmanKaufman 2019: 1383). As Emma Kaufman has found, CAR prisons not only segregate people by citizenship, but also house almost exclusively Latinos, the majority of whom are Mexican (Reference KaufmanKaufman 2019: 1382).

3.3.2 State IHP

During the early years of the IHP, the INS implemented a “Five State Criminal Alien Model” focused on California, New York, Texas, Florida, and Illinois (GAO 1990: 2; Jack Shaw, Assistant Commissioner, INS, H. of Rep. 1995a: 26; IHP Working Group 1995). In 1987, 99% of all state IHP proceedings took place within these five states (Figure 2). As the IHP expanded during the early 1990s, the immigrant-destination states of Arizona, New Jersey, and Washington took on a larger share of the state IHP cases (Figure 2).

Figure 2. Top Eight State IHP Programs (Fiscal Years 1986–2019)

During the Clinton administration, the INS sought to implement an “enhanced” IHP (INS 1996; BOP 1996: 12; 1997: 15; Peggy Philbin, EOIR, H. of Rep. 2001: 14). The overarching goal of these Clinton-era enhancements was to increase the number of IHP sites and eventually to “complete deportation proceedings on virtually 100 percent of the criminal aliens detained in Federal and State prisons before the end of their sentences” (T. Alexander Aleinikoff, INS General Counsel, H. of Rep. 1995b: 4). By 1995, the IHP operated in forty-one states, the District of Columbia, and the Virgin Islands, with plans to expand further (5).

Today there are far fewer states participating in the IHP (EOIR 2017a, 2018b). In 2019, the final year of our study period, there were only nineteen states with active IHP programs. Among the top eight states already discussed (Figure 2), Washington no longer had an active state IHP program (Figure 3). In 2019, the IHP was also active in Connecticut, Georgia, Iowa, Maryland, Massachusetts, Michigan, Minnesota, Nebraska, New Hampshire, Ohio, Pennsylvania, and Wisconsin. Figure 3 depicts these active participating states with light shading.Footnote ⁷ The remaining states, appearing in dark gray shading, did not participate in the IHP in 2019. These patterns reveal considerable state-level variation in IHP program participation.

Figure 3. State Correctional Facilities Participating in the IHP (Fiscal Year 2019)

Each dot on Figure 3 represents a city with one or more state correctional facilities that participated in the state's IHP as of 2019. Each cross represents the hearing location where judges adjudicated these cases. Sometimes, the court's hearing location was inside the state correctional facility, but other times the hearing location was an immigration court outside of the prison that was connected via videoconference to the respondents held in the state prison. In other instances, incarcerated individuals were brought to a nearby court location outside of the prison facility to attend their hearings (e.g., Reference Keenan and SeamonKeenan and Seamon 2016: 2).

The dot and cross patterns in Figure 3 highlight how participating states have advanced different models for structuring their prisons around IHP participation. Specifically, six states maintained a centralized model for the IHP: California, Georgia, Nebraska, Texas, Maryland, and New Hampshire. By dedicating bed space within one or two facilities to use for the IHP and physically housing inmates in only those centralized IHP sites, these states created prison spaces segregated by citizenship and national origin in ways that advanced immigration enforcement goals and ensured swift deportations. For example, in California all persons targeted for inclusion in the IHP in 2019 were transferred to the Calipatria State Prison located near the U.S.–Mexico border.

The remaining thirteen states adopted a decentralized model in which three or more facilities housed individuals as they participated in immigration court hearings (Figure 3). Rather than build centralized housing for the IHP, typically these decentralized state systems operated by heavy reliance on videoconferencing, which linked judges in one or more hearing sites to respondents in scattered facilities. New York, for instance, housed individuals undergoing IHP hearings in facilities located across thirty-six different cities. These findings of a patchwork system of state-level IHP participation contribute to the emerging scholarship on how state and local policy shapes federal immigration enforcement (e.g., Reference MoinesterMoinester 2018; Reference Ryo and PeacockRyo and Peacock 2020).

3.3.3 Municipal IHP

Following the 1992 protests of the brutal beating of Rodney King by Los Angeles police officers, the city responded by increasing its policing of immigrant communities (H. of Rep. 1995a: 257–61). In violation of established policy, the Los Angeles Police Department collaborated with the INS and the Los Angeles County Sheriff to charge and deport hundreds of immigrants, the majority of whom were from Mexico, on protest-related charges (ACLU of Southern California 1992: 7–9, 50). That same year, the County Sheriff became the very first sheriff to open a municipal jail to host the IHP (Reference OstrowOstrow 1992; Reference SenguptaSengupta 1992). These developments were not an anomaly for Los Angeles. As historian Kelly Hernández has shown, Los Angeles had long served as “a hub of incarceration, imprisoning more people than any city in the United States,” with a particular focus on criminalizing Mexican immigrants (Reference HernándezHernández 2017: 1, 154).

In the Los Angeles IHP pilot, a local immigration judge would travel to the jail one day per week to hear deportation cases (U.S. Senate 1994: 18–19, 69). In identifying cases for the IHP, as one IHP judge told a Senate Committee, the “intent” of the INS was to present “quick deport” cases, meaning cases that could be resolved in a single hearing (69). Immigrant rights advocates spoke up against the new program, warning that it would “speed things up so quickly that no one will know what happened to them” and put “pressure on people to sign away their rights” (Reference OstrowOstrow 1992).

Our analysis of IHP court data shows that 90% of those targeted for the Los Angeles jail program during the first three years of its operation were Mexican, and 9% were from the so-called Northern Triangle countries of El Salvador, Guatemala, and Honduras (Table 4). We also find that 99% of those in the Los Angeles IHP court were not represented by counsel, none applied for relief, and almost all cases were decided in only one hearing (Table 4). Yet, 11% of proceedings inside the county jail did not end in deportation (Table 4). The majority of these proceedings ended in transfer, suggesting that the judges simply handed contested cases over to non-IHP courts rather than resolve them within the IHP as “quick deports.” Another 8% ended in termination, suggesting that these individuals had lawful permanent resident status and the charge was not one that made them deportable, or that they were U.S. citizens erroneously held for immigration enforcement. This error rate also raises questions about whether lawful permanent residents or citizens may have been accidentally deported in the county's quick deport program. As political scientist Jacqueline Stevens has shown, “the government of the United States has been misclassifying its own citizens as aliens and deporting them for over 100 years” (Reference StevensStevens 2011).

Table 4 Descriptive Statistics for IHP Proceedings in the Los Angeles County Jail (Fiscal Years 1992–1995)

As we explore further in the next section, over the years very few municipal jails have participated in the IHP.Footnote ⁸ The lack of development of the IHP at the local level is partially due to the fact that the time individuals spend in jail is often too brief to complete the immigration court hearing process (U.S. Senate 1994: 19; Rep. Lamar Smith, H. of Rep 1997: 5). It may also reflect resistance by some localities to get involved in immigration enforcement (Reference Lasch, Linus Chan, Eagly, Haynes, Lai, McCormick and StumpfLasch et al. 2018).

3.4.1 The Rise and Post-1997 Decline of the IHP

An immigration court proceeding consists of one or more hearings that end in the decision of an immigration judge to terminate the case, grant relief, issue administrative closure, or order removal. In analyzing the court data, we find that 250,731 IHP proceedings were completed between 1980 and 2019 (Figure 4). The number of IHP proceedings rose remarkably fast through 1996, reaching a high of 19,171 proceedings completed in 1997 (EOIR CASE Data December 11, 2019).

Figure 4. IHP Proceedings, by Program Type (Fiscal Years 1980–2019)

Over the almost forty years of our study, 72% (n = 181,892 of 252,594) of all IHP proceedings occurred in state prisons. The state program was boosted by the Immigration Act of 1990, which required states to provide the INS certified records of convictions of noncitizens within thirty days of conviction (T. Alexander Aleinikoff, INS General Counsel, H. of Rep. 1995b: 8). Additional support came in 1994 when the federal government authorized funding to reimburse states and localities for incarcerating all noncitizens under the State Criminal Alien Assistance Program (VCCLEA 1994; Reference MorseMorse 2013; 8 U.S.C. § 1231(i) 2018), building on the existing program to partially reimburse states for incarcerating Mariel Cubans (Act of August 30, 1984; 49 Fed. Reg. 38,719 1984; see also Reference ClarkClark 1991).

Figure 4 spotlights a sharp overall decline in the number of IHP proceedings completed since 1997. This downward slide has been attributed to a number of factors, including insufficient staffing (GAO 1997: 1; H. of Rep. 1997: 2, 4; OIG Audit 2002: 14–15), misdirection of funds and resources away from the IHP to other INS programs (H. of Rep. 1997: 2, 4; Reference SchuckSchuck 2013: 612), the short length of some prison stays (Reference FordFord 1992; Reference RobieRobie 1992; Reference DufresneDufresne 1994; Reference CreppyCreppy 1995b; INS et al. 1996: 4), and the lack of cooperation of some state prison systems (H. of Rep. 1997: 77–79; Reference Schuck and WilliamsSchuck and Williams 1999: 407–17; Reference Lasch, Linus Chan, Eagly, Haynes, Lai, McCormick and StumpfLasch et al. 2018: 1743–45). The post-1997 fall in IHP cases is also a result of changes in the law that eliminated the right to a hearing for many incarcerated individuals. In 1994 Congress allowed for administrative removal of certain undocumented immigrants convicted of aggravated felonies, without access to a hearing before an immigration judge (VCCLEA 1994; 8 U.S.C. § 1228(b) 2018). At the same time, the number of crimes that constitute aggravated felonies has grown dramatically (Reference Chacón, Coutin, Bosworth, Parmar and VázquezChacón and Coutin 2018; Reference GoodmanGoodman 2020: 176). Also in 1994, the Immigration and Nationality Technical Corrections Act allowed for the judicial deportation of immigrants by order of a U.S. District Court judge at the time of sentencing (INTCA 1994: § 224; see also BOP 2006: 4). Two years later, Congress invented reinstatement of removal, a process by which a prior order of removal can be reissued through administrative channels without allowing for further scrutiny by an immigration judge (IIRIRA 1996; 8 U.S.C. § 1231(a)(5) 2018). Together, these programs of administrative removal, judicial deportation, and reinstatement of removal have eliminated the ability of many incarcerated noncitizens to have their cases reviewed by immigration judges (Reference FamilyFamily 2009; Reference Sivaprasad WadhiaSivaprasad Wadhia 2014; Reference KohKoh 2017), shrinking the size of the IHP (H. of Rep. 2001: 16).

Although the Trump administration tried to increase reliance on the IHP (Reference KellyKelly 2017: 3; TRAC 2017; AIC 2019), program participation declined from 3,433 completed proceedings in 2017 to 3,021 in 2019 (Figure 4). This continued reduction in IHP court proceedings under the Trump administration may be partially due to the fact that President Trump ordered that administrative procedures “shall be used in all eligible cases” in lieu of formal court proceedings (TRAC 2017).

3.4.2 Access to Counsel

Unlike in criminal court, in immigration court there is no right to appointed counsel. Thus, although noncitizens have a right to counsel, they must hire an attorney or find a pro bono volunteer to take on their case (8 U.S.C. § 1229a(b)(4)(A) 2018). Federal prisons participating in IHP proceedings are generally located in remote areas of the country, such as Oakdale, Louisiana, that are far from immigration counsel. In addition, by definition, IHP participants are incarcerated during their court proceeding. As studies on migrant detention have identified, those who remain detained during their court case rarely find lawyers (Reference Eagly and ShaferEagly and Shafer 2015; Reference RyoRyo 2016).

Representation by counsel within the IHP is dismal. Between 1988 and 2019, only 10% of IHP participants were represented by counsel at their initial case completion (n = 20,140 of 196,262).Footnote ⁹ Notably, these representation rates differed across program type. From 1988 to 2019, persons participating in the federal IHP were the least likely to have counsel at their initial case completion (only 7%, n = 4,032 of 54,710), while those in the municipal IHP (18%, n = 741 of 4,153) and state IHP (11%, n = 15,367 of 137,399) were somewhat more likely to find lawyers.

Within the federal IHP, representation rates also varied by facility type. While 10% of those in BOP facilities found counsel (n = 2,829 of 27,414), only 3.5% of those sent to for-profit contract facilities did (n = 919 of 26,194) (initial completions from 1988 to 2019). This statistic raises serious concerns regarding the availability of lawyers at for-profit CAR facilities, which are known for poor conditions and lack of programming (ACLU of Texas and ACLU 2014).

The different state IHP programs also varied widely in their level of attorney involvement (Figure 5). At the high end were Connecticut (31%) and Minnesota (30%), where almost one-third of respondents were represented by counsel. At the low end were California (0.11%) and Arizona (4%), where almost nobody found a lawyer. Although Texas had an overall representation rate of 21% (Figure 5), representation was highly concentrated in cases that ended in stipulated removal. This outcome was due to a Texas state program that appointed counsel to IHP respondents only if they agreed to stipulated removal, a topic that we address in more depth in Section 3.5.2. When stipulated removal cases were eliminated from the analysis, the Texas IHP representation rate fell to 10%.Footnote ¹⁰

Figure 5. Representation Rates in Active State IHP Programs, Initial Case Completion (Fiscal Years 1988–2019)

3.4.3 Deportation Rates

It is important to evaluate respondent outcomes within the IHP program. Research on pretrial detention in the criminal justice system has consistently found that persons who are not released prior to trial are more likely to be convicted and also more likely to receive longer sentences (Reference Scott-Hayward and FradellaScott-Hayward and Fradella 2019: 135–43). IHP participants are by definition incarcerated throughout the court process, and analysis of the IHP data shows, perhaps not surprisingly, that 93% of IHP participants were deported (n = 182,059 of 196,262 initial completions 1988–2019). This IHP removal rate was greater than the 83% removal rate for non-IHP detained removal proceedings during the same period (n = 1,626,348 of 1,948,514) (p < .001 two-tailed difference of proportions test) (see Appendix).

Despite high overall deportation rates, outcomes did vary by program type: 86% were ordered removed in the municipal IHP (n = 3,564 of 4,153), 92% in the state IHP (n = 126,901 of 137,399), and 94% in the federal IHP (n = 51,594 of 54,710) (at initial case completions 1988–2019). Also striking was the difference in removal rates between for-profit federal contract facilities and federal facilities run by the BOP. While only 4% of those placed in for-profit contract facilities avoided deportation (n = 25,189 of 26,194), 7% of IHP participants in BOP-run facilities were not deported (n = 25,476 of 27,414).

Finally, as seen in Figure 6, IHP participants in every jurisdiction were less likely to be deported when they had a lawyer. Among IHP participants, counsel was associated with the biggest reduction in removal rate for those in the municipal IHP. This outcome makes sense because those in the municipal program may have been held pretrial and thus not yet convicted, or may have had only a misdemeanor conviction that left open more possibilities to qualify for relief from removal. Greater success by counsel in municipal cases may also mean that there was a stronger pool of qualified lawyers able to take on these cases in urban jail locations (Reference Miller, Keith and HolmesMiller et al. 2015b).

Figure 6. IHP Removal Rate by Representation Status, Initial Case Completion (Fiscal Years 1988–2019)

3.4.4 Segregated Prisons

Since its inception, the IHP has been based on the creation of what Jack Shaw, Assistant Commissioner for Investigations of the INS under President Clinton, called “chokepoints” or “centralized locations” at participating correctional facilities (H. of Rep. 1995a: 15). These chokepoints relied on INS agents working inside prisons to identify individuals to place in the IHP. Prisons participating in the IHP therefore had to create the physical space within their facilities dedicated to the identification and deportation of noncitizens. For example, in New York, corrections officials built “model courtrooms in two of its male reception centers and provided new office space for use by INS personnel” (Anthony J. Annucci, Deputy Commissioner and Counsel, New York Department of Correctional Services, H. of Rep. 1997: 73). They also arranged for segregated bed space within the facility to house persons awaiting their immigration court hearings (73). Similarly, the Texas state corrections department built an entire facility dedicated exclusively to housing noncitizens for its IHP program, complete with office space for INS officials and immigration courtrooms (Catherine McVery, Assistant Director, Programs & Services Division, Texas Department of Criminal Justice, H. of Rep. 1997: 83–84).

This “chokepoint” system has meant that prison management has a great deal of influence over the IHP because the prison controls the size and location of housing capacity and directs where inmates are assigned to live. When prisons reserve beds exclusively for noncitizens to participate in the IHP, by definition they also engage in residential segregation of noncitizens targeted by the immigration agency for deportation. For example, when the Mariel Cubans were sent to the federal prison in Atlanta (Rep. John Lewis, H. of Rep. 1989: 12), the BOP chose to “consolidate this population” of noncitizens at one location, rather than have them “dispersed throughout the Federal prison system” (Kathleen Hawk, Director, Federal Bureau of Prisons, H. of Rep. 1994: 167). Before long, the BOP began to “operate several institutions in which more than half of the population consist[ed] of non-U.S. citizen inmates” (167). Within these institutions, federal prison wardens participating in the IHP were required to set aside bed space exclusively for noncitizens going through immigration proceedings (Reference CarterCarter 1996: 1; Reference McGoingsMcGoings 1998: 1).

In the most complete form of citizenship segregation, IHP locations have been created inside federal prisons that were built to hold exclusively noncitizens serving their sentences. Researchers have documented that a certain type of immigrant prison—known as known Criminal Alien Requirement or CAR prisons—began in 1999 (Reference Greene and MazónGreene and Mazón 2012: 16; Reference KaufmanKaufman 2019: 1401). Our investigation of the IHP reveals somewhat earlier start of all-noncitizen federal prisons managed by private contractors. In 1994, the Director of the BOP testified that her agency ran four contract prison facilities and all individuals held there (except for two) were noncitizens (H. of Rep. 1994: 167).Footnote ¹¹ The IHP was the glue that supported and helped to justify the creation of these early all-immigrant prisons.

Our analysis brings to light that the IHP has not only separated immigrants from the rest of the prison population, but also engaged prisons in worrisome patterns of residential segregation based on national origin and, by extension, race and ethnicity. As previously established, the original IHP experiment that took place in the early 1980s in the Atlanta penitentiary was limited to Cuban men (Table 1), many of whom were Black or Afro-Cuban (Reference StephensStephens 2016). Similarly, the Oakdale and La Tuna initiatives that followed in 1989 also segregated people by national origin. As internal DOJ documents confirm, the BOP purposefully transferred non-Cuban, non-Mexican males to the centralized Oakdale, Louisiana, location for their hearings (INS et al. 1996: 2; BOP 2002: ch. 10, 8B–9; see also Doris Meissner, INS Commissioner, U.S. Senate 1994: 46), while those identified as Mexican were sent to the La Tuna facility in Texas (Table 3).Footnote ¹² The strategic expansion of the IHP to the Donovan facility in California along the Mexican border allowed the IHP to further its goal of deporting Mexican immigrants (Table 2), or as U.S. Representative Lamar Smith put it, “illegal aliens who speak English or Spanish and who are willing to be deported” (H. of Rep. 1997: 5). At times the targeting of Latinos was even more explicit. For example, in 1997, a BOP official told members of Congress that they planned to work with the INS to establish additional IHP locations “targeted at the Mexican citizens” (John J. Clark, Assistant Director, Community Corrections and Detentions, H. of Rep. 1997: 91).

We also find that Mexican IHP participants were disproportionately placed in facilities operated by for-profit private contractors. As seen in Figure 7, 77% of IHP participants held in for-profit contract facilities were Mexican nationals (n = 20,108 of 26,194), compared to only 35% of IHP participants in BOP facilities (n = 9,688 of 27,414). These patterns are especially concerning given the subpar conditions documented in these private, for-profit facilities (Reference GreeneGreene 2001; ACLU of Texas and ACLU 2014). Also of note are clear patterns of residential and court segregation along lines of national origin, with, for example, the vast majority of Jamaicans, Nigerians, and Colombians being held in BOP facilities rather than for-profit contract facilities (Figure 7).

Figure 7. Percent National Origin by Federal IHP Type, Initial Case Completions (Fiscal Years 1988–2019)

Tables 5 and 6, which feature only those facilities associated with 250 or more IHP proceedings since 1980, delve into a more detailed examination of residential segregation within the federal prison facilities where IHP participants lived. Table 5 focuses on federal facilities run by the BOP, while Table 6 includes federal facilities run by private for-profit contractors. Here, we show once again stark patterns of segregation among participating federal prisons. In particular, while Mexicans were 52% of all participants in these IHP programs since 1980, they have been highly concentrated in select facilities. For example, 91% of all IHP participants at La Tuna have been Mexican, 90% at Reeves, 87% at Eden, and 84% at Eloy (Table 6). Cuban immigrants have been concentrated at Adams County (14%) and D. Ray James (49%) (Table 6), while excluded from Oakdale (0%) and La Tuna (1%) (Table 5). Jamaicans, who are only 4% overall of federal IHP participants, have been disproportionately sent to Allenwood (9%) and Oakdale (9%) (Table 5). This analysis reveals that, rather than integrate immigrants into the national prison system, federal prison officials—who have discretion to house migrants across the country—have crafted a residential and court system that divides participants along lines of national origin, often within private, for-profit facilities that are also entirely segregated by citizenship (Reference KaufmanKaufman 2019: 1387–1408).

Table 5 Descriptive Statistics for Federal IHP, Federal Bureau of Prisons Facilities (Fiscal Years 1980–2019)

Note: Facilities listed in Table 5 are Federal Bureau of Prisons (BOP) facilities that completed 250 or more proceedings since 1980. “All Federal” includes all federal IHP proceedings in all facilities completed between 1980 to the end of the study period that completed 250 or more proceedings. Active years include those years with a consistently larger numbers of completed proceedings. Statistical significance tests are based on a 2-tailed difference of proportions test; base category is “All Federal” excluding comparison group; all differences statistically significant at p < .05, unless indicated as non-significant (NS).

Table 6 Descriptive Statistics for Federal IHP, Federal For-Profit Contract Facilities (Fiscal Years 1980–2019)

Note: Facilities listed in Table 6 are private contract facilities that completed 250 or more federal IHP proceedings since 1980. “All Federal” includes all federal IHP proceedings in all facilities completed between 1980 to the end of the study period that completed 250 or more proceedings. Active years includes those years with a consistently larger numbers of completed proceedings. Statistical significance tests are based on a 2-tailed difference of proportions test; base category is “All Federal” excluding comparison group; all differences statistically significant at p < .05, unless indicated as non-significant (NS).

3.5.1 Videoconferencing Technology

When the IHP first began, immigration judges would generally travel to the facilities where noncitizen inmates were housed and hold court inside the prison (Alex C. Moscato, Director, EOIR, H. of Rep. 1995b: 22; Michael J. Creppy, Chief Immigration Judge, H. of Rep. 1997: 41). Sometimes, rather than having judges travel, respondents were transported from remote prison locations to a central court on the days of their hearings. As the IHP grew, however, officials sought to expand capacity and therefore looked for ways to reduce the time judges and respondents spent traveling. An early pilot program at the Marion County Jail in Oregon experimented with adjudicating IHP cases over the telephone (Cynthia Wishinsky, Director, INS Criminal Alien Branch, H. of Rep. 1995a: 282; 1997: 41). In 1992, the EOIR set in motion its first pilot project with videoconferencing technology by linking a federal BOP facility in Lexington, Kentucky, with immigration judges in Chicago (BOP 1992; Reference PetersburgPetersburg 1994; H. of Rep. 1995a: 282; GAO 1998: 5).

Reliance on videoconferencing eliminated the need for travel to conduct an in-person hearing. Instead, the immigration judge could remain in the downtown Chicago court and be connected via video link to the federal prison in Kentucky (Cynthia Wishinsky, Director, INS Criminal Alien Branch, H. of Rep. 1995a: 293; Paul W. Virtue, INS Acting Executive Associate Commissioner for Programs, H. of Rep. 1997: 36–37). Soon, other IHP courts began to experiment with videoconferencing (Reference AleinikoffAleinikoff 1995; Reference FloresFlores 1996; Reference PearsonPearson 1999).

Influenced by the success of the IHP pilot, Congress formally incorporated videoconferencing into the immigration law in 1996 (IIRIRA 1996: § 304). The technology could now be used not only within the IHP, but also more broadly within all immigration courts. At first, immigration judges limited their use of videoconferencing to initial hearings in a case, also known as “master calendar” hearings (Michael J. Creppy, Chief Immigration Judge, H. of Rep. 1997: 41). However, video gradually spread to include “individual” hearings, that is, the trial stage of an immigration case in which the merits of any motion or application for relief is adjudicated (Reference EaglyEagly 2015: 945). In 2019, the majority (56%) of all scheduled hearings within the IHP were conducted by video or telephone. Notably, in 2019 video was also used in 20% of all hearings in detained immigration courts outside of the IHP.Footnote ¹³

3.5.2 Stipulated Removal

As mentioned earlier, a desire to complete higher numbers of IHP cases drove INS officers to prioritize individuals who would not contest their deportation (U.S. Senate 1994: 18–19). Mexicans and Central Americans were targeted for the program as so-called “quick deports” (19). Senate staff observing IHP hearings in the early 1990s found that immigration judges typically spent less than five minutes of court time on each case. In these quick proceedings, the respondent would state their name, the judge would read the charges and the rights, the respondent would typically make no objections, and the judge would order the respondent deported (19).

The IHP's sustained focus on cases that could be resolved quickly was a subject of some concern by the Senate Committee on Governmental Affairs. Although INS had claimed it was removing “the worst of the worst,” the committee noted that the program was “actually a fast-track home for the ‘best of the worst’ criminal aliens” (Senator William Roth, U.S. Senate 1995: 3). In their view, cases that may have been difficult to complete before sentences ended were “excluded from the program in favor of less complicated, uncontested cases” (3).

With pressure building to boost deportation numbers, an exploratory practice quietly emerged within the IHP in Huntsville, Texas, whereby noncitizens would, through counsel, sign an agreement to waive their right to an in-person IHP court hearing (Reference HetrickHetrick 1994). The result was that individuals were ordered deported without ever coming to court so that judges could “more efficiently handle [their] caseload” (Gerald Hurwitz, Counsel to the Director of EOIR, H. of Rep. 1994: 184). During the Huntsville pilot in the early 1990s, more than half of IHP participants waived their right to a court hearing (T. Alexander Aleinikoff, INS General Counsel, H. of Rep. 1995b: 11). The Huntsville program relied on state-funded staff attorneys to facilitate these waivers of rights. Individuals who agreed to deportation without seeing the judge were represented by a staff attorney to complete all the required paperwork (11). In contrast, individuals who chose instead to go to court were not given a state-funded lawyer.

Other trial programs with stipulated removal were also underway in the early 1990s. In Florida, those charged with misdemeanors or non-violent offenses could agree to deportation. In exchange, state court prosecutors would drop their criminal charges (H. of Rep. 1995b: 11). Similarly, under a memorandum of understanding between INS and the Governor of Florida in effect as of 1995, Florida gave conditional clemency to persons convicted of nonviolent offenses if they agreed to stipulate to their deportation (41).

Building on these state experiments, the EOIR began to work with the INS to draft “a regulation permitting immigration judges to enter uncontested stipulated orders of deportation or exclusion without a hearing” (Sheila F. Anthony, Assistant Attorney General, H. of Rep. 1995a: 592; see also 59 Fed. Reg. 24,976 1994). In 1996, the procedure for stipulated removal was formally codified into the immigration law (IIRIRA 1996). Today, even individuals without counsel can be removed through stipulation without ever appearing before the immigration court (8 U.S.C. § 1229a(d) 2018; Reference KohKoh 2013: 497, 503). From 2004 to 2019, 5% (n = 3,628 of 67,305) of removal orders at an initial completion in the IHP were stipulated.Footnote ¹⁴ Importantly, stipulated removals have now spread into the non-IHP detained population of cases. During this same time period of 2004 to 2019, stipulated removals constituted 17% of all removal orders in the non-IHP detained population (n = 202,898 of 1,216,183).

Within the IHP, stipulated removals were most common in the federal IHP, where 13% (n = 2,696 of 20,530) of initial removal orders were stipulated since 2004. Moreover, stipulated removals were concentrated in three for-profit contract federal prison locations: Correctional Institution Big Spring in Big Spring, Texas (31%, n = 1,174 of 3,838), Correctional Institution D. Ray James in Folkston, Georgia (20%, n = 29 of 143), and the McRae Correctional Facility in McRae-Helena, Georgia (22%, n = 48 of 220). At the state IHP level, stipulated removals were less common during this period, constituting 13% of IHP removals in Ohio (n = 65 of 517), 10% of removals in Texas (n = 712 of 6,962), and 6% of removals in New Hampshire (n = 5 of 82).

3.5.3 Repeal of 212(c) Relief

The deportation process is generally understood to contain two stages. In the first stage, individuals are informed of the charges against them and their rights in the removal process, including the right to be represented by counsel (GAO 1990: 8; 1997: 4). A hearing is held for this purpose, during which the person charged may “immediately accept an order of deportation” (GAO 1997: 5). Alternatively, the individual may contest the ground for deportation or continue to the second stage of removal and apply for relief from deportation, including for cancellation of removal, as well as for asylum, withholding of removal, or protection under the Convention Against Torture (GAO 1990: 9; 1997: 5). If relief is sought, an evidentiary hearing will be held (Immigrant Rights Clinic, Stanford Law School 2019: 4).

In the early years of the IHP, the most common type of relief sought by respondents was under section 212(c) of the INA (Figure 8). This form of relief allowed lawful permanent residents with an aggravated felony conviction to obtain relief from removal if they had lived in the United States for seven years and had not served a sentence of five years or more for a felony (INA 1952: § 212(c); Immigration Act of 1990: § 511). Time spent in custody was counted toward the seven-year residency requirement (Jere Armstrong, Assistant Chief Immigration Judge, U.S. Senate 1994: 89; Senator William Roth, U.S. Senate 1995: 4). In 1992, just over 10% of IHP cases included a claim under 212(c) (n = 761 of 7,345) (Figure 8).

Figure 8. Applications for Relief from Removal in IHP, Initial Case Completion (Fiscal Years 1988–2018)

In a 1993 Senate hearing, Assistant Chief Immigration Judge Jere Armstrong testified that the cases of “criminal aliens” in the IHP could be quite time consuming “due to case complexity,” including the fact that 212(c) applications took time to adjudicate (U.S. Senate 1994: 65–68, 88; 1995: 28). In an effort to speed up IHP deportations, Judge Armstrong recommended that Senate members consider changes to “simplify” the deportation standard and “expedite the process” of deportation (U.S. Senate 1994: 66). Both Judge Armstrong and his colleague, Los Angeles Immigration Judge Thomas Fong, concluded that in their “personal opinion” the 212(c) law “probably need[ed] reexamining” (67–68).

In 1996, Congress responded by eliminating 212(c) eligibility for lawful permanent residents with certain types of convictions (AEDPA 1996: § 440(d)). The next year, Congress repealed 212(c) entirely (IIRIRA 1996: § 304(b)). As shown in Figure 8, applications for 212(c) declined sharply after the statute was repealed, but did not completely disappear due to a subsequent ruling of the U.S. Supreme Court that preserved the right to seek 212(c) relief for those who pleaded guilty prior to the statue's repeal (INS v. St. Cyr 2001). In its place, IHP participants began to pursue other forms of relief. By far the most common has been relief under the Convention Against Torture, followed by withholding, asylum, and cancellation of removal for lawful permanent residents (Figure 8).

3.5.4 Jail-Based Immigration Enforcement

Finally, one of the IHP's most consequential imprints has been to insert federal immigration agents into carceral institutions to screen for immigration status and expedite deportations (GAO 1998: 4). Prisons and jails participating in the IHP not only built immigration courts but also opened their door to allow INS agents to establish physical offices within their institutions and to roam their facilities looking for deportable immigrants. In Los Angeles, for example, the County Sheriff's office supplied the then-INS with a daily list of individuals in the jail allegedly born outside the United States, and INS officers would then go about investigating who might be deportable (U.S. Senate 1994: 19). These early collaborations between the INS and prisons and jails established the foundation for later innovations in what has often been termed “crimmigration,” or the merging of immigration and criminal enforcement (Reference StumpfStumpf 2006).

In 1996, as part of the Illegal Immigration Reform and Immigrant Responsibility Act, the immigration law was amended to allow state and local police departments to enter into formal agreements with the federal immigration agency to perform the functions of immigration agents (IIRIRA 1996). The resulting program, known as 287(g), borrowed directly from the IHP's signature design of conducting immigration screening inside prisons and jails. By thus growing and expanding prison- and jail-based enforcement, 287(g) and other related programs such as Secure Communities have placed “ever larger numbers of removable noncitizens—both unauthorized and authorized—in the pipeline for removal” (Reference Meissner, Kerwin, Chishti and BergeronMeissner et al. 2013: 7). Although states and localities have passed sanctuary laws and begun to question the propriety of mixing immigration enforcement with state and local criminal enforcement (U.S. Senate 1994: 22–23; Reference Lasch, Linus Chan, Eagly, Haynes, Lai, McCormick and StumpfLasch et al. 2018), the IHP has retained its foothold even within many so-called sanctuary states such as California and New York (Figure 3).