June 14, 2011

IV. Data on Detainee Transfers

Number, Gender, and Nationality of Transferred Detainees

The data represent all 2,271,911 non-citizens held in ICE detention between October 1, 1998 and April 30, 2010. These non-citizens account for 2,869,323 episodes of detention, as some individuals were detained by ICE multiple times. The data show that 1,159,568, or 40 percent of all detainees, experienced at least one transfer during their detention (they are referred to here as “transferred detainees”). Over 46 percent of transferred detainees (505,787 detainees) were transferred two or more times. We found that 16.7 percent of transferred detainees experienced three or more transfers, with over 3,400 detainees experiencing 10 or more transfers. One detainee was transferred between facilities 66 times.

Most (1,045,620) transferred detainees were male, 113,814 were female, and 134 were listed as unknown gender. Most transferred detainees were from Mexico (511,945), Honduras (133,062), and Guatemala (131,691). Shockingly, 38 were from the United States (2 females, 36 males). Why so many of those transferred were listed as being United States nationals is beyond the scope of our analysis. However, we note that previous analysis performed by Human Rights Watch on ICE datasets has revealed serious problems with ICE data management.[27]Moreover, our recent research into the experiences of detainees with mental disabilities revealed several troubling cases of US citizens who were kept in immigration detention for years, and experienced multiple transfers during their time in detention, despite the fact that their US citizenship should have negated any ICE involvement.[28] 

Transfers over Time

As shown in Figure 1, below, transfers have increased steadily over time. Cumulatively, between the beginning of fiscal year 1999 and April 2010, 41 percent of all detention episodes included at least one transfer between facilities. In fiscal year 1999, 23 percent of detention episodes included one or more transfers between facilities, but in fiscal year 2009, 52 percent of detention episodes included one or more transfers. Figure 2 illustrates the percentage of detention episodes experiencing transfer between 1998 and 2010. Table 1 shows the total number of transfer movements for each year between 1998 and 2010.

Figure 1

  

Figure 2 

 

Table 1 – Total Transfer Movements by Fiscal Year

In February 2009 ICE wrote to Human Rights Watch to state its intention to minimize detainee transfers. As shown in Figure 3, below, other than a peak in March 2010, which also saw an overall increase in detention episodes, the number and rate of detention episodes utilizing transfers has decreased slightly since ICE stated its intention to reduce transfers in February 2009. While this may show an informal commitment to reducing the practice, there have been no official policy reforms aimed at reducing transfers. This very slight decrease is therefore unlikely to indicate a continuing trend.

Figure 3

 

Geographical Distances Covered

The dataset contained a total of 2,040,103 movements of detainees between detention facilities (referred to here as “transfer movements”), corresponding to 1.07 million individual transferred detainees.[29] Of these transfer movements, 564,209 (27 percent) were interstate.

Perhaps the most important finding of our data analysis is the ongoing use of long-distance transfers for detainees. As Table 2 below shows, transfers between Pennsylvania and Texas were thethird most frequent type of transfer used, requiring detainees to travel 1,642 miles. In addition, large numbers of detainees were moved between North Carolina and Georgia, between Pennsylvania and Louisiana, and between southern California and Arizona. Such long-distance transfers cannot realistically be accomplished without use of an airplane. The 99 longest transfer segments originated or ended in Guam, Hawaii, Alaska, or Puerto Rico. The longest continental US transfers occurred between Florida and Washington.

Table 2 — Ten Most Frequent Interstate Transfer Movements[30]

Since many detainees were transferred multiple times during each period spent in detention, transfers are best counted and measured per detention episode. For all transferred detainees, the average distance transferred per detention episode was 369.81 miles. As shown in Figure 4, there is what approximates a bell curve when examining distance as related to the number of transfers, where distance traveled seems to peak around 10 transfers. It is unknown why average distance moved decreases when detainees experience more than 10 transfers.  It is possible that when detainees experience this number of transfers, they are being frequently transferred over short distances.

Figure 4

Table 3 – Top 10 Facilities Utilizing Transfers

 

Originating Facility

 

 

State

 

 

Circuit

 

Total Transfer Actions

# of Facilities Transferring To

Total Outbound Transfers

# of Facilities Receiving From

Total Inbound Transfers

Los Custody Case Holding Facility, SAN PEDRO

CA

9th Cir.

283,218

301

150,200

251

133,018

FLORENCE STAGING FACILITY

AZ

9th Cir.

205,673

212

120,744

271

84,929

HARLINGEN STAGING FACILITY

TX

9th Cir.

179,190

181

112,890

197

66,300

MIRA LOMA DET. CENTER

CA

9th Cir.

119,118

97

56,433

94

62,685

FLORENCE SPC

AZ

9th Cir.

111,553

141

28,508

150

83,045

YORK COUNTY JAIL

PA

3rd Cir.

100,776

184

42,869

225

57,907

WILLACY COUNTY DETENTION CENTER

TX

5th Cir.

97,040

66

29,585

102

67,455

LAREDO CONTRACT DET. FACILITY

TX

5th Cir.

76,069

126

59,627

165

16,442

SOUTH TEXAS DETENTION COMPLEX

TX

5th Cir.

74,023

92

17,780

194

56,243

HOUSTON CONTRACT DET. FACILITY

TX

5th Cir.

71,456

141

25,507

314

45,949

Table 4 shows the variation between the states that received transfers and those from which transfers originated. Only Pennsylvania, Texas, New York, and Alabama were in the top ten states for both receiving and sending transfers. Louisiana received 19 percent while California sent 12 percent of interstate transfers.

Table 4 – Top Ten Originating/Receiving States for Detainee Transfers

Originating State

 

# of Tranfers

% of Originating Transfers

Receiving State

 

# of Transfers

% of Receiving Transfers

CALIFORNIA

70,002

12%

LOUISIANA

104,703

19%

PENNSYLVANIA

42,166

7%

ARIZONA

79,030

14%

NEW YORK

37,978

7%

TEXAS

77,784

14%

NORTH CAROLINA

35,299

6%

GEORGIA

48,383

9%

TEXAS

28,452

5%

PENNSYLVANIA

35,827

6%

ALABAMA

26,689

5%

WASHINGTON

27,172

5%

OREGON

23,751

4%

ALABAMA

26,765

5%

NEW JERSEY

22,923

4%

NEW MEXICO

23,158

4%

FLORIDA

22,043

4%

NEW YORK

15,646

3%

LOUISIANA

19,146

3%

ILLINOIS

15,160

3%

Intra- and Inter-Federal Circuit Court Transfers

Most transfers (84.8 percent) were between facilities within the jurisdiction of the same Federal Circuit Court of Appeals (which defines the applicable law for each detainee’s case).   For example, of the 483,425 transfers originating in the Fifth Circuit, 93 percent were to other Fifth Circuit facilities. As shown in Table 5, most circuits originate and receive about the same percentage of transfers.

However, when examining only interstate transfers, shown in Figure 5 below, we find the Fifth Circuit receives, by a large margin, the most interstate transfers. Originating circuits are color-coded.

Table 5 – Transfers Originating in, and Received by, Each Circuit Court

Figure 5 – Interstate Transfers by Circuit Court

Table 6 below shows that the Fifth Circuit is the federal circuit court district with the worst ratio in the country (510:1) for immigration attorneys to received transferred detainees (as measured by the number of attorneys in the circuit who are members of the American Immigration Lawyers Association).

Table 6 – Ratio of Attorneys to Transferred Detainees by Federal Circuit Court Districts

 

Circuit

 

 

Rank by Transferred Detainee to Attorney Ratio

 

Received Transfers 1998 – 2010

 

 

AILA Members as of May 2011

 

 

Transferred Detainee to AILA Member Ratio

 

5th

12

483,457

947

510.51

9th

11

760,606

2674

284.45

10th

10

90,898

424

214.38

3rd

9

126,855

634

200.09

11th

8

187,275

1242

150.79

8th

7

58,927

455

129.51

4th

6

104,738

812

128.99

6th

5

73,624

646

113.97

7th

4

47,177

633

74.53

1st

3

40,143

557

72.07

2nd

2

66,287

1576

42.06

D.C.

1

89

307

0.29

Source for AILA membership numbers: email from Amanda Walkins, Member Outreach Associate, American Immigration Lawyers Association to Human Rights Watch, May 2, 2011.

Facility Type

More than half of all transfers involved a facility that has an Intergovernmental Service Agreement with ICE to hold immigration detainees. These facilities are most commonly state or local criminal jails and prisons, intended to house people awaiting criminal trial or persons serving criminal punishments. After analyzing transfers by facility type, Table 7 shows that the breakdown of facilities involved in transfers is almost identical between originating and receiving facilities.

Table 7 – Number of Transfers by Facility Type

Length of Detention

The length of detention was determined using the dates on which detainees entered (were booked into) and left (were booked out of) detention. Individuals that experienced transfers were held on average over three times as long as those that were never transferred, either measured by the mean or median days in detention, as illustrated in Table 8.[31] 

Table 8 – Length of Detention for Transferred and Never-Transferred Detainees

Female detainees (median: 13 days, mean: 33 days) and male detainees (median: 14 days, mean: 39 days) were held in detention for similar lengths of time. When examining length of detention by nationality, Table 9 shows that there were clear differences, as citizens of countries such as Vietnam and Haiti were held for over five times as long as Mexican nationals. This is likely a result of diplomatic and humanitarian problems causing delays in deportations to those countries.

Table 9 – Length of Detention by Country

Deportation or Termination of Detention for Transferred Detainees

The dataset contained a variable coded as “release reason,” which described the reason for each detainee’s departure from immigration detention. Of the 2,271,911 detainees contained in the dataset (both transferred and never transferred), 62 percent, or 1,413,500, were ultimately deported.[32]               

The next most common reason for the termination of detention was voluntary departure, in which 343,557 people agreed to leave the US voluntarily. Another 421,538 people were released to undergo their immigration court proceedings outside of the confines of detention, either on bond, on an order of recognizance, or on an order of supervision.[33]

Some 44,110 people had their cases terminated. Many things can lead to termination of a case. One documented in our previous research is the termination of the cases of people with mental disabilities who have been transferred multiple times between detention facilities over many months or years. Ultimately, some of these cases are terminated by judges who decide they cannot continue with the deportation of someone with serious disabilities.

For another 33,439 detainees, their odyssey in immigration detention did not end during the time period. This group lacked a book-out date, or release code, or had a final movement recorded as another transfer. Of this group, 35 percent had already experienced at least one transfer during their current stay in immigration detention.

Finally, 206 immigrant detainees died, and 940 escaped from immigration detention.[34]

As shown in Table 10, detainees who were never transferred had more favorable reasons for their release from detention than those who experienced one or more transfers. Among detainees who were never transferred, 54 percent were ultimately deported, whereas 74 percent of transferred detainees were deported. In addition, a larger percentage of immigrant detainees who were never transferred (16 percent) were released on orders of voluntary departure, as compared with 6 percent of those who were transferred.

Finally, a larger percentage of detainees who were never transferred (16 percent) as compared with those who were transferred (14 percent) benefitted from bond or other forms of release from detention while their immigration court proceedings were still ongoing. The ability to remain near communities of support may help explain why more detainees who were never transferred were able to benefit from release from detention while their court cases were still underway. Judges may only order release for persons who are not considered at risk of absconding from proceedings, and this is often proved through strong ties to the community, which transferred detainees rarely have in their post-transfer locations.

Table 10 – Release Reason for Transferred Detainees Compared with Never Transferred[35]

Release Reason

 

Never Transferred

% of Never Transferred

Experienced One or More Transfers

% of Transferred

Deported/Removed

910,818

54%

840,254

74%

Voluntary departure

274,685

16%

68,872

6%

Released with Proceedings Ongoing

270,074

16%

151,464

14%

U.S. Marshals or Other Agency

87,832

5%

25,580

2%

Paroled

64,748

4%

17,144

2%

Outstanding Detention Hanging Closure

39,365

2%

12,469

1%

Proceedings Terminated

25,241

1%

18,869

2%

Withdrawal

11,443

1%

747

0%

Cost Analysis

ICE provides no publicly available analysis of the savings or costs associated with detainee transfers. There is no public accounting for the costs of bed space in every part of the country in which ICE operates or subcontracts for detention space. The agency also does not provide information on the rationales for transfers in particular cases, which might help the agency and others to better understand the savings or costs associated with its practices.

For example, although none of the detainees interviewed for our 2009 report Locked Up Far Away had been transferred for medical reasons, it is certainly the case that some percentage of transfers are completed in order to provide immigrant detainees with necessary medical care, and that providing such care prevents illness, loss of life, and costly lawsuits. However, there is no way to estimate these savings since the agency does not make public, or even record in a centralized database, the reasons for detainee transfers.[36]  

Therefore, while we have no way to estimate savings associated with transfers, we can roughly estimate some transportation costs associated with transfers, based on information provided to Human Rights Watch by the US Marshals Service and the IRS. Two cost estimates were used, assuming air travel was used for transfers greater than 475 miles and ground transportation used for transfers less than 475 miles. Further details on the information used for these calculations can be found in the Methodology section.

According to our estimates, the over two million transfers that occurred between October 1998 and April 2010 cost approximately $366,832,842 in total. We believe that these transport costs only represent a fraction of the total costs of transfers: since transferred detainees spend on average more than three times longer in detention than those who are not transferred, the most significant financial costs may come from court delays and unnecessarily long periods of detention.

As illustrated in Table 11, the most costly transfer “segment” has been from York County Jail in Pennsylvania to Harlingen Staging Facility in Texas. Over 11,000 transfers have been sent the 1,642 miles from Pennsylvania to Texas, costing an estimated $13.2 million. The most costly single transfer movements have occurred between Guam and the continental US.  These transfers are nearly 8,000 miles long and likely cost several thousand dollars per detainee. Figure 6 provides the costs of transfers originating from each of the 50 states.

Table 11 – Most Costly Transfer Segments

Figure 6

As noted above, our conservative estimate of $366 million dollars for detainee transfers between 1998 and 2010 does not include other costs that may be associated with transfers, such as flights made by carriers more costly than the Justice Prisoner and Alien Transportation System (JPATS); personnel time spent on paperwork or other administrative tasks; costs of additional court time, court delays, or lengthened detention caused by transfers; costs associated with needless transfers of persons who are found to be eligible for bond and therefore are unnecessarily detained; or costs associated with duplicative medical screenings or tests. Therefore, without better public information on ICE’s operational budget related to transfers, it is impossible to conclude whether transfers result in net costs or savings for the agency.

 

[27]Human Rights Watch, US: Forced Apart (By the Numbers): Non-Citizens Deported Mostly for Nonviolent Offenses, April 15, 2009, http://www.hrw.org/en/node/82159/section/6.

[28]Human Rights Watch, Deportation by Default: Mental Disability, Unfair Hearings, and Indefinite Detention in the US Immigration System, July 25, 2010, http://www.hrw.org/en/reports/2010/07/26/deportation-default-0.

[29] We are not including within our total transfer movements the 99,687 movements identified in the dataset as transfers, but which were actually intra-facility movements in which the detainee never left a particular detention facility. 

[30] If two facilities were in the same city and we could not distinguish separate longitude and latitude values, we used a proxy distance estimate of 1 mile to estimate the distance between facilities.

[31] The mean is our common perception of average. The median is the value in the middle of a data set. When there are outliers, they will drive the mean up or down but will have less effect on the median. Both numbers tell a story regarding length of detention. When the mean is higher than the median, it can result from a few people that were held for a long time, thus driving the mean higher. But it can also indicate that many people were held for a short time (i.e., 0-1 days), which may result in a median that is lower than the mean.

[32]Some people were deported multiple times: 1,413,500 deportees experienced 1,754,443 deportations.

[33]The ICE District Director has discretion to issue an “order of release on recognizance or supervision,” which releases an immigrant from detention, subject to certain conditions as determined by the Director, such as regular reporting to the ICE district office. See Immigration and Naturalization Act, Section 236.

[34] While there has been significant press coverage of the deaths of immigrant detainees, including information provided by ICE itself, we were unable to locate any public information provided by ICE nor any press reports documenting instances of escape from immigration detention.

[35]This table only includes the most common release reasons and excludes cases with data entry issues ( ~0.3 percent).

[36]Even if one accepts the notion that transfers for medical care provide cost savings to the agency, it is also true that transfers for medical care are not adequately addressing detainee medical needs, since ICE’s failure to care for the medical needs of non-citizen detainees (resulting in deaths in several cases) has been the subject of numerous lawsuits, press investigations, and congressional action. See, e.g., “ACLU Sues U.S. Immigration Officials and For-Profit Corrections Corporation Over Dangerous and Inhumane Housing of Detainees,” ACLU press release, January 24, 2007, http://www.aclu.org/prison/conditions/28127prs20070124.html (accessed May 11, 2011); “ACLU Sues Over Lack of Medical Treatment at San Diego Detention Facility,” ACLU press release, June 13, 2007, http://aclu.org/immigrants/detention/30095res20070613.html (accessed May 11, 2011); Dana Priest and Amy Goldstein, “Careless Detention: System of Neglect,” Washington Post, May 11, 2008, http://www.washingtonpost.com/wp-srv/nation/specials/immigration/cwc_d1p1.html (accessed May 11, 2011); “In Custody Deaths,” New York Times, http://topics.nytimes.com/top/reference/timestopics/subjects/i/immigration_detention_us/incustody_deaths/index.html (accessed May 11, 2011) (collecting articles published by the New York Times about immigrant detainee deaths and failure to provide medical care from 2005 to 2009); U.S. House of Representatives, Committee on the Judiciary, Subcommittee on Immigration, Citizenship, Refugees, Border Security, and International Law, “Hearing on Problems with Immigration Detainee Medical Care,” June 4, 2008, http://judiciary.house.gov/hearings/hear_060408.html (accessed May 11, 2011).