The Life and Death of Administrative Closure

In August 2020, the Executive Office of Immigration Review (EOIR)[1] proposed a new rule that would effectively eliminate administrative closure as a docket management tool for Immigration Judges. The EOIR justified this proposed rule by claiming that administrative closure has "exacerbated both the extent of the existing backlog of immigration court cases and the difficulty in addressing that backlog in a fair and timely manner." TRAC analyzed the EOIR's claims as well as the historical data on administrative closure from 1986 to present, and produced the following report. The case-by-case court records used in this report were obtained directly from the EOIR through Freedom of Information Act (FOIA) requests.

Report Highlights: Four Key Findings

TRAC's detailed analysis of the court records on administrative closure yields four key findings. First, administrative closure has been routinely used by Immigration Judges to manage their growing caseloads as well as manage the unresolved overlapping of jurisdictions between the EOIR and other immigration agencies. Second, TRAC finds that far from contributing to the backlog, administrative closure has helped reduce the backlog. Third, data from the Immigration Courts show that immigrants who obtain administrative closure are likely to have followed legal requirements and obtain lawful status. Fourth, the EOIR significantly misrepresented the data it used to justify this rule. Specifically, the agency claims to show low numbers of case completions during the Obama Administration and high numbers of case completions during the Trump Administration. In reality, the data behind this argument artificially eliminates cases that were administratively closed. Its argument also fails to recognize that average annual case completions per Immigration Judge have actually declined not increased during the past four years perhaps due to the changes introduced by this Administration. We elaborate on these findings below.

Contrasting Obama and Trump on Administrative Closure

Court records document that administrative closures have been an important case management tool used by Immigration Judges for decades. Administrative closures have allowed judges to temporarily close cases and take them off their active docket either because judges wish to focus limited resources on higher priority removal cases or because jurisdictional issues were prolonging the case. Often administrative closure is used when a case's outcome is affected by applications pending for decision before another government body, such as the U.S. Citizenship and Immigration Services (USCIS). Administrative closure does not terminate a case, it does not provide permanent relief from deportation, and it does not confer lawful status of any kind. If a case is administratively closed, either party (i.e. the DHS trial attorney or the immigrant respondent) or the Immigration Judge can at any time request that the case be placed back on the court's calendar for final resolution through the act of "recalendaring."

TRAC's analysis of Immigration Court records show that administrative closure is a long-standing practice of Immigration Judges. Case-by-case case records tracking the use of administrative closures trace this history back to the establishment of the Executive Office for Immigration Review (EOIR) in 1983[2]. Analysis of these court records document 376,439 cases when administrative closure was used to remove the proceeding at least temporarily from the court's active workload and associated hearing calendar. About 6.1 percent of the 6,147,987 cases completed by the court during this period from FY 1986 through July of 2020 were administratively closed at some point.

During the Obama Administration, administrative closure was also used to close cases through the explicit exercise of prosecutorial discretion. Originally announced in August 2011, the program sought to clear low priority cases from the Court's clogged docket to make more room to expedite decisions on higher priority cases involving "people who have been convicted of crimes or pose a security risk." As a result, an additional 88,249 cases used administrative closures based on prosecutorial discretion. These cases were concentrated between FY 2012 through early FY 2017 when President Trump assumed office.

Under the Trump Administration, the Department of Justice has attempted to restrict and eliminate the use of administrative closure in three key moves. First, as soon as Donald Trump assumed office, the Trump Administration immediately ended the practice of administratively closing low priority deportation cases. Second, in May 2018 Attorney General Jeff Sessions bypassed the Board of Immigration Appeals in the Matter of Castro-Tum[3] and decided that, despite the long history of administrative closure, Immigration Judges "lack[ed] a general authority to administratively close cases." Third, in August 2020 the Department of Justice proposed new rules in the Federal Register seeking to eliminate administrative closure altogether[4].

The EOIR's proposed rules aim to "amend the regulations to make clear that there is no freestanding authority of line immigration judges or BIA members to administratively close cases." Beyond its legal arguments, the agency justified the promulgation of these new rules with the claim that the use of administrative closures since 2012[5] had actually been counterproductive. The agency alleged that their statistics demonstrated that "as a policy matter," the use of administrative closures had "exacerbated both the extent of the existing backlog of immigration court cases and the difficulty in addressing that backlog in a fair and timely manner." EOIR further claimed that the increased use of administrative closures was correlated with a "marked decline in the productively" of immigration judges and "unquestionably exacerbated the growth in the pending caseload."

In what follows, this report analyzes the empirical evidence surrounding the use of administrative closures. It examines how the availability of this tool actually has affected the Immigration Court's backlog and the so-called "productivity" of Immigration Judges.

Use of Administrative Closures Over Time

The relative usage of regular administrative closures has varied markedly over time since FY 1986, from a low of 1 percent annually to a high of 30 percent. Prosecutorial discretion closures by themselves reached a peak of 13 percent of all closures in FY 2016 after this initiative was first announced by the Obama Administration in 2011. Figure 1-A displays the number of administrative closures by type for each fiscal year. Figure 1-B plots the corresponding percentage of all cases closed through administrative closures. (Appendix Table 1 provides the year-by-year statistics on which these graphs were based.)

Figure 1-A. Immigration Court
Use of Administrative Closures
FY 1986 - FY 2020 (through July 2020)
(Click for larger image)

Figure 1-B. Administrative Closures
as Percent of All Cases
FY 1986 - FY 2020 (through July 2020)
(Click for larger image)

A striking feature of these time series graphs is that the 2012-2017 period is not the only time when administrative closure usage climbed precipitously. There was an earlier peak back in the FY 1987-1991 period shortly after President Ronald Reagan signed the 1986 Immigration Reform and Control Act which provided a path to legalization for most undocumented immigrants who had arrived in the country prior to January 1, 1982. Indeed, in percentage terms, usage of administrative closures was higher back then than the more recent FY 2012-2017 peak.

The relative use of administrative closures to help manage the court's workload is naturally impacted by the nature of the cases before the court[6]. These, of course, vary over time. The composition of the Court's workload is also influenced by changes in the law or in court precedents which provide new grounds to regularize undocumented immigrants' status, or for gaining asylum or other relief. And, as demonstrated by differences between the Obama and Trump Administrations, the use of administrative closures can be greatly impacted by the shifting priorities of the President in office at the time.

Recalendared Cases and Their Outcome

Administrative closure of cases typically removes the matter from the Court's active workload for a considerable period of time. The average time between when a case was administratively closed and when it was recalendared was 1,296 days, or over three years. The median elapsed time—where half of cases took more time and half took less—was 659 days. Once a case was recalendared, it usually did not go to the end of the queue but was scheduled for a hearing promptly and a decision was rendered typically within just four months.

When cases were administratively closed, recalendared, and decided, most immigrants met the legal standard to remain in the country lawfully. For example, for those cases in which the government was seeking removal orders, six out of ten (60.1%) immigrants met the high legal threshold of remaining in the country. The largest proportion of these had their cases terminated since the Court ultimately found there were no longer valid grounds to deport them. Just three out of ten (30.3%) immigrants were ultimately ordered removed. In addition, one out of ten (9.6%) received a voluntary departure order[7]. See Figure 2. (Appendix Table 2 provides the numbers on which Figure 2 is based.)

Figure 2. Outcome of Recalendared Removal Cases
Previously Administratively Closed
(Click for larger image)

As of the end of July 2020, a large proportion of administrative closures had not as yet been recalendared. Indeed, some 292,042 remained closed. Of these, just 69,430—less than one-fourth—had been closed because of the exercise of prosecutorial discretion associated with Obama-era policies. In the majority, 222,612 administrative closures had taken place for other reasons. Figures 3-A and 3-B provide year-by-year breakdowns of the number of regular and prosecutorial discretion administrative closures, respectively, that have been subsequently recalendared, while Appendix Table 3 provides the supporting year-by-year counts.

Figure 3-A. Current Status of Regular
Administratively Closed Cases
FY 1986 - FY 2020 (through July 2020)
(Click for larger image)

Figure 3-B. Current Status of Prosecutorial Discretion
Administratively Closed Cases
FY 1986 - FY 2020 (through July 2020)
(Click for larger image)

If the cases that are currently administratively closed and not yet recalendared were brought back onto the Court's active docket, this would suddenly increase the Court's active workload from its current backlog at the end of July 2020 of 1,233,307 cases to 1,525,349 cases. This would produce a 24 percent jump in the court's already clogged hearing schedules, pushing the resolution of other backlogged cases off for many additional months if not years.

The actual use of administrative closures, as discussed above, clearly has not contributed to the Court's growing backlog of cases. Quite the reverse. Had administrative closures not been used, the Court's current active backlog of cases awaiting decision would certainly be at least 24 percent higher. While administratively closed cases remain closed, this also relieves some of the pressure on the Court's clogged hearing schedules. In practice, administrative closures allowed Immigration Judges to resolve other cases faster. For instance, judges could avoid unnecessary periodic status hearings while applications were pending before the U.S. Citizenship and Immigration Services and other government bodies.

EOIR Uses Flawed Statistics to Justify Its Proposed Rule

The EOIR alleged that agency statistics demonstrated that "as a policy matter," the use of administrative closures had "exacerbated both the extent of the existing backlog of immigration court cases and the difficulty in addressing that backlog in a fair and timely manner." However, the statistics the EOIR points to are themselves flawed in several ways.

First, while the table EOIR points to is titled, "New Cases and Total Completions," the completion counts arbitrarily exclude all cases involving administrative closures[8]. Readers could easily be misled since the notes at the bottom of the table which otherwise define what was included in the counts fail to mention this important detail. This approach obviously contributes to a serious undercount of actual case completions in those years where there are peaks in administrative closures, such as during the FY 2012-2017 period. Using this flawed methodology, the EOIR then points to the low counts starting in FY 2012 to argue that it was the use of administrative closures which reduced productivity and contributed to increasing case backlogs when precisely the opposite was true.

Second, there is a further problem when the table is used to draw inferences concerning changes in the Court's backlog: the EOIR doesn't count correctly. The statistics in this table allow each case to be counted only once in the "incoming" column, but allow each case to count multiple times in the "case completion" column. For example, if a case were opened, decided by an Immigration Judge, appealed, then remanded, that case would count just once in the "incoming" workload column, but twice in the "case completion" column. This alternative approach to counting gives the impression that more workload is being completed relative to the amount of incoming workload.

EOIR mistakenly tries to compare the incoming case counts with the outgoing case completion counts as if they were based on comparable enumeration rules to infer that when the completed count is greater than the incoming count the backlog is somehow being reduced. But no such conclusion can be drawn. The incoming "receipts" column ignores important components of the incoming workload.

EOIR Ignores the Impact of New Judges on Case Closures

The EOIR also claimed that the increased use of administrative closures corresponds to a "marked decline in the productively" of Immigration Judges and "unquestionably exacerbated the growth in the pending caseload." By excluding administrative closures in the agency's case completion counts, the EOIR artificially suppressed actual case closures during the latter years of the Obama Administration which gives the mistaken impression that case closures suddenly increased once Trump assumed office and the use of administrative closure was curtailed.

But there is a further problem with the agency's logic. Simply looking at the increase in case completions does not imply that judge productivity has necessarily increased. More judges have recently been hired. With more judges, more hearings can be scheduled and more cases should be decided as a natural result. To assess judge "productivity," increases in the number of judges have to be taken into consideration. EOIR failed to do this. The EOIR doesn't explicitly define what it means by judge "productivity" in this context, but in other contexts the agency relies on a simplistic metric: the number of cases a judge completes in a year[9]. The EOIR inexplicably ignores their practice of using this metric in their document.

Once the total number of Immigration Judges is taken into account, the EOIR's own measure[10] fails to provide any support for the contention that Immigration Judge productivity has risen with the changes introduced since President Trump assumed office. Quite the reverse is true. During the FY 2012 through FY 2016 period, average annual case completions per Immigration Judge "on board" at the beginning of each fiscal year averaged 737 closures per judge. Since FY 2017, the average annual case completions per Immigration Judge fell to 657. Even if we exclude FY 2020 in these calculations because case completions have been suppressed due to the partial court shutdown for the pandemic, the average annual case completions per Immigration Judge for the FY 2017-FY 2019 period was still only 674—again well below the average of 737 for the FY 2012-FY 2016 period. See Figure 4. Supporting figures year-by-year are found in Appendix Table 4.

Figure 4. Average Case Completions Per Immigration Judge by Administration,
FY 2012 - FY 2020 (through July 2020)
(Click for larger image)


Contrary to EOIR's published claims, the use of administrative closure has not contributed to the court's growing backlog of cases. Quite the opposite: administrative closure is a longstanding docket control tool for Immigration Judges and eliminating it altogether would result in a considerable rise in the number of cases in the Immigration Court backlog. Likewise, in the wake of the many changes that have been implemented since the Trump Administration assumed office (including curtailing administrative closures), Immigration Judge "productivity" overall appears to have declined, not improved, when measured as a count of annual case completions per judge. Moreover, taking administrative closure away from the Immigration Judges as a tool for managing their caseloads would disproportionately impact precisely those immigrants who are likely to obtain relief from deportation, adjust status, or otherwise obtain lawful status.

Appendix Table 1. Use of Administrative Closures
FY 1986 - FY 2020 (thru July 2020)
Fiscal Year Total Case Closures* # of Administrative Closures* Percent of Total Closures
Regular Pros. Discretion Total Regular Pros. Discretion Total
1986 32,864 6,656 - 6,656 20.3% 0.0% 20.3%
1987 52,190 15,464 - 15,464 29.6% 0.0% 29.6%
1988 75,093 16,500 - 16,500 22.0% 0.0% 22.0%
1989 94,962 22,731 - 22,731 23.9% 0.0% 23.9%
1990 95,828 19,432 - 19,432 20.3% 0.0% 20.3%
1991 102,786 11,445 - 11,445 11.1% 0.0% 11.1%
1992 89,083 7,740 - 7,740 8.7% 0.0% 8.7%
1993 92,989 7,716 - 7,716 8.3% 0.0% 8.3%
1994 109,026 6,616 - 6,616 6.1% 0.0% 6.1%
1995 140,628 6,036 - 6,036 4.3% 0.0% 4.3%
1996 184,703 9,492 - 9,492 5.1% 0.0% 5.1%
1997 193,678 8,949 - 8,949 4.6% 0.0% 4.6%
1998 190,979 9,163 - 9,163 4.8% 0.0% 4.8%
1999 166,387 10,126 - 10,126 6.1% 0.0% 6.1%
2000 165,123 5,571 - 5,571 3.4% 0.0% 3.4%
2001 162,770 6,256 - 6,256 3.8% 0.0% 3.8%
2002 166,121 7,504 - 7,504 4.5% 0.0% 4.5%
2003 186,400 6,968 - 6,968 3.7% 0.0% 3.7%
2004 200,574 6,567 - 6,567 3.3% 0.0% 3.3%
2005 257,839 5,542 - 5,542 2.1% 0.0% 2.1%
2006 264,956 6,527 - 6,527 2.5% 0.0% 2.5%
2007 211,923 7,589 1 7,590 3.6% 0.0% 3.6%
2008 221,423 7,972 - 7,972 3.6% 0.0% 3.6%
2009 231,235 7,457 - 7,457 3.2% 0.0% 3.2%
2010 225,851 8,564 - 8,564 3.8% 0.0% 3.8%
2011 229,971 5,988 8 5,996 2.6% 0.0% 2.6%
2012 211,271 8,580 9,036 17,616 4.1% 4.3% 8.3%
2013 192,677 16,401 15,299 31,700 8.5% 7.9% 16.5%
2014 199,881 20,896 12,796 33,692 10.5% 6.4% 16.9%
2015 185,228 27,761 17,543 45,304 15.0% 9.5% 24.5%
2016 197,054 27,389 24,974 52,363 13.9% 12.7% 26.6%
2017 210,541 22,868 8,491 31,359 10.9% 4.0% 14.9%
2018 230,759 9,895 79 9,974 4.3% 0.0% 4.3%
2019 320,495 503 7 510 0.2% 0.0% 0.2%
2020 254,699 1,575 15 1,590 0.6% 0.0% 0.6%
6,147,987 376,439 88,249
* Case closure counts cover the initial case closure to avoid counting the same case more than once. Similarly, administrative closure counts are limited to the initial administrative closure in the case.
Appendix Table 2. Outcome of Recalendared Cases by Major Case Types
FY 1986 - FY 2020 (through July 2020)
Case Type = Removal* Case Type = Deportation* Case Type = Exclusion*
Outcome Cases Percent Outcome Cases Percent Outcome Cases Percent
Allowed to Remain 59,650 60.1% Allowed to Remain 24,088 55.4% Allowed to Remain 2,297 45.7%
Termination 43,764 44.1% Termination 11,005 25.3% Termination 1,171 23.3%
Relief 15,468 15.6% Relief 12,717 29.2% Admit 423 8.4%
Other 418 0.4% Other 366 0.8% Other 703 14.0%
Deported 39,677 39.9% Deported 19,404 44.6% Deported 2,724 54.3%
Removal Order 30,095 30.3% Deported 13,714 31.5% Exclude 2,724 54.3%
Voluntary Departure 9,582 9.6% Voluntary Departure 5,690 13.1% Voluntary Departure na na
TOTAL 99,327 100.0% TOTAL 43,492 100.0% TOTAL 5,021 100.0%
* Until April 1, 1997 the two major types of cases adjudicated by Immigration Courts were Exclusion and Deportation cases. Provisions of the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 replaced these two with Removal as the case type for subsequent cases.
na - not applicable
Appendix Table 3. Current Status of Administratively Closed Cases by Type
FY 1986 - FY 2020 (through July 2020)
Fiscal Year Regular Administrative Closures Pros Discretion Admin Closures*
Cases Still Closed Percent Cases Still Closed Percent
1986 6,656 4,260 64% - -
1987 15,464 11,863 77% - -
1988 16,500 12,713 77% - -
1989 22,731 16,227 71% - -
1990 19,432 13,447 69% - -
1991 11,445 7,821 68% - -
1992 7,740 4,981 64% - -
1993 7,716 4,760 62% - -
1994 6,616 3,928 59% - -
1995 6,036 3,492 58% - -
1996 9,492 5,643 59% - -
1997 8,949 5,467 61% - -
1998 9,163 5,896 64% - -
1999 10,126 6,643 66% - -
2000 5,571 3,113 56% - -
2001 6,256 3,461 55% - -
2002 7,504 3,951 53% - -
2003 6,968 2,899 42% - -
2004 6,567 2,626 40% - -
2005 5,542 2,157 39% - -
2006 6,527 2,659 41% - -
2007 7,589 3,398 45% - -
2008 7,972 3,537 44% - -
2009 7,457 2,894 39% - -
2010 8,564 4,547 53% - -
2011 5,988 2,572 43% 8 5 63%
2012 8,580 4,289 50% 9,036 6,402 71%
2013 16,401 8,699 53% 15,299 11,555 76%
2014 20,896 12,066 58% 12,796 9,996 78%
2015 27,761 15,227 55% 17,543 13,989 80%
2016 27,389 16,036 59% 24,974 20,525 82%
2017 22,868 13,044 57% 8,491 6,888 81%
2018 9,895 6,368 64% 79 50 63%
2019 503 399 79% 7 5 71%
2020 1,575 1,529 97% 15 15 100%
* Initiative began in 2011 and was discontinued after President Trump assumed office.
Appendix Table 4. Number of Case Completions per Immigration Judge
FY 2012 - FY 2020 (through July 2020)
Fiscal Year Case Completions* Immigration Judges Case Completions Per Immigration Judge
2012 205,068 273 751 Obama Average: 737
2013 189,330 267 709
2014 176,749 262 675
2015 190,172 249 764
2016 197,472 254 777
2017 195,753 289 677 Trump Average: 657*
2018 206,442 338 611
2019 283,217 395 717
2020 271,936 442 615
* Because of the partial court shutdown during the pandemic, FY 2020 figures are unusually low. Excluding FY 2020, the annual average case completions per immigration judge for FY 2017-2019 was 674.
Notes: Case completions identical to EOIR statistics but then add back administrative closures. Immigration Judges on board based upon EOIR published numbers at the beginning of each fiscal year. FY 2020 case completions annualized based upon first ten months.


[1] The Executive Office for Immigration Review (EOIR) is the agency within the Department of Justice that manages and oversees the U.S. Immigration Court system. It is responsible for managing immigration proceedings, which are civil legal matters that begin when the Department of Homeland Security believes that an individual is in the country unlawfully or in violation of their visa. The EOIR also oversees the Board of Immigration Appeals (BIA), an appellate-level body that adjudicates appeals from Immigration Judges' decisions.

[2] Given the many changes in recording practices and database technology over the ensuing years, some caution should be exercised in the use of records for early time periods as the likelihood that records become lost or are dropped increases the farther back in time one goes. But even if these early records represent an undercount of what actually took place, they document extensive use of administrative closures even during this early period.

[3] Matter of Castro Tum, 27 I&N Dec. 187 (A.G. 2018)

[4] On August 26, 2020, the EOIR "propose[d] to amend the regulations to make clear that there is no freestanding authority of line immigration judges or BIA members to administratively close cases."

[5] EOIR points to the BIA decision in Matter of Avertisyan, 25 I&N Dec. 688 (BIA 2012) as triggering the rise in administrative closures. However, there were other forces at play. In addition to the implementation of prosecutorial discretion closures at this time, the court's caseload was also impacted by a growing wave of asylum seekers—including many unaccompanied children—that started arriving in 2012 along the southwest border with Mexico. For example, while cases involving unaccompanied children arriving from countries beyond Mexico went to the Immigration Court, judges there had to wait for these children's asylum applications and other claims to allow them to stay in the U.S. to be decided by the USCIS and sometimes other government bodies before they could legitimately decide whether removal was warranted.

[6] See earlier footnote 5 concerning the wave of asylum seekers that started arriving along the country's southwest border with Mexico starting in 2012.

[7] Voluntary departure allows immigrants who are determined to be in the country unlawfully to depart the country on their own without a removal order.

[8] EOIR also excluded cases that are classified not as "administrative closures" but as "other completions."

[9] Each judge now has a dashboard that displays a cumulative count of its "case closures" so this can be compared against the productivity standard set by the agency. To get a "satisfactory" rating on their performance evaluations, judges will be required to clear at least 700 cases a year. There are several problems with this simple metric. Significant parts of a judge's workload, such as conducting bond hearings or deciding some types of cases such as credible fear and reasonable fear cases are apparently excluded from these counts. Further, no consideration is given to the nature of the cases a judge is assigned to hear. Because the nature of cases vary, the time a judge needs to spend to dispose of a case can vary markedly. The federal court system has developed a system which, rather than counting each case the same, weights cases by the time required on average to handle that particular type of case. Clearly, Immigration Judges require on average more time to dispose of their workload where there are a large proportion of cases involving applications for asylum or other forms of relief than cases where many immigrants present no defense at their initial master calendar hearing and concede removability. Yet none of this is taken into consideration in EOIR's measure of a judge's "productivity."

[10] Since EOIR based the statistics it cites on total completions counting "initial case completions plus subsequent case completions," this approach is also used in computing these IJ "productivity" measures. Case completions per judge would naturally be somewhat lower if double counting of completions when cases were reopened or remanded were disregarded, and only the latest (or just the earliest) case completion was counted.

TRAC is a nonpartisan, nonprofit data research center affiliated with the Newhouse School of Public Communications and the Whitman School of Management, both at Syracuse University. For more information, to subscribe, or to donate, contact or call 315-443-3563.