Skip navigation

Vera Institute of Justice Evaluation Performance and Outcome Measurement of Usdoj Eoir Legal Orientation Program May 2008

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.
LEGAL ORIENTATION PROGRAM
Evaluation and Performance and Outcome
Measurement Report, Phase II

Nina Siulc
Zhifen Cheng
Arnold Son
Olga Byrne
Vera Institute of Justice
May 2008

© 2008 Vera Institute of Justice. All rights reserved. Prepared by the Vera Institute of
Justice’s Center on Immigration and Justice for the Executive Office for Immigration
Review, in fulfillment of Task Order DJJ­6F­EIR01­0220 T0007.
Additional copies can be obtained from the communications department of the Vera
Institute of Justice, 233 Broadway, 12th floor, New York, New York, 10279,
(212) 334­1300. An electronic version of this report is available for download on Vera’s
web site, www.vera.org.
Requests for additional information about the research described in this report should be
directed to CIJcoordinator@vera.org.

Vera Institute of Justice ii

Executive Summary
Since 2003 Congress has funded the Executive Office for Immigration Review (EOIR) of the
U.S. Department of Justice to administer the Legal Orientation Programs (LOP). The LOP seeks
to educate detained persons in removal (deportation) proceedings so they can make more
informed decisions, thus increasing efficiencies in the immigration court and detention
processes. The LOP provides detained persons with basic information on forms of relief from
removal, how to accelerate repatriation through the removal process, how to represent
themselves pro se, and how to obtain legal representation. The LOP is designed to provide this
information to detained persons prior to the first hearing in their removal proceedings before
EOIR (the “immigration courts”). The LOP is offered nationally by nonprofit legal service
providers who work collaboratively with local immigration courts, detention facilities, and
Immigration and Customs Enforcement (ICE).
The LOP involves four levels of service:
•	 Group orientations are presentations by attorneys or paralegals (under attorney
supervision) that offer a broad overview of the immigration court process and basic
information on relief from removal or ways to expedite the removal process.
•	 Individual orientations are one­on­one meetings generally following the group
orientation. In these meetings, detainees ask LOP attorneys and paralegals more detailed
questions about process, specific defenses, or forms of relief from removal.
•	 Self­help workshops are small workshops led by LOP staff for detainees who will be
handling their cases pro se. In these workshops, individuals can prepare and practice with
other persons who will be pursuing similar defenses or applications for relief from
removal.
•	 Referrals to pro bono attorneys are made for some indigent detainees who are unable to
proceed pro se or whose cases could benefit from the assistance of legal representation.
Since 2005, EOIR has contracted with the Vera Institute of Justice (Vera) to manage the
LOP. Vera subcontracts to nonprofit organizations to provide LOP services, and Vera staff
monitor, oversee, and measure the performance of the program. The contract also required Vera
to implement a Performance Outcome and Measurement Plan and undertake a program
evaluation to document LOP services, assess if the LOP is working as intended, determine any
impact of the program and the significance and extent of any impact, and make
recommendations for ongoing program improvements. This report summarizes research
activities and findings as of September 2007.
Highlights from the LOP Evaluation

In 2006, the LOP reached more than 25,500 detainees. From the program’s inception in 2003
through September 2007, the program has reached more than 100,000 detained persons. As the
Vera Institute of Justice iii

use of detention—and bed space in many of the facilities hosting the LOP—has expanded, the
program has continued to serve more people each year. However, as the expansion of detention
has outpaced the expansion of funding for the Legal Orientation Program, the numbers of people
receiving LOP services represents a shrinking percentage of the overall detained immigration
court population each year.
Vera’s analysis—a combination of statistical analysis and interviews with LOP stakeholders,
including participants, providers, immigration judges, court administrators, detention facility
staff, and ICE employees—identified numerous differences in case outcomes between LOP
participants and “comparison groups” of detained persons who did not participate in the LOP.
These differences, described below, suggest possible benefits of the LOP for those detained
persons it is able to serve.
LOP participants move through the courts faster
Detained LOP participants have immigration court case processing times that are an average of
13 days shorter than cases for detained persons who did not participate in the program. This
suggests that the LOP may have important resource­saving benefits for the immigration courts
and immigration detention system. The faster detained cases are completed, the sooner detained
persons are eligible to be released from custody or removed from the United States. This can free
available bed space at detention facilities and, at least in theory, substantially reduce costs for the
federal government.
LOP participants receive fewer in absentia removal orders
Nationwide, very few detained persons are released on bond or recognizance. However, when
released from detention prior to the completion of their immigration court cases, LOP
participants received 7 percent fewer in absentia removal orders, meaning that they appeared for
court hearings at greater rates than comparison groups, especially when pursuing relief from
removal.1 Low rates of in absentia removal orders were even more pronounced for LOP
participants who received intensive levels of LOP service (meaning they participated in more
than group orientations). Immigration court and detention system stakeholders are concerned
with reducing the numbers of persons who receive in absentia removal orders. Our analysis of
the LOP supports conclusions from studies of other court systems that when respondents have
access to legal information and understand the court process, they are less likely to receive in
absentia removal orders.
The LOP can effectively prepare detained respondents to proceed pro se
The LOP is not a substitute for legal representation. However, some detained persons who
received intensive LOP services (more than group orientations) and represented themselves pro
se achieved case outcomes approximating those associated with legal representation. LOP
1

An in absentia removal order occurs when a person fails to appear in immigration court, provided the government
shows that the person is removable and that required procedures occurred.

Vera Institute of Justice iv

participants who represented themselves pro se were also more likely to receive grants of
voluntary departure than detainees who did not participate in the LOP.2
Detention facility staff state that the LOP improves detention conditions
Detention facility employees at LOP sites reported that they have observed a reduction in
behavior problems when detainees have access to legal information. They also stated that the
LOP makes detention “safer” and “more humane.”
Immigration judges state that the LOP increases immigration court efficiency
Immigration judges at LOP sites report that respondents who have attended the LOP appear in
immigration court better prepared, are more likely to be able to identify the relief for which they
are statutorily eligible, to not pursue relief for which they are ineligible, and to have a better
understanding of the immigration court process, thus helping to improve court efficiencies.

2

Voluntary departure is a procedure that allows an individual to leave the United States, usually within 60 or 120
days, at his or her personal expense, thus avoiding some of the negative consequences of a removal order.
Vera Institute of Justice v

Table of Contents
Introduction................................................................................................... 7

History of the Legal Orientation Program ................................................. 7

Measuring Performance and Impact of the LOP ..................................... 11

Legal Orientation Program Services............................................................... 19

How the Legal Orientation Program Works ............................................
Management and Oversight of the Legal Orientation Program.................
Who Does the Legal Orientation Program Serve.....................................
Nationwide Trends for Cases Beginning in Detention ..............................

19

24

27

36


Measuring the Impact of the Legal Orientation Program ................................. 43

Analyzing Trends in Immigration Court Data ..........................................
Tracking LOP Participants in the Courts .................................................
Comparisons Between LOP Participants and Other Detained Persons.......
Qualitative Interviews...........................................................................
Key Findings ........................................................................................

44

45

46

46

47


Recommendations and Next Steps ................................................................ 69

LOP Impact ......................................................................................... 70

Appendix I: Five­Year Life of Program Statistics (2003­2008).......................... 73

Appendix II: Data Organization and Analysis ................................................. 74


Vera Institute of Justice vi

I. Introduction
This report begins with a two­part introduction designed to place it in proper context. The first
part of the introduction presents a history of the Legal Orientation Program (LOP). The second
provides background pertinent to measuring the LOP’s performance and impact.

History of the Legal Orientation Program
In 2002, Congress appropriated $1 million to the U.S. Department of Justice to carry out Legal
Orientation Programs (LOP)—programs that refer cases to volunteer attorneys and conduct
individual and group orientations on immigration law and procedure—for detained persons in
removal proceedings (“detained proceedings”) before the Executive Office for Immigration
Review (EOIR), the office within the Department of Justice that manages U.S. immigration
courts. EOIR, confronted with a significant increase in the number of detained proceedings
resulting from an expanded use of detention, wanted to explore innovative ways to ensure that
cases were processed in a timely manner while also increasing access to pro bono legal programs
for detained persons. Congressional interest in funding the LOP was in part motivated by the
success of a pilot program run by EOIR in 1998. In an evaluation of that pilot program, EOIR
concluded that “rights presentations” for detained persons helped the Department of Justice
ensure that all respondents had a clear understanding of their procedural rights, led to cases being
completed more quickly, and increased availability of representation, usually pro bono, for
detainees with “potentially meritorious claims to relief.”3
EOIR modeled the LOP, which is housed within the agency’s Legal Orientation and Pro
Bono Program, on a project that relied on independent nonprofit legal advocates to advise
individuals in immigration detention of their rights. The Florence Immigrant and Refugee Rights
Project (“Florence Project”) developed a “rights presentation” model in response to a local
immigration judge’s concern that indigent persons detained by the legacy Immigration and
Naturalization Service (INS) at its Florence, Arizona, Service Processing Center were at risk of
having their statutory rights violated.4 The project recognized that pro bono attorneys from local

3

Executive Office for Immigration Review (EOIR). Evaluation of the Rights Presentation. Washington, DC: No
Date. http://www.usdoj.gov/eoir/statspub/rtspresrpt.pdf (last accessed August 31, 2007). “Relief” encompasses a
variety of requests to the immigration court that, if granted, will prevent the respondent from being ordered removed
from the United States. Examples include asylum, withholding of removal, cancellation of removal, and adjustment
of status. Voluntary departure, which if granted still requires the respondent to leave the country, is considered by
some to be partial relief.
4

When Congress created the Department of Homeland Security in 2003, the functions of the Immigration and
Naturalization Service were devolved into two new agencies: the United States Citizenship and Immigration
Services (USCIS) and Immigration and Customs Enforcement (ICE). Enforcement functions formerly performed by
the INS were transferred to ICE, whose responsibilities include administration and oversight of detention and
removal operations.

Vera Institute of Justice 7

firms could help ensure that detainees’ rights were protected.5 Throughout the early 1990s, the
Florence Project pioneered and refined a service model that encouraged people in detention to
play an active role in their own cases, whether or not they were represented by counsel. Rather
than focusing their limited resources on representing a small number of detainees, project
attorneys strove to provide all detained individuals with accurate legal information from which to
make more informed decisions about how to proceed with their immigration court cases. The
project also worked to dispel common misconceptions about the immigration court process and
thus decrease anxiety, confusion, and discomfort about immigration proceedings.
As the Florence Project started to gain recognition, federal officials became interested in
exploring its impact on the immigration system. In 1992, the General Accounting Office (GAO)
conducted a study of the project and concluded that its rights presentations resulted in substantial
time­savings for the government.6 The GAO report concluded that immigration hearings at the
Florence immigration court took less time because the detainees who appeared in immigration
court were already familiar with the removal proceeding process and their eligibility for forms of
relief from removal.
A 1994 bipartisan Senate resolution commended the Florence Project’s work and
recommended that the Department of Justice test similar programs at other INS Service
Processing Centers.7 In the fall of 1998, the Department of Justice established three 90­day pilot
projects that provided daily rights presentations to INS detainees at the Port Isabel Detention
Center in Los Fresnos, Texas (administered by the South Texas Pro Bono and Asylum
Representation Project, or ProBAR); the San Pedro Detention Facility in San Pedro, California
(administered by Catholic Legal Immigration Network, Inc., or CLINIC); and the Florence
Service Processing Center in Florence, Arizona (administered by the Florence Project). The
detention centers served by these programs were among the three largest in the country. In all
three centers, the majority of detained individuals spoke either English or Spanish—a
circumstance which simplified logistics.
In 2002, in response to an EOIR evaluation which found that the pilot programs had resulted
in cost savings and more efficient immigration courts, Congress appropriated $1 million in fiscal
year 2002 to develop the LOP by expanding the pilot project model to detention facilities across
the nation. EOIR contracted with Norwich University to oversee LOP operations and administer
subcontracts to local nonprofit organizations carrying out these operations. In early 2003,
Norwich selected six nonprofit organizations to provide LOP services to detained adults. Three
of these organizations had participated in the Department of Justice’s 90­day pilot projects,
though only one of these—ProBAR—continued to provide services at the original pilot project
site; the Florence Project began providing services at the Eloy Contract Detention Facility in
Eloy, Arizona, while CLINIC relocated the LOP from San Pedro to the Mira Loma Detention
5

Florence Immigrant and Refugee Rights Project, "Florence: The 'Justice and Efficiency Model,'" April 5, 2003.

6

General Accounting Office (GAO). Immigration Control: Immigration Policies Affecting INS Detention Efforts.
Washington, DC: GAO, 1992.
7

Senate Resolution 284, 103d Congress 2d Session (1994).
Vera Institute of Justice 8

Facility in Lancaster, California. The other three organizations to which Norwich University
awarded subcontracts included the Erie County Bar Association Volunteer Lawyers Project
(VLP), which worked at the Buffalo Federal Detention Facility in Batavia, New York; the
Northwest Immigrant Rights Project (NWIRP), which served detainees in Seattle, Washington;
and the Rocky Mountain Immigration Advocacy Network (RMIAN), which worked at the
Aurora Contract Detention Facility in Aurora, Colorado. LOP operations were introduced
incrementally at these six sites over the course of 2003. In 2004, Norwich University expanded
the LOP to the El Paso Service Processing Center in El Paso, Texas; services at this site were
provided by CLINIC.
Criteria for LOP Site Selection

EOIR and Norwich University worked together to select the original six LOP sites. They used a
number of criteria to do so. Among the most important of these was the requirement that a site
have low rates of representation and limited availability of legal services, thus ensuring that the
government invested in those sites with the greatest need. EOIR and Norwich also looked for
sites with high numbers of detained persons in removal proceedings; sought out sites in diverse
locations to determine whether the LOP could achieve success in different environments; and
made a strategic decision to place the LOP in sites that primarily served English and Spanish
speakers. (Because English and Spanish were the predominant language groups among people in
immigration detention, EOIR planned to identify and implement best practices for serving these
groups before expanding the program to locations with greater linguistic diversity.) Finally, the
likelihood that staff from the local immigration court, the local INS (and later ICE) office, and
the local detention facility would support the program was an important consideration in site
selection.
In evaluating proposals from potential subcontractors, EOIR and Norwich University
considered the experience of local nonprofit staff and the availability of “matching resources” or
other in­kind program services already in place at local nonprofits.
Involvement of the Vera Institute of Justice

In 2005, replacing Norwich University, EOIR made the Vera Institute of Justice the primary
LOP contractor. Vera agreed to administer, monitor, and provide technical assistance to the LOP
and to carry out research on the performance and impact of the LOP. Due to budgetary
constraints in 2005, Vera and EOIR ended the LOP program at the Buffalo Federal Detention
Facility. In late 2006, Congress doubled the appropriation for the LOP; consequently, EOIR
expanded the program to include six additional sites and also reinstituted the Buffalo program.
Figure 1, below, lists the sites at which LOP programs have operated since 2003 (listed by
site name assigned by Vera), hearing location code (used to indicate the location of the
applicable immigration court), facility name and location, subcontractor (the “LOP provider”
referred to in this report), and program start date.

Vera Institute of Justice 9

Figure 1: Legal Orientation Program

Subcontractors, Hearing Locations, and Facilities, 2003­2008

Site
Name

Court
Hearing
Location*

Facility Name

Port
Isabel

PIS

Port Isabel Service
Processing Center

Batavia

BTV

Buffalo Federal
Detention Center

Batavia, NY

Erie County Bar Association
Volunteer Lawyers Project (VLP)

February 21,
2003**

Eloy

EAZ

Eloy Detention Center
(CCA)

Eloy, AZ

Florence Immigrant and Refugee
Rights Project (FIRRP)

March 7,
2003

Seattle

AIR

Northwest Detention
Center (Corrections
Services Corporation)

Tacoma, WA

Northwest Immigrant Rights
Project (NWIRP)

March 17,
2003

Mira
Loma

LAN

Mira Loma Detention
Facility

Lancaster, CA

Catholic Legal Immigration
Network (CLINIC)

May 27, 2003

Denver

WSI

Aurora Detention
Facility (GEO)

Aurora, CO

Rocky Mountain Immigrant
Advocacy Network (RMIAN)

June 22,
2003

El Paso

EPD

El Paso Service
Processing Center;
Otero County Prison

El Paso, TX

Diocesan Migrant & Refugee
Services (DMRS)***

June 2, 2004

Houston

HOD

Houston Service
Processing Center
(CCA)

Houston, TX

University of Houston Immigration
Law Clinic

October 1,
2006

Newark

NEW

Middlesex County
Correctional Facility;
Hudson County
Correctional Facility

New Jersey

Legal Services of New Jersey
(LSNJ)

October 1,
2006

Laredo

LAR

Laredo Processing 
Center (CCA)

Laredo, TX

Lutheran Immigration and Refugee
Services (LIRS); Bernardo Kohler
Center (BKC)

October 1,
2006****

San
Antonio

SAD

Political Asylum Project of Austin
(PAPA)

October 1,
2006

San Pedro

SPD

San Pedro Service
Processing Center

San Pedro,
CA

Legal Aid Foundation of Los
Angeles (LAFLA)

October 1,
2006****

York

YOR

York County Prison

York, PA

Pennsylvania Immigration
Resource Center (PIRC)

October 1,
2006

San Diego

CCA

Otay Mesa Detention
Facility (CCA)

San Diego,
CA

ABA Immigration Justice Project
(IJP) of San Diego

January 2,
2008

Detention
Location

Subcontractor

Program
Start Date

Los Fresnos, ABA South Texas Pro Bono Asylum February 18,
Representation Project (ProBAR)
2003
TX

Pearsall, TX;
Pearsall Immigrant
San Antonio,
Detention Center (GEO)
TX

Notes: * This column lists the EOIR code for the court hearing locations in question.
** The Batavia program was not operational between May 19, 2005, and October 1, 2006.

***The subcontract for the El Paso LOP was devolved from CLINIC to DMRS in 2006.

**** Sites highlighted in gray are no longer operational. Services in Laredo were terminated in September

2007. The San Pedro Service Processing Center was temporarily closed in October 2007, and LOP
services were terminated at that time.

Vera Institute of Justice 10

Measuring Performance and Impact of the LOP
During the LOP’s first year of operation, in 2003, Norwich University reviewed program data
and conducted a small stakeholder survey. The survey and data review suggested that the LOP
had increased the efficiency of immigration court proceedings; improved access to legal services
for detainees in removal proceedings; decreased case completion times; and increased the
number of meritorious applications for relief filed with the immigration court.8 These findings
were consistent with EOIR’s evaluation of the 1998 pilot projects and the 1992 GAO study of
the Florence Project. However, none of these studies were comprehensive enough to rigorously
assess the potential impact of the LOP.
As a result, when EOIR contracted with Vera in 2005, it also asked that the Institute conduct
a systematic study of the ways in which the LOP might benefit program participants as well as
the immigration court and detention systems. In addition, EOIR asked Vera to develop and
document performance measurement methods for the LOP, to identify best practices, and to
formulate recommendations for program changes and improvements.
To ensure that the programs they support are cost­effective, private entities and government
agencies alike increasingly require that those programs be rigorously evaluated. Measuring the
performance and impact of programs can also lead to recommendations for change, thereby
making programs more efficient, ensuring that they are sustainable, and facilitating the process
of replicating effective practices elsewhere.9 In fact, the Organization for Economic Cooperation
and Development (OECD) has argued that program evaluation and performance measurement
are essential to good governance, as these procedures help ensure that government resources are
being used effectively and spent on activities with measurable outcomes.10 The Department of
Justice’s 2007­2012 strategic plan similarly emphasizes the importance of evaluating the
programs it sponsors, noting that formal, methodologically rigorous program evaluations that
examine fundamental questions of program design, implementation, and impact cannot be
substituted by internal audits, inspection, and review processes.11 While these latter activities are
essential for ensuring accountability, their methods are distinct from those used in social
scientific studies. Additionally, agencies can strengthen their credibility when they contract with
independent researchers to evaluate a program.
In general, there are three types of evaluations: process, performance, and outcome
evaluations. A process evaluation documents how a program was intended to work (its
blueprint), how it actually works (fidelity to the blueprint), and who is involved in program
8

Unpublished survey questionnaire results submitted by Norwich University to EOIR (July 2004).

9

Aucoin, Peter. Decision­Making in Government: The Role of Program Evaluation. Discussion Paper, 2005.
http://www.tbs­sct.gc.ca/eval/tools_outils/Aucoin/Aucoin_e.asp

10

Odile Sallard. Organisation for Economic Cooperation and Development. The OECD Observer. Paris: Nov 2005
(252/253), p. 9.
11

Department of Justice Strategic Plan 2007­2012, accessed September 9, 2007.
http://www.usdoj.gov/jmd/mps/strategic2007­2012/appendix_c.pdf
Vera Institute of Justice 11

activities such as implementation, organization, and administration. One can think of a process
evaluation in terms of what was done. A performance evaluation studies the extent to which
programs function as they were intended to (in other words, how the blueprint translates into
practice—especially with regard to services delivered or other outputs). One can think of a
performance evaluation in terms of how much was done. Finally, an outcome evaluation assesses
whether the outcomes changed over the course of the program and whether the program was
responsible for any observed changes. One can think of an outcome evaluation in terms of how
well the program did what it was supposed to do.
EOIR asked Vera to develop a plan for incorporating all three types of evaluation in its study
of the LOP. In particular, Vera was asked to address the following research questions:
•	 Process: What were/are the planned and expected activities of the LOP?
•	 Performance: To what extent is the LOP working as intended? What services is it
providing, and to whom? How could it work more effectively?
•	 Outcome(s): Did any desired change occur that might be attributed to the program? To
what extent can the desired changes be attributed to the LOP?
EOIR additionally requested that Vera draw on its research findings to make recommendations
for program modifications or improvements.
Evaluation Questions

EOIR asked Vera to collect and analyze data on how the LOP is working and its impact on the
immigration courts and detained persons; to identify legal services that are likely to improve the
efficiency of the courts and affect case outcomes; to conduct interviews with detainees and other
project stakeholders; and to help the agency identify data sources that might be used in a cost
assessment.
Prior to addressing these questions directly, Vera researchers worked with LOP providers,
EOIR staff, Vera program managers, detention facility staff at LOP sites, and detainees to define
the goal (desired impact) and objectives (activities necessary to achieve that impact) of the LOP.
These discussions resulted in agreement on two primary objectives and nine subsidiary
objectives, described below.
Primary objectives.
The LOP should improve
(a) legal access for detained persons in removal proceedings by providing impartial, accurate
orientations to the immigration court process and providing detainees with information to
help them determine how to proceed in immigration court and
(b) efficiencies in the immigration detention and immigration court process for detained
persons in removal proceedings. (Efficiencies may be defined as the best possible
Vera Institute of Justice 12

allocation of resources [maximum benefits for minimum costs] or as enhancements of
systems and processes that enable them to work more smoothly.)
Subsidiary objectives.
1.	 Ensure detainees have a general understanding of what the removal and immigration
court processes entail and the rights of detained persons in these processes; how to self­
screen for eligibility for relief (using information learned in orientations as well as
handouts and law library materials); how to access forms of relief when eligible; how to
prepare for the immigration court process; how to access legal representation if available;
and what representation entails.
2.	 Teach detained persons who want to leave the United States how to do so as quickly as
possible, and should educate them about the consequences of removal from and unlawful
return to the United States.
3.	 Help detained persons identify their eligibility under the law for forms of relief from
removal, and should inform ineligible detainees of the risks of filing “frivolous claims.”
4.	 Help detained persons who want to fight their cases access either legal representatives
(paid or pro bono) or self­help (pro se) services.
5.	 Teach detained persons how to effectively act on the decisions they have made about how
to proceed in their removal proceedings.
6.	 Educate detained persons on how to respond to routine questions asked by an
immigration judge at a Master Calendar Hearing (e.g., answers should be given orally
and in a voice loud enough to be picked up by a tape recorder; respondents should answer
truthfully).12
7.	 Teach unrepresented persons how to effectively represent themselves pro se.
8.	 Reduce detainee anxiety about the removal and immigration court processes by providing
factual information and treating detained persons with dignity and respect.
9.	 Be carried out by independent legal representatives who work in collaboration with local
detention facility, ICE, and EOIR staff.
In addition to studying the LOP’s progress in meeting these objectives, Vera researchers
considered the extent to which the LOP might also assist EOIR in meeting some of the agency’s
priorities as outlined in its 2005­2010 Strategic Plan. In particular, we focused on three agency­
wide objectives described in the EOIR Strategic Plan that intersect with the objectives of the
LOP. First, objective 1.1 of that plan states that EOIR “must eliminate case backlogs by the end
of fiscal year 2008,” and “must render ‘expeditious decisions’ and continue to reduce ‘frivolous’
applications.” Objective 1.2 of the plan commits the agency to “implement improved caseload
12

Master Calendar Hearings deal with procedural matters. At an initial Master Calendar Hearing, respondents are
informed of the relevant rights and charges and are asked whether they dispute and want to oppose the government’s
allegations. Master Calendar Hearings are distinguished from merits hearings where testimony is taken and issues
are tried before the immigration judge.
Vera Institute of Justice 13

management practices,” which includes “studying failure to appear rates.” Finally, EOIR’s plan
commits the agency to “encourage pro bono representation,” noting that, “effective
representation can add value to the adjudicative process.” In our discussion of findings and
recommendations, we detail ways in which the LOP may be an efficient and cost­effective
method of aiding EOIR in meeting these objectives.
Evaluation Methods

Any evaluation requires a clear statement of the research question or problem and a baseline of
comparison, which in this case asks what would have happened in the absence of the LOP. That
is, what if the LOP had not been available? In order to accurately answer “what if” questions,
program evaluations generally employ an experimental research design. This classic method of
obtaining a baseline of comparison enlists some people in the program or “experiment,” while
others, randomly assigned to a control group, receive no program services. In theory, both groups
are subject to the same institutional and environmental influences except for the program or
intervention being tested. Generally, control group experiences and outcomes are measured in
parallel with participants’ experiences. In the case of evaluating the impact of the LOP, however,
we determined the use of a classic experimental design to be inappropriate for two major
reasons: (1) feasibility and (2) ethics.
First, the LOP was designed as a voluntary program and was operating as such at all sites. It
was not feasible to alter this basic tenet of the program. In essence, programmatic demands
trumped research preferences.
Second, there is an ethical issue involved in research with detained or incarcerated
populations. The LOP aims to provide all detained persons in immigration court proceedings
with information on immigration law and procedure, to offer guidance to those individuals
appearing pro se, and refer cases with potential relief to pro bono attorneys. When we designed
our research, LOP providers expressed a concern that it would be unethical to only provide legal
services to one group of detainees while depriving detainees assigned to the control group access
to legal information.
Given the institutional constraints beyond our control and ethical concerns expressed by
program providers, we determined that randomly assigning individuals to control or
experimental groups, though the most scientifically sound method, was not feasible in this case.
The alternative to an experimental design is to employ a quasi­experimental design, which
divides participants into groups for comparative purposes but employs a non­random assignment
of persons to these groups. A quasi­experimental design does not permit the same levels of
certainty that an outcome is the product of the program being tested, but it is the next best way of
measuring program impact when random assignment is not practical or feasible.
The non­experimental configuration of this research, therefore, makes all findings and
recommendations based on them descriptive, suggestive, and exploratory rather than confirmed.
This was a shortcoming Vera and EOIR jointly accepted when undertaking this study.
Subsequent work will be needed to strengthen the study design and to lead to more robust,
Vera Institute of Justice 14

confirmatory analyses and inferences. That stated, and the limitations of quasi­experimental
designs noted, quasi­experimental designs are commonly used to organize first­generation
research in the absence of experimental alternatives or where experimental alternatives are not
attractive or possible because of the heavy investment in time and resources they usually entail.
These exploratory designs comprise comparative rather than control­group analyses that can
nonetheless preliminarily and tentatively reveal patterns across groups—patterns that are of great
interest and are often noteworthy to program staff and policymakers alike and that often deserve
a more focused, sustained, and substantial investment of resources than may be available.
Throughout the analytical portions of this report, our findings are couched in comparative terms
whenever possible, which is so for virtually all analyses. Sometimes we compare statistics before
and after the introduction of the LOP at particular sites, while in other instances we compare
outcomes based on participation in the LOP.
In summary, we determined, in consultation with LOP stakeholders, that the best alternative
to random assignment was to analyze patterns at LOP sites in the years immediately before and
after the program began, and then compare LOP participants with comparison groups comprising
persons who were also detained but who did not participate in the LOP and attended immigration
court in sites not hosting the LOP. Below and throughout this report we describe in greater detail
our methods for organizing LOP participants and other detained persons into groups for analysis.
Evaluation Work Plan

Because the LOP is dependent on Congress to allocate funds to the program each fiscal year,
Vera and EOIR divided the contractually required Performance Outcome and Measurement Plan
(POMP) and evaluation activities into three discrete phases of work, each of which would build
on the previous, and each of which could stand alone if need be (in the absence of any additional
funding from Congress). This report represents our interim findings following completion of
Phase II of the POMP.
In the first phase of research, Vera documented LOP performance, and “cleaned” data
compiled by Norwich University, the prior contractor, in order to produce statistics for EOIR that
showed the numbers of persons served by the LOP over the life of the program. In that phase of
research, Vera also developed and implemented a program service database, LOPster, that
subcontractors use to track and monitor their performance by recording information on all LOP
participants in standardized ways. Vera researchers also began to interview program stakeholders
in order to document their impressions of where the LOP might have an impact. This was
important in order to narrow the focus of the evaluation of the LOP’s impact, particularly
because EOIR had very limited funds to assign to evaluation activities. Finally, in the first phase
of evaluation, researchers analyzed aggregate administrative data from EOIR to identify trends in
the immigration courts in the 36 months immediately before and after the LOP was
implemented. By comparing macro­level differences between immigration courts hosting the
LOP and other immigration courts around the country, Vera researchers were able to identify
different patterns and trends that were occurring nationwide before and after the LOP began.
Vera Institute of Justice 15

Working in close consultation with EOIR staff, we plotted variables such as case processing
time, representation rates, and grant rates. Knowing these patterns and trends enabled the
research staff to gauge the extent to which any observed positive outcomes associated with LOP
might simply reflect trends that should not be attributed to the LOP. We compared different
variables and immigration courts before identifying a few key points of difference between
immigration courts hosting the LOP and other immigration courts and a few key patterns and
trends that appeared to begin around the time LOP services began.
With additional funding from EOIR for a second phase of research, Vera researchers
continued to monitor and document program services and statistics. We matched data collected
by LOP subcontractors with immigration court data in order to track LOP participants’ cases in
the immigration courts. We examined several variables but focused our attention, at EOIR’s
request, on case processing time, representation rates, rates at which LOP participants pursued
various forms of relief from removal, case outcomes (defined in our study as the immigration
judge’s final case decision in a removal proceeding), and in absentia removal orders.13 We
generated descriptive information about LOP participants’ immigration court cases in general,
and then analyzed this information for different subgroups of LOP participants.14 After analyzing
13

For the purposes of this research, Vera defines a case as the sum of all the proceedings involving a single
respondent before the immigration courts. This means in our analysis a single case may contain numerous
proceedings and numerous applications for relief that have been initiated and decided in the time between an initial
Master Calendar Hearing and the final decision issued by the immigration judge in the last proceeding in the case. In
many reports authored by EOIR, cases are evaluated and reported at the proceeding level. Vera researchers
determined that for the purposes of our study, it would be confusing to report on proceedings as opposed to what we
defined as cases. This is because in EOIR’s case management system, each case—from initial Master Calendar
Hearing to final case decision—may be composed of several proceedings, which are distinguished in EOIR’s data
by generation numbers that descend from 99. When a respondent’s detention status or hearing venue changes, the
case before the immigration courts is typically transferred from one hearing location or immigration judge to
another. When this occurs, the first proceeding in the case is closed, and a new proceeding is opened in the
immigration court records. However, the case has not been concluded and reopened. For example, if an asylum
seeker is detained at the initial Master Calendar Hearing and is later released and granted a motion to change of
venue to a different hearing location, the pending asylum application remains active as it moves from one
immigration court and judge to another, but a new proceeding is opened in EOIR’s records. Thus, the number of
days in each proceeding does not reflect how long a case was active in the immigration courts.
Additionally, because we wanted to measure any potential impact of the LOP on the immigration courts using
the most consistent measures, we made a decision not to include in case processing time any days that might have
accrued after the immigration judge’s decision was issued. When either party (ICE or the respondent) reserves the
right to appeal, the case is not completed until the appeal deadline has passed with no appeal filings, until a decision
has been issued on the appeal by the Board of Immigration Appeals, or, if a case is remanded to an immigration
judge, until a decision has been issued. There are other scenarios that might also prevent a case from being
immediately completed (or closed) after the immigration judge issues a decision. Because of all these reasons, our
definition of case processing time may not match definitions used by EOIR or other researchers. However, we
believe our definition does allow us to most accurately assess time for our purpose, which is to see if LOP is
correlated with any reduction or increase in the number of days a matter remains before the immigration courts (see
Appendix II for more detailed description of how the data was organized and analyzed).
14

As described later in the report, we defined subgroups by detention status (detained throughout the immigration
court process versus released before the final decision was issued), representation status (legal representation at
some point in the case versus no legal representation), type of application for relief from removal filed with the
immigration courts, and in some instances, by nationality, language, or the statute used by ICE to charge the
respondent’s “removability.” Many research studies create subgroups according to age (generally determined by
Vera Institute of Justice 16

data on LOP participants, we created a comparison group comprising detained persons who did
not participate in the LOP and whose initial Master Calendar Hearings were scheduled in courts
without the LOP. In our discussions and findings of the program’s process, performance, and
outcome, we present information about the immigration courts before and after LOP and contrast
LOP participants with comparison groups in order to make statements about similarities and
differences between LOP participants and other detained persons in removal proceedings.15
In addition to studying immigration court data on LOP participants and other detained
persons in removal proceedings, we conducted a total of 53 qualitative interviews with LOP
stakeholders, including immigration judges, court administrators, detention facility staff, and
local ICE employees. These interviews focused on documenting stakeholder impressions of the
LOP and any observations they had about the impact of the LOP. We also asked stakeholders for
input on some of our preliminary findings; these interviews helped us identify “confounding
factors” that might prevent us from seeing an actual impact of the LOP in an analysis of the
variables specified above. Findings from these qualitative interviews are included throughout the
report and often complement findings from our analysis of court records.
Finally, EOIR requested that Vera conduct interviews with detainees in order to assess the
impact of the LOP from their perspective. In the second phase of the research, we conducted
preliminary interviews with 33 detainees but determined that their experiences were too diverse
to be able make generalizable statements from short interviews with a small group of detainees.
The logistics and resources (travel, permission from multiple detention facility operators,
inability to conduct phone interviews or record interviews in detention settings) required to
conduct intensive, longer interviews with large numbers of detainees deterred us from carrying
out more formal interviews with detainees. Rather, we drew on these preliminary interviews,
combined with data in our monthly reports and interviews with stakeholders to draft a set of site
self­evaluation materials that could be administered on an ongoing basis to test what LOP
participants are learning in orientations and workshops and to measure how program participants
are applying this knowledge. These draft materials, once validated for reliability, will enable
LOP providers to measure and report on the impact of the LOP on detainees on an ongoing basis,
across sites and over time.
As explained above, program evaluations focus on documenting process, performance, and
outcomes. In the next section we detail the process and performance of the LOP, describing how
date of birth) or sex, but we were unable to do so because EOIR did not record this information in 2006. We hoped
to create two additional categories of subgroups that we were not able to successfully construct because of
incomplete EOIR data in one case (time in the United States) and, in the second case, due to the time that would
have been required to create subgroups according to residency status (legally present versus unlawfully present).
Because the immigration courts do not record information about residency status, we would have had to create
proxy measures for residency status by analyzing the charges and types of relief from removal sought by each
respondent, which is possible but beyond the scope of what we were able to deliver with limited funding in the
second phase of work.
15

For the purpose of clarity and to preserve the integrity of the research design, when we make comparisons
between LOP participants and other detained persons who did not participate in the LOP, we are excluding any
detainees at LOP sites who did not participate in the LOP (see Appendix II).
Vera Institute of Justice 17

the program was intended to work and how it works in practice; services provided to LOP
participants in 2006 and over the life of the program; and basic demographic information on
program participants, such as nationality, language, and immigration charges.16 In Section III, we
describe observations of program impact and discuss ways in which the LOP is and is not
meeting its primary objectives of (1) improving legal access for detained persons in removal
proceedings by providing impartial, accurate orientations to the immigration court process and
providing detainees with information to help them determine how to proceed in immigration
court; and (2) improving efficiencies in the immigration detention and immigration court process
for detained persons in removal proceedings. In Section IV we detail recommendations and next
steps for additional research and for program improvements that might emerge from this
evaluation.

16

Given the length and focus of this report, we do not include detailed information on demographics at each LOP
site, but Vera is producing site­specific reports showing detailed information about LOP participants and services.
Vera will use these reports to develop management plans in consultation with each LOP site’s program manager.
Vera Institute of Justice 18

II. Legal Orientation Program Services
In this section of the report, we describe the basic program operation plan and how it is
implemented across LOP sites, noting relevant variations that may exist in program
implementation. We then discuss program oversight activities carried out by Vera, including
analysis of program statistics. After describing how we collect and gather program service data,
we discuss whom the program is serving, as well as some of the ways in which the program is
and is not reaching the population it intends to serve. Finally, we provide recommendations for
improving the program’s reach.

How the Legal Orientation Program Works
The Legal Orientation Program is a court­based legal education program for detained noncitizens
in immigration court proceedings. The program provides a range of services, including group
orientations, individual orientations, self­help workshops, dissemination of written legal
educational materials, and, for a limited number of cases, pro bono recruitment.17
A significant feature of the LOP is that its services are limited to legal orientation (as
opposed to legal representation). Program providers are not permitted to use LOP funds to
engage in legal representation. (They may do so with funds from other sources.) Before
providing any services, LOP presenters make this clear, explaining that their role is to provide
participants with information on immigration law and procedure—not to represent them. In
addition, LOP providers ask detainees who take part in individual orientations and self­help
workshops to sign a statement indicating that they understand that LOP providers are not serving
as legal counsel.
The LOP offers the same basic services at each site where it operates, but there is some
variation in the methods used for delivering these services. As an immigration court­based
program, the LOP’s primary goal is to provide legal orientations to detained persons in removal
proceedings. However, detained persons who do not have active immigration court cases are not
prohibited from attending presentations, provided that their attendance is logistically feasible and
does not prevent detainees in removal proceedings from receiving LOP services.18
17

Pro bono counsel may include attorneys, accredited representatives, and supervised law students affiliated with
nonprofit organizations (including the same nonprofits who administer the LOP, using different funding sources).
18

Participants without immigration court cases may be subject to stipulated orders of removal, expedited removal,
or post­removal order review. A stipulated order of removal is a written agreement between a person and the
government that agrees that the person will be removed from the United States. A stipulated order must be approved
by an immigration judge, but a court proceeding is not required. The immigration judge has the option to inquire
into the validity of the individual’s waiver of the right to contest removal. In our research, we included cases
involving stipulated orders of removal when analyzing certain patterns and trends (such as the distribution of
nationalities across LOP sites), but we omitted observations about stipulated removals when reporting on
representation rates and case processing times since the detained person generally does not appear in court on these
cases. Where relevant, we indicate whether stipulated removal cases were included or omitted from the analysis
being presented.
Vera Institute of Justice 19

Group Orientations

The group orientation is an essential component of the LOP; nearly everyone who participates in
the program takes part in one. Group orientations are designed to give detained persons—
regardless of whether they have access to legal representation—a general overview of
immigration law, their legal rights, and the immigration removal process. These orientations
explain the removal hearing process and provide general information about the statutory
requirements for various defenses and forms of legal relief.19 Additional topics include the notice
to appear (the document that specifies the charges in a removal proceeding and directs the
individual so charged to appear in immigration court); procedures for assessing eligibility for
and, in suitable cases, applying for voluntary departure or release under bond; and the
consequences of re­entering the United States after a removal order has been issued.
Group orientations are led by immigration attorneys or paralegals under the supervision of
attorneys. Orientations generally last from 30 minutes to an hour but may run longer. (The length
of a presentation depends on the size and needs of the group; in many cases, factors beyond the
presenter’s control—such as how much time participants have before they are required to return
to their dormitories for detention center “counts”—determine the length of the presentation.)
Information is presented in a manner consistent with the principles of adult education in mind,
taking into consideration differences in language, culture, and levels of formal education. Most
providers use visual aids such as flip charts or PowerPoint presentations to make the material
more accessible. (Examples of common visual materials include enlarged notice to appear forms,
presentation outlines, and pictorial representations of procedural concepts.) Group orientations
are conducted in the language or languages most suitable to a majority of those attending,
typically Spanish or English. However, LOP providers are required to provide taped or written
orientation materials to detainees who speak other common languages.
While the material covered in the group orientation is similar from one site to another, LOP
providers customize the orientation to meet the needs of the detention center population and the
session participants. For example, sites that see large numbers of individuals facing charges
based upon their criminal convictions may cover certain forms of relief more in depth, while
sites serving large numbers of individuals who recently arrived in the country and have not been

Expedited removal is a process that allows an immigration inspector, rather than an immigration court, to
remove from the United States certain classes of inadmissible non­citizens; the Illegal Immigration Reform and
Immigrant Responsibility Act of 1996 (IIRIRA) mandates that persons who arrive at a U.S. port of entry without
travel documents or who present fraudulent documents must be detained and placed in expedited removal. Post­
removal order review is a multi­step process that evaluates the likelihood of removal in circumstances where a
person is detained for a lengthy period after a removal order becomes final and the possibility of release if removal
is not likely in the foreseeable future. Delays in removal are often caused by the reluctance or refusal of the
receiving country to issue the documents and permission necessary to effect a removal. See “Immigration Court
Process in the US,” US Department of Justice, EOIR, April 28, 2005,
http://www.usdoj.gov/eoir/press/05/ImmigrationCourtProcess2005.htm
19

For examples of legal orientation program materials, see the “Know Your Rights” publication series of the
Florence Immigrant and Refugee Rights Project. <http://www.firrp.org/kyrindex.asp>

Vera Institute of Justice 20

charged with a crime may skip those topics. Similarly, the presentation style used for a group of
10 might vary considerably from that used for a group of 100.20
Different sites also see detainees at different stages of the immigration court process,
depending on what has been negotiated with local ICE or detention center personnel.21 At some
sites, providers receive copies of the immigration court hearing calendar, or docket. These
dockets are then redacted to generate a list of names of persons with upcoming initial hearings in
the immigration court. The redacted lists are used by the detention facility to assemble detainees
for group orientations. At sites that use this method, LOP presenters typically meet with the
group orientation participants a day or two before their initial Master Calendar Hearings. At
other sites, providers use a “new arrival” list, which allows them to provide the LOP to all
detainees soon after they are admitted into the detention facility. At sites that use this method,
providers may see detained persons who are subject to expedited removal, reinstatement of
removal, or administrative removal proceedings.22 They may also see detained persons who will
request (or who have already signed) stipulated orders of removal and, thus, may choose to cover
these processes in their presentations.
Coordinating group orientations—and, indeed, all LOP services—requires the cooperation
and assistance of many stakeholders. Immigration court personnel must agree to share
immigration court dockets, or facility staff must agree to share lists of “new arrivals” (persons
who have recently been admitted to the detention center) so that the LOP providers can generate
a list of potential participants. Working from this list, detention facility staff must then bring
potential participants to the LOP. The facility must also provide security clearance and access for
LOP staff, set aside space for the orientations, assemble detainees at the scheduled times, and
escort detainees to and from LOP sessions. The success of the program depends on all of these
logistical elements working together smoothly.
20

When designing our research, we considered any major differences that might affect outcomes, such as the size of
the group in a presentation session. In addition, program managers are constantly evaluating the performance of
LOP providers through site visits and other program oversight activities, as described in the next section of this
report. However, while we do discuss differences in program outcomes based on different levels of intensity of
participation in the LOP below, we have not had the resources to assess the effect of individual presenters or
particular presentation styles on program outcomes, in part because the presentation’s focus changes from day to day
depending on the participants in each session.
21

Under the LOP contract, sites are required to “provide group orientations as soon as feasible to all detained aliens
(with reasonable exceptions to be approved by the EOIR Contract Officer’s Technical Representative), regardless of
representation status, who are or may be placed in immigration removal proceedings, prior to their initial Master
Calendar Hearing in the Immigration Court.” For various strategic and logistical reasons, however, some sites have
made the decision to see all new arrivals.

22

“Reinstatement of removal” refers to cases where an ICE officer reinstates a prior order of removal against a
noncitizen who has been previously removed (and who subsequently re­entered) the United States. “Administrative
removal proceedings” may be brought by ICE against a person who is not a lawful permanent resident and who has
an aggravated felony conviction. Administrative removal proceedings are determined by ICE rather than the
immigration courts. (An aggravated felony is a crime that Congress has deemed serious, although some
misdemeanors under state law are classified as aggravated felonies under immigration law. Persons with aggravated
felony convictions have fewer remedies available to them than others in removal proceedings.)

Vera Institute of Justice 21

While the LOP has by and large operated smoothly, running a program that involves two
government agencies and numerous contract facilities—and that works inside a detention
environment—can be logistically complicated and inevitably leads to challenges. At various
times over the course of the program, LOP providers have encountered difficulties obtaining
immigration court dockets or new arrival lists. There have been many instances of detention
facility staff, particularly those unfamiliar with the program, failing to bring the detainees to the
orientation location in a timely manner or, in some cases, failing to bring them at all. At some
sites, providers have also experienced difficulty finding a suitable location for the group
orientation. While some LOP providers present the group orientation in noisy recreation or
multipurpose rooms, others are able to use empty immigration courtrooms or detention center
libraries. For obvious reasons, background noise and other distractions can impede the
participants’ ability to concentrate on the orientation. LOP providers have worked with detention
facility staff and ICE to address these and other logistical issues.23
Individual Orientations

Individual orientations are also integral to the LOP model and are conducted at each LOP site for
detained persons without representation who have participated in a group orientation. Individual
orientations aim to provide detailed and specialized information about particular forms of legal
relief or components of the immigration court process. This is necessary because group
orientations only provide a broad overview of the law and do not go into detail. In an individual
orientation, the presenter may also respond to individual questions.24
In individual orientations, the amount of time that the presenter spends with each detainee
varies according to the detainee’s needs, the number and complexity of the questions posed, and
the number of detainees who want an individual orientation. In some sites, additional program
staff are available to assist with individual orientations following the group presentation. In other
sites, presenters or other staff members conduct individual orientations on the same day as the
group presentation so as not to interfere with detention center censuses (or “counts”) or meals
and in an effort to make individual presentations available to as many group orientation
participants as possible. The amount of time spent on individual orientations is also affected by
site­specific strategic decisions. For example, at the Eloy detention facility, the Florence Project
gives a relatively concise group orientation, then devotes a significant amount of time to repeated
in­depth individual orientations with a majority of the detainees. At Port Isabel, on the other
23

LOP providers report to Vera any changes in program implementation, such as changes in the location of the
group orientations, challenges they encounter with new staff, or successes such as receiving permission to use a
laptop computer to show a PowerPoint presentation. Vera tracks and reports to EOIR on all of these implementation
changes and works with each LOP site to address any changes that create challenges to successful implementation of
the program.
24

Individual orientation presenters must always take care to distinguish between general legal information and legal
advice. While they are encouraged to provide general information, providing legal advice would mean crossing the
line into “legal representation,” which is prohibited when using LOP funds. (As noted above, though, some LOP
providers receive funding from other sources—a circumstance which permits them to provide direct representation.)
Vera Institute of Justice 22

hand, ProBAR conducts a lengthy and thorough group orientation followed by relatively short
individual orientations with a smaller group of individuals.
The interplay between the group and individual orientations also depends on the size of the
group in the group orientation. If the group is relatively small, the group presentation might
resemble an individual orientation. Under such circumstances, detainees are more likely to feel
comfortable talking about their situation in front of the group, and the presenter may decide to
address individual questions in the group session.
Self­Help Workshops

Self­help workshops—interactive, classroom­like sessions that provide detainees with the skills
they need to represent themselves in immigration court (pro se)—are conducted on an
intermittent basis in response to the needs of the detained population. Self­help workshops
usually feature groups of three or more unrepresented detainees and cover such topics as the
collection and presentation of evidence, general information about how to properly fill out
applications, and other legal advocacy skills, as needed. Recent workshops have focused on a
number of additional topics as well, including how to pursue specific forms of relief or defenses
from removal (including voluntary departure); custody redetermination (bond) hearings; special
procedures such as temporary protected status; reinstatement of a previous order of removal or
deportation; “reasonable fear” or “credible fear” proceedings; and post­removal order review.25
To determine whether LOP participants would benefit from a self­help workshop, LOP
providers usually keep track of the relief applications or defenses being pursued by each
participant and consider whether there are any special needs related to those applications that
might be served by a self­help workshop. For example, if there are five Spanish speakers who are
interested in pursuing cancellation of removal, an LOP provider might coordinate a self­help
workshop on that topic in Spanish.26 Self­help workshops are only available to persons who have
not retained counsel.
Dissemination of Written Legal Education Materials

To supplement the assistance offered in group orientations, individual orientations, and self­help
workshops, LOP providers distribute written materials to program participants. As required by
the LOP contract, all written materials are approved by EOIR prior to distribution and are made
available to all interested persons. All providers make available large pro se packets (modeled on
25

“Temporary protected status” allows nationals of a particular country to remain in the United States in cases
where the Secretary of the Department of Homeland Security has determined that it is unsafe to return to that
country. “Credible fear” and “reasonable fear” proceedings are interviews conducted by asylum officers to
determine whether an individual meets the threshold for asserting a claim of asylum, withholding of removal, or
withholding under the Convention Against Torture.

26

Cancellation of removal is a form of discretionary relief available to lawful permanent residents and non­
permanent residents. (Note that there are different eligibility requirements for each of these two groups.) When
adjudicating cancellation proceedings, immigration judges use a list of statutory criteria to weigh the equities related
to removing an individual from the country.
Vera Institute of Justice 23

packets originally developed by the Florence Project) for detainees who are applying for relief
without legal representation. Providers also distribute standard sample motions, briefs, and
letters, as well as concise fact sheets that describe various forms of relief. Written materials are
continuously updated based on the needs and preferred languages of the detained population. At
some sites, LOP staff have also worked with detention facility administrators to update legal
reference materials in facility libraries, which are mandated under the ICE detention standards.
Pro Bono Coordination

To increase access to legal representation, LOP providers also provide pro bono coordination
services. The degree to which providers engage in pro bono coordination varies from one site to
another, but at all sites, the bulk of this work consists in reaching out to local attorneys or other
representatives to cultivate an interest in pro bono work and to provide opportunities for training,
sometimes for Continuing Legal Education (CLE) credit. At several sites, the remote location of
the detention facilities and immigration courts, short immigration court deadlines, a shortage of
mentors for legal representatives without experience in immigration law, and a lack of pro bono
counsel have all presented significant obstacles to this work. While pro bono coordination is an
important component of the LOP program, it is the least­funded of all LOP services: only a few
providers have dedicated staff resources or designated budget line items for pro bono
coordination.

Management and Oversight of the Legal Orientation Program
Vera staff work in close collaboration with EOIR to manage the LOP and to coordinate the
nationwide implementation of the program. Vera’s management strategy is multi­pronged,
drawing on the skills of program management, technical assistance, research, fiscal, and legal
staff to ensure that the project meets its objectives in a timely manner; to respond quickly and
effectively to unanticipated changes and problems; to disburse funds to subcontractors; and to
conduct performance measurement and evaluation. To monitor the activities of subcontractors,
Vera staff use a comprehensive approach that combines quantitative program data with feedback
from and communication with LOP sites. Finally, Vera staff and LOP providers work together to
document best practices, improve program performance, and ensure contract compliance. In
what follows, we describe the key management activities performed by Vera staff on an ongoing
basis.
Site Visits

Vera managers and EOIR staff make annual visits to each LOP site to meet with LOP providers,
observe presentations or workshops, and discuss program performance. During these site visits,
Vera managers and EOIR staff also meet with stakeholders (such as local immigration court and
ICE personnel) to obtain their input, and record detailed comments and observations about
program performance and challenges to implementation. After the visit, Vera managers and
Vera Institute of Justice 24

EOIR staff discuss their observations and subsequently provide feedback to LOP providers. In
addition, Vera research staff—social scientists not directly involved in the day­to­day
administration of the LOP—make periodic site visits to conduct confidential qualitative
interviews with LOP staff and stakeholders, as required by EOIR.
Standardized Reporting of Program Data

Soon after EOIR contracted with Vera, Institute staff built a customized database (LOPster) to
ensure that program service data is recorded in a standardized manner by all LOP providers.
Providers are required to record basic data in LOPster. Vera also created a LOPster user guide,
which is constantly updated and includes rules for reporting on different types of services and
participant data.
At the beginning of each month, subcontractors use LOPster to create data reports, which are
then submitted to Vera. These data sheets allow Vera researchers to track the following statistics
for each site:
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	
•	

The number of group orientations provided;
The alien identification numbers (A­numbers) of group orientation participants;
The number of group orientation participants whose A­numbers were not recorded;27
The average number of participants at group orientations;
The range in the number of participants at group orientations (for example, group

orientations might have ranged from 4 to 12 participants in a given month);

The number of individual orientations provided;
The number of individual orientation participants who are potentially eligible for relief
from removal;28
The frequency with which individuals were potentially eligible for each form of relief
from removal;
The sex of individual orientation participants;
The nationality of individual orientation participants;
The language of individual orientation participants;
Any changes in location of group orientations;
Any changes in topics presented at group orientations;
Self­help workshop topics;
Number of self­help workshop attendees;
The language in which each workshop was conducted;
Program challenges, progress, or changes;
Staff changes and recruitment;

27

In some detention centers, detained persons are given a unique detention center identification number and are thus
not required to carry their A­numbers with them. Other participants have not yet been received or have misplaced
the charging document (notice to appear), which lists the A­number they have been assigned.

28

For confidentiality reasons, this information is not linked to identifiers when reported to Vera.
Vera Institute of Justice 25

•	 Pro bono recruitment and training efforts;
•	 Any departure from the program operation plan, together with the explanation for such
changes;
•	 Stories of participants who obtained relief from removal, whether by representing

themselves (pro se) or with the assistance of pro bono counsel;

•	 A­numbers of any participants who received pro bono referrals; and
•	 For cases where a detained person has legal representation, the type of representation (a
retained attorney; the LOP provider, using funds from a source other than the LOP;
another nonprofit; pro bono).
Vera researchers compile the information from these monthly data reports in a master
database, highlight any changes they observe, and send the reports back to each LOP provider
with any questions they may have. After addressing the researchers’ questions, providers return
the reports to Vera for review and submission to EOIR. This monthly process provides a constant
feedback loop—sites track and monitor their own performance, Vera reviews monthly data
reports and follows up as needed with LOP providers, and EOIR receives program statistics
(including highlights of any new trends) and written explanations of any changes in services or
participants. When relevant, Vera also submits notes from meetings, conference calls, and
monthly site visits to EOIR, as well as information about any potential delays in the project
timetable.
Vera program managers use the data from these monthly reports to monitor how well each
LOP provider is meeting the obligations and objectives set forth in its contract. The data and
program notes form the basis for regular discussions about each provider’s performance. When
the data indicate that a provider’s performance is moving in an unfavorable direction, Vera
managers work with the site director to determine the cause of the problem and develop
strategies for addressing it. When the data indicate a positive performance, Vera managers work
with providers to identify effective practices that might be replicated at other sites.
One­on­One Conference Calls

Another way that the Vera monitors the LOP is through regular one­on­one conference calls with
the project director for each site. These calls provide an opportunity to discuss contract
compliance, program performance, new developments, site­specific concerns and concerns
raised in monthly reports, data analysis, and site visits. They also give Vera managers a chance
to help trouble­shoot any challenges that arise. (When problems or challenges arise, Vera staff
communicate with site staff on a daily or weekly basis, as needed. On occasion, EOIR staff are
brought into these conversations as well.)
Monthly Conference Calls

Vera also coordinates monthly conference calls that involve personnel from all LOP providers
and EOIR staff. These monthly conference calls provide an opportunity for providers to update
Vera Institute of Justice 26

the entire group on any new developments, problem solve with regard to site­specific challenges,
and share strategies for implementing the LOP and collaborating with local detention facility,
ICE, or EOIR staff. Monthly conference calls also give Vera and EOIR a chance to update site
staff on changes that affect the program as a whole.
Annual Peer­to­Peer Retreat

Each year, Vera organizes a peer retreat for LOP providers. The peer retreat provides an
opportunity for skill development, ongoing training, and collective problem solving. By coming
together at one location, participants can share effective practices (such as techniques for
improving orientation sessions and workshops), discuss challenges they face, hear feedback on
the program from government stakeholders and former LOP participants, and manage the stress
that often comes with work in detention settings.
In preparation for each annual retreat, Vera convenes a retreat planning committee composed
of EOIR personnel and staff from several LOP sites. Through surveys of other LOP providers,
the committee identifies key areas of interest and need. Vera staff then choose a retreat location
in the vicinity of one of the LOP sites, which gives other providers an opportunity to visit and
observe the site. The site visits proved so effective in promoting an exchange of ideas among
providers that in 2006, Vera implemented a series of on­site training programs (see below).
On­Site Training Program

The on­site training program, which was introduced in 2006, evolved from the annual site visits
in LOP peer retreats. The on­site training program enables staff who are new to the LOP to
receive guidance from more experienced providers in the intervals between annual retreats. On­
site training programs are hosted by LOP providers and facilitated by Vera staff.
Materials Development

Vera program managers work closely with EOIR and LOP providers to ensure that the written
materials distributed by LOP providers are carefully developed and are reviewed and approved
by EOIR. In 2006, Vera staff worked with LOP providers to translate existing documents into
languages other than English and Spanish.

Who Does the Legal Orientation Program Serve?
As the use of immigration detention—and bed space in many of the facilities that host the
LOP—has expanded, the program has continued to serve more and more people each year.
Between its creation in mid­2003 and the end of September 2007, the program had served more
than 100,000 detained persons. However, as the expansion of detention has outpaced the
expansion of funding for the Legal Orientation Program, the numbers of people receiving LOP

Vera Institute of Justice 27

services represents a shrinking percentage of the overall detained immigration court population
each year.
As discussed above, Vera managers monitor LOP providers on a monthly basis to improve
program services. One way Vera does this is by studying the numbers and demographic profiles
of participants who receive each type of program service at each LOP site. In this section, we
discuss this process in detail, using data from calendar year 2006.
Group and Individual Orientation Participants

Figure 2 shows program statistics from calendar year 2006. For each subcontractor (and
corresponding immigration court), we tracked the total number of people who participated in
each of the LOP services: group orientations, individual orientations, workshops, and pro bono
referrals.29 Columns 1 through 3 show, respectively, the total number of group orientations
conducted; the total number of group orientation participants; and the average number of
participants per group orientation. As the last row of column 2 illustrates, 25,852 people attended
LOP group orientations in 2006.30
As columns 4 and 9 illustrate, the number and percentage of those who participated in both
group and individual orientations varied widely from site to site. Seven percent of LOP
participants attended individual orientations at Port Isabel, whereas 43 percent attended
individual orientations at Denver. This is a result of differences in both population and program
service models across sites: Ideally, participants self­select for the individual orientation based
on the knowledge they receive in the group orientation. It appears that this did occur in Port
Isabel, where, in 2006, a large number of detained persons who attended the LOP were not
involved in immigration court proceedings and thus would not have needed individualized pro se
assistance.
This is one example of how program statistics need to be combined with qualitative data and
feedback from each site in order to determine how well the LOP is working: in other cases, low
rates of participation in the individual orientation reflect the fact that few detainees at those
locations have viable claims to relief and thus do not have a need individual orientations.
Workshop Participation

Columns 6 and 7 (in Figure 2) illustrate the number of pro se workshops and the number of
individuals who participated in those workshops in 2006. Fewer than 5 percent of LOP
participants (981 out of 25,111) benefited from pro se workshops in 2006. Two conclusions can
be drawn from this figure: the percentage of detained persons who need the sort of in­depth
29

The LOP expanded to six new sites in late 2006, but those sites were only required to report 2006 data for
December of that year. Therefore, numbers do not yet illustrate the increase in numbers served that accompanied the
expansion.

30

The last row in column 8, “Total Unique Participants for All LOP Sessions,” shows that a total of 25,111 unique
persons participated in the LOP in 2006. The reason for this disparity is that a small number of detainees attended
more than one group orientation.
Vera Institute of Justice 28

assistance offered by pro se workshops is relatively small, and LOP providers have a limited
capacity to offer these workshops, probably due to a lack of funding.
Pro Bono Referrals

In 2006, only a small percentage of LOP participants were referred to pro bono representatives
by LOP providers. As column 5 shows, 257 pro bono referrals were made in total—
approximately four referrals per site each month over the course of the year.31 Although this may
seem like a small number, one should consider the fact that very few detained persons (less than
20 percent in our analysis) pursue forms of relief other than voluntary departure.
Figure 2: Legal Orientation Program Services Provided, January 1 – December 31, 2006
1

2

3

4

Unique
Average Number
Total Group
of Participants Per Individual
Presentation
Orientation
Group
Participants
Participants
Presentation

5

6

7

8

9
Percentage of Group
Participants with
Individual
Orientations*

Number of
Pro Bono
Referrals

Number of
Workshops

Number of
Workshop
Participants

Total Unique
Participants
for All LOP
Sessions

60

48

198

2,770

43%

681

8

30

102

3,555

19%

2,735

54

16

114

3,593

74%

11

62

142

3,452

23%

Site Name

Court
Hearing
Location

Number of
Group
Presentations

Denver

WSI

193

2,605

13

El Paso

EPD

160

3,759

23

Eloy

EAZ

252

3,877

15

Mira Loma

LAN

144

3,744

26

941

Port Isabel

PIS

239

8,577

36

637

49

24

118

8,340

7%

Seattle

AIR

411

2,598

6

1,038

64

152

272

2,724

29%

1,276

Batavia

BTV

15

142

9

109

5

2

9

144

75%

Houston

HOD

17

162

10

38

1

0

0

157

24%

Newark

NEW

10

79

8

16

3

0

0

81

7%

84

14

13

0

0

0

74

12%

Laredo

LAR

6

San Antonio

SAD

9

79

9

43

2

1

3

81

52%

San Pedro

SPD

6

90

15

58

0

6

23

84

65%

York

YOR

4

56

14

13

0

0

0

56

23%

TOTAL

N/A

1,466

25,852

18

7,598

257

341

981

25,111

27%

Notes: Sites above the bar reported data from January to December 2006. Sites below the bar reported data only in December
of 2006.
* The percentages are calculated based on the unique participants for regular individual orientations and the unique
participants for group presentation.

Is the LOP Reaching All Detained Persons in Immigration Court Proceedings?

EOIR requires LOP providers to make their services available to anyone who is 1) detained at
the detention facility served by that provider, and 2) involved in removal proceedings at the
immigration court associated with that detention facility. These services are to be provided
regardless of whether an individual has legal representation.
31

Because researchers did not collect the A­numbers (identification numbers) for referrals that were made during the
first two months of 2006, there is a small degree of uncertainty in this figure.
Vera Institute of Justice 29

To measure the LOP’s success in reaching detained persons, we conducted two separate
analyses of data submitted by LOP providers in Denver, Eloy, El Paso, Mira Loma, Port Isabel,
and Seattle for calendar year 2006. In both analyses, we matched program data with different
sets of immigration court records.
The first analysis examined how many participants had active immigration court cases (as
opposed to closed cases or no case). This gives us a sense of how many LOP participants are
involved in pending removal proceedings. The second analysis sought to determine how many
individuals with active court cases received LOP services. This tells us how effective the LOP is
at reaching people in removal proceedings at the sites it serves. We discuss both of these
analyses in greater detail below.
LOP Participants with Active Court Cases

We began by matching the A­numbers of LOP participants, as supplied by LOP providers in
Denver, Eloy, El Paso, Mira Loma, Port Isabel, and Seattle with immigration court records. As
Figure 3 illustrates, not all participants had active immigration court cases at the time they
received LOP services. At four out of six LOP sites, more than 90 percent of participants had
active immigration court cases. At the two remaining sites (Port Isabel and Mira Loma), the
percentage of participants with active cases was much lower: 76 percent and 49 percent,
respectively. In total, only 73 percent of those who took part in the LOP in 2006 had active
cases.
In order to learn more about the 27 percent of LOP participants who did not have active
immigration court cases, we examined the data for LOP participants who did not have active
immigration court cases but whose A­numbers appeared in old court cases. As shown in Figure
3, nine percent of all LOP participants fit this description. Interviews with LOP providers suggest
that many of these people were apprehended as a result of stepped­up ICE enforcement and
detained under prior removal orders, though we cannot confirm this view without access to ICE
data.32 In addition, some of the participants with old cases may have been detained subject to
post­removal order review.33 Others may have been transferred to immigration detention from a
state or federal prison after having their case concluded under the Institutional Hearing
Program.34 It is noteworthy that at Port Isabel, where only 49 percent of LOP participants had
32

Immigration court records do not track anything about the subsequent custody status of persons whose cases have
already been concluded by either an immigration judge of the Board of Immigration Appeals.
33

As noted above, post­removal order review is a multi­step process that evaluates the likelihood of removal in
circumstances where a person is detained for a lengthy period after a removal order becomes final and the possibility
of release if removal is not likely in the foreseeable future. Delays in removal are often caused by the reluctance or
refusal of the receiving country to issue the documents and permission necessary to effect a removal.
See “Immigration Court Process in the US,” US Department of Justice, EOIR, April 28, 2005,
http://www.usdoj.gov/eoir/press/05/ImmigrationCourtProcess2005.htm
34

The Institutional Hearing Program identifies individuals serving criminal sentences and seeks to complete
removal proceedings while the individual is serving the criminal sentence. If a removal order is issued, the
individual will be removed from the United States promptly following release from prison.
Vera Institute of Justice 30

active immigration court cases in 2006, 16 percent had prior immigration court cases—three to
five times the figure at other sites.
As the last column in Figure 3 shows, 18 percent of all LOP participants had A­numbers that
could not be matched with immigration court records. (As is evident from Figure 3, a significant
portion of these individuals were detained at either Mira Loma or Port Isabel.) Interviews with
LOP providers and other stakeholders suggest that many of these individuals were subject to
expedited removal or other types of administrative detention that did not involve immigration
court proceedings.35 It is also likely that data entry errors at LOP sites contributed to the number
of participants whose A­numbers could not be matched with immigration court records.
The significant difference between the numbers in column 3 for Port Isabel and Eloy or El
Paso is noteworthy, given that all three sites are located near the Mexican border in regions that
account for large numbers of apprehensions by federal immigration authorities. In large part, this
discrepancy can be attributed to the fact that in 2006, LOP providers at Eloy and El Paso focused
on detained persons with active immigration court cases (cases that appear on the docket lists
available to LOP providers). In contrast, the providers at Port Isabel served all new arrivals to the
detention center.
Serving detained persons who do not have active immigration court cases has pluses as well
as minuses. Some providers believe that providing legal information to all detained individuals is
the best way to make the system more efficient. Indeed, at some sites detention center personnel
and ICE staff have requested that the LOP be offered more widely—to individuals subject to
stipulated and expedited removal proceedings, for example. On the other hand, offering LOP
services to all detained persons may distract from the program’s primary function—namely, to
serve people in removal proceedings who are in need of pro se assistance. Thus far, Vera and
EOIR have worked with each LOP provider to determine which model works best at that
provider’s site. At some sites, providers have a limited scope of action, since logistical
considerations at the local detention facility largely determine whether it is possible to serve
everyone. For example, using the new arrivals list (which typically includes each detained
person’s dormitory assignment) to plan and arrange orientations—thus offering LOP services to
everyone—may be the simplest course of action at a particular facility. At other facilities, it may
make more sense to use names from the court docket, serving only those with active cases. The
issue of whether or not to see new arrivals is one that Vera and EOIR continue to discuss with
ICE personnel and detention facility staff on both a national and local basis.

35

As mentioned above, not everyone who is subject to removal from the United States is entitled to appear before an
immigration judge. The Illegal Immigration Reform and Immigrant Responsibility Act of 1996 (IIRIRA) mandates
that persons who arrive at a U.S. port of entry without travel documents or who present fraudulent documents must
be detained and placed in expedited removal. The expedited removal process allows an immigration inspector to
remove from the United States certain classes of noncitizens who are inadmissible. See “Immigration Court Process
in the US,” US Department of Justice, EOIR, April 28, 2005,
http://www.usdoj.gov/eoir/press/05/ImmigrationCourtProcess2005.htm.
Vera Institute of Justice 31

Figure 3: LOP Participants with Active, Old, and No Immigration Court Cases

at the Time of LOP Services, January 1 – December 31, 2006

1

2

3

Site Name

Court Hearing
Location

Active Court Case
(%)

Old Court Case
(%)

Not Found in Court
Records (%)

Denver

WSI

91

4

5

EPD

91

3

5

Eloy

ELZ

90

5

5

Mira Loma

LAN

76

6

18

Seattle

AIR

5

3

16

35

9

18

El Paso

Port Isabel

PIS

92
49

Total

N/A

73

Active Court Cases That Included LOP Participation

While our first analysis examined LOP participants with cases in the immigration courts—active
cases as well as cases that had been concluded—our second analysis looked at the percentage of
individuals with active court cases at LOP sites who actually received LOP services in 2006. In
this second analysis, we examined data for all individuals who had a case in one of the
immigration courts associated with an LOP site and who were in detention at the time of the
initial Master Calendar Hearing (in calendar year 2006). The results of this analysis tell us how
well the LOP is succeeding in reaching all individuals in removal proceedings.
Figure 4, below, shows the results of this analysis. Column 1 lists the total number of initial
Master Calendar Hearings for “detained cases” (cases that began while the respondent was in
detention) at each of the six immigration courts served by an LOP provider in 2006. Column 2
lists the percentage of respondents from column 1 who participated in the LOP, by site. At
Denver, for example, 80 percent of initial Master Calendar Hearings in 2006 involved persons
who had participated in the LOP, compared with only 21 percent at Eloy. Overall, the LOP
reached only 42 percent of detained persons whose initial hearings were held at immigration
courts served by the LOP in 2006 despite the fact that the program served more people in 2006
than in any other year since the program began.
There are a number of possible explanations for the fact that some LOP sites serve large
numbers of individuals (indicated in Figure 2, column 8) yet nonetheless have low rates of
program participation (indicated in Figure 4, column 2). For one, individuals who do obtain legal
counsel may conclude that it is not in their best interest to participate in the LOP. As column 3
(in Figure 4) shows, 58 percent of all detained respondents at LOP sites did not participate in the
LOP. Also, as illustrated in column 3(2), 5 percent of all detained respondents at LOP sites did
not participate in the LOP and had legal representation. In other words, 9 percent of the
respondents who did not receive LOP did have legal representation. While this figure is not
represented in Figure 4, it can be obtained by dividing the total percentage of persons with
representation and no LOP (column 3[2]) by the total percentage of persons with no LOP
(column 3). While immigration court records do not allow us to determine whether legal
Vera Institute of Justice 32

representation was retained before or after LOP services were offered, it is possible that some of
those who did not participate in the LOP chose not to do so because they were receiving legal
advice from their counsel.
A second possible explanation for the low rates of LOP participation involves stipulated
removal orders. As column 3(1) of Figure 4 indicates, 27 percent of the detained individuals who
attended immigration court hearings at an LOP site but did not take part in the LOP had signed
stipulated orders of removal. One should exercise caution in drawing conclusions about
stipulated removals from this table, though, since it does not represent the total percentage of
detainees at LOP sites with stipulated orders of removal.36 Rather, it simply reveals that 27
percent of immigration court cases at LOP sites involved stipulated removal and no participation
in the LOP. These 27 percent (of all court cases) in turn represent 47 percent of all persons with
initial Master Calendar Hearings at LOP sites who did not participate in the LOP. It is not
surprising that the LOP would fail to serve such individuals with stipulated removals: Because
they do not require a court appearance, stipulated removal cases are typically settled within a few
days and are thus unlikely to be listed on the immigration court dockets given to LOP providers.
Also, people who have already signed stipulated removal orders have, ipso facto, indicated a
desire to be repatriated and may not have any interest in taking part in the LOP.
Column 3(3) of Figure 4 indicates that 25 percent of all detained persons with initial Master
Calendar Hearings at LOP sites did not participate in the LOP, had not signed stipulated orders
of removal, and were not represented by legal counsel. We do not know for certain why these
individuals did not take part in the LOP. However, there are several possible explanations.
Several LOP providers told us that detention facility staff sometimes deny them access to the
facility during facility lockdowns.37 It also happens that detained individuals who are interested
in taking part in the LOP are not given an opportunity to attend; that there are no announcements
about upcoming programs; that announcements are inaudible; that staffing shortages prevent
facilities from providing the necessary logistical support for orientations; and that interested
individuals are overlooked due to poorly organized lists.
It also happens that detained individuals refuse to participate in the LOP. Again, there can be
many reasons for this. An individual may already have representation, as discussed above, or he
36

Our analysis revealed that nearly 25 percent of all completed cases that began “detained” in 2006 were recorded
as stipulated removal cases by the immigration courts. At four LOP sites, the number of stipulated removal cases
was even greater percent: At Mira Loma, for example, 51 percent of all completed cases that started in detention
were identified as stipulated removals, as were 48 percent at Eloy; 43 percent at Seattle; and 39 percent at Port
Isabel. On the other hand, stipulated removals accounted for only 4 percent of all cases beginning in detention at El
Paso and less than 1 percent of cases at Denver. These figures should not be interpreted to mean that one in four
persons in immigration detention facilities signed stipulated orders of removal in 2006 because we are looking at a
cohort of cases that involved initial Master Calendar Hearings in 2006 and a case completion by the time of our
analysis in early 2007. Thus, 25 percent of the completed cases we studied involved stipulated removals. Calculating
the percentage of all cases represented by stipulated removals would require following a cohort of cases until all
cases in that cohort had been completed, which could take several years. When this report was being written (spring
2008), statistical reports produced by ICE and EOIR did not report on immigration court cases or removals in this
way.
37

When this occurs, LOP providers report to Vera the dates when they were unable to provide services.
Vera Institute of Justice 33

or she may be misinformed about the nature of the services provided. (Some detained individuals
have reported being told that they were being assembled for a religious program or a class.) Of
course, some individuals simply may not be interested in participating. Providers also report that
some detained persons are unable to understand the orientation because of mental illness, while
others miss the LOP because they have been sent to “segregated housing” as a result of a
behavior problem or health concerns. Finally, it is possible that some persons who are scheduled
for hearings in the immigration courts in which the LOP operates may not actually be detained in
those facilities. Instead, they may be transported to the immigration court for their hearings. This
practice has become increasingly common in several facilities where the LOP currently
operates.38
Figure 4: Participation in LOP Services at LOP Sites: Percentage of Detained Persons with Initial
Master Calendar Hearings at LOP Sites Who Did and Did Not Participate in the LOP, January 1 –
December 31, 2006
1
Site Name

Court Hearing
Location

2

Total Number
Cases with Initial
of Initial MCH
MCH and LOP
at Each Site

3

3(1)

3(2)

3(3)

Cases with Initial MCH and No LOP

Denver

WSI

2,903

80%

20%

Stipulated
Removal
1%

7%

13%

Eloy

ELZ

11,294

21%

79%

43%

3%

33%

El Paso

EPD

5,481

45%

55%

6%

14%

34%

6,641

32%

68%

42%

3%

22%

2,987

82%

AIR

3,509

60%

18%
40%

5%
20%

5%
3%

8%
17%

N/A

32,815

42%

58%

27%

5%

25%

Mira Loma
Port Isabel
Seattle
Total

LAN
PIS

Total

Represented

Others

After reviewing these statistics, Vera managers met with LOP providers, EOIR staff, and
other program stakeholders to discuss how the LOP might boost participation rates among
detained persons in removal proceedings, as well as how program services might be expanded to
ensure that the program is accessible to all individuals in immigration detention with active
immigration court cases. One question that was raised during these discussions was whether the
LOP was reaching speakers of languages other than English and Spanish. In the next section, in
an effort to shed more light on this question, we discuss what is known about the national and
linguistic backgrounds of LOP participants.
Nationalities and Languages of Cases at LOP Sites

Our analysis of the rates at which different national groups accessed LOP services found that
Mexicans accounted for the greatest numbers of participants, although non­Mexicans had higher
rates of participation at some LOP sites. In 2006, 59 percent of all individuals whose
immigration court cases began while they were in immigration detention and 73 percent of all
38

In future analysis, it may be possible to identify which detainees are housed elsewhere by reviewing the person’s
address recorded by the immigration courts. Vera did not analyze this information in the research reported on here.

Vera Institute of Justice 34

detained at LOP sites were Mexican.39 Both nationwide and at LOP sites, El Salvador,
Guatemala, and Honduras accounted for most of the non­Mexicans (often referred to as “Other
Than Mexicans” [OTMs] in reports on immigration enforcement efforts). Citizens of the
Dominican Republic, China, Cuba, Haiti, and Jamaica each accounted for slightly more than 1
percent of detained persons nationally and less than 1 percent at LOP sites. Caribbean nationals,
who account for large numbers of detained persons in some parts of the country, were less
prevalent at LOP sites in 2006.40
As noted earlier in this report, when EOIR began the program in 2003, program managers
deliberately selected detention locations with predominantly bilingual, English­ and Spanish­
speaking populations. As the program expanded beyond the original sites and into detention
centers with larger numbers of speakers of languages other than English and Spanish, the
program managers tried to increase their ability to serve all detained persons, regardless of their
language spoken. In 2006, LOP served speakers of more than 60 languages. However, at the six
sites that were fully operational in 2006 (Denver, Eloy, El Paso, Mira Loma, Port Isabel, and
Seattle), Spanish speakers continued to account for the overwhelming majority of participants.
Similarly, according to immigration court records that track the language used by each
respondent in the courtroom, Spanish continued to be the predominant language spoken by
respondents with initial Master Calendar Hearings in 2006.41 Nationally, 79 percent of detained
persons with initial Master Calendar Hearings in 2006 were listed by the immigration courts as
Spanish speakers versus 86 percent of cases at LOP sites. In comparison, English speakers
accounted for 14 percent of all detained persons with initial Master Calendar Hearings in 2006
versus 11 percent at LOP sites.
Additional languages were also represented nationally and at LOP sites, although in far
smaller numbers. Mandarin speakers accounted for 1 percent of “detained” immigration cases
nationally versus 0.4 percent of cases starting at LOP sites. All other languages occurred in less
than 1 percent of all initial Master Calendar Hearings nationally and at LOP sites. Creole,
39

These figures will not match statistics provided by EOIR or ICE on all immigration court cases or all persons in
immigration detention or removed from the United States. We are presenting statistics only on persons who had
initial Master Calendar Hearings in 2006 whose cases began “detained.” In other words, we are including persons
who were detained throughout their immigration court cases and persons who were detained but later released; we
are not including persons who immigration court records list as “never detained.” Moreover, we are not reporting on
all “detained” immigration court cases in 2006, but, rather, only those cases that began (had an initial Master
Calendar Hearing) in 2006.
40

Apprehension and detention patterns shift rapidly, and these statistics from 2006 may not accurately depict current
population demographics. Since 2006, the distribution of nationalities at LOP sites has reflected shifting trends in
detention practices.
41

Immigration court clerks use the “language” field of their administrative database to track the language of
interpreter needed at each hearing (or will list English if no interpreter is needed). This means that in most instances
the language in immigration court records will reflect the language that the respondent requested at the last or latest
hearing. However, in some immigration courts, the language field is used to track the language of interpreter needed
for a witness, which may be different from the language spoken by the respondent. There is no way of knowing
when looking at data which languages may be for witnesses instead of respondents, though we were assured by
immigration court administrators that this occurs infrequently.
Vera Institute of Justice 35

Portuguese, Arabic, Korean, Russian, Vietnamese, and French were listed relatively more
frequently than other languages occurring in less than 1 percent of initial Master Calendar
Hearings, nationally and at LOP sites. Despite the relative infrequency with which these other
languages were used in immigration court hearings, LOP providers have nonetheless worked in
close collaboration with Vera program staff to translate LOP materials into languages other than
English and Spanish. Vera program managers continue to prioritize expansion of program
services and creation of program materials to serve as diverse a range of linguistic groups as
feasible.
Because the detained population served by the LOP is continually shifting, largely due to
ICE enforcement and placement decisions (and to a lesser extent fluctuations in migrant
populations), LOP providers cannot always predict the demographics of people in detention from
one week to the next. Experience has shown us that a large enforcement action by ICE can lead
to a sudden influx at detention centers of language groups that the LOP has rarely served before.
Similarly, at sites where the group orientation is offered to new arrivals (many of whom may be
subject to expedited removal), there may be times when large numbers of LOP group orientation
participants do not have cases pending in the immigration court. This as well might fluctuate
considerably week­by­week or month­to­month.

Nationwide Trends for Cases Beginning in Detention
As the discussion above illustrates, in addition to assessing whether the LOP is meeting its
objectives, we also analyzed the relationship between the LOP and larger trends and patterns in
immigration detention and the immigration courts. This enabled us to better understand the
broader context in which the LOP is being implemented.
A recent report from the Government Accountability Office notes that over the past few
years EOIR has experienced a 44 percent increase in the number of new cases brought by ICE
against noncitizens, in part a result of enhanced border and interior enforcement activities.42
Analysis of ICE data conducted by other researchers shows a 22­percent increase in
apprehensions between 2002 and 2005.43 Additionally, ICE reported an increase in immigration
detention beds of 6,300 in fiscal year 2006, bringing the total number to 27,500 nationwide by
2007. In addition to increases in apprehensions and detention beds, ICE reported housing
237,667 immigration detainees in 2005—an increase of more than 28,000 people since 2001.
Immigration enforcement officials project that these numbers will continue to rise.44

42

Government Accountability Office. Executive Office for Immigration Review: Caseload Performance Reporting
Needs Improvement, August 2006.
43

Batalova, J., and D. Konet (March 2007), Spotlight on Immigration Enforcement in the United States. Migration
Information Source. Accessed June 11, 2007
http://www.migrationinformation.org/USFocus/fisplay.cfm?ID=590#top
44

Detainees under Bureau of Immigration and Customs Enforcement (ICE) Jursidiction. Sourcebook of Criminal
Justice Statistics Online. Accessed June 13, 2007 from http://www.albany.edu/sourcebook/pdf/t6612005.pdf
Vera Institute of Justice 36

Immigration Charges Among Detained Persons in Removal Proceedings

EOIR databases do not track immigration or residency status. However, from the charges listed
on the notice to appear, which are recorded in immigration court databases, we can discern
which removal cases were filed by ICE based on non­criminal immigration status violations, and
which were brought based on past criminal convictions. Understanding the types of immigration
charges facing LOP participants—and detainees nationally—can help us assess if LOP providers
are focusing their orientation presentations on topics affecting the greatest percentages of
participants and can further contextualize program service numbers.
The overwhelming majority of all new detained removal proceedings in the immigration
courts in 2006 were for immigration status violations not linked to prior criminal convictions. In
other words, most detainees with new immigration court cases were accused of immigration
violations that did not render them “criminal aliens” or “aggravated felons.” Detained persons
charged with immigration status violations were predominantly accused of unlawful entry or
presence without admission or parole under INA §212(a)(6)(A)(i). In 2006, LOP participants
were more likely to be charged pursuant to this provision than other detainees in removal
proceedings nationwide (61 percent of all LOP cases versus 45 percent of all comparison group
cases). This is not surprising given that a few of the original LOP sites in this evaluation operate
in large facilities near the southern U.S. border. This was particularly true of the programs in
Port Isabel and El Paso in 2006, although demographics here have since shifted and more
persons with criminal convictions are being transferred to these facilities from elsewhere in the
U.S.
Detained LOP participants and comparison groups alike were charged as “aggravated felons”
in fewer than 10 percent of the immigration court cases we analyzed. Most of the detained
persons in new removal proceedings in 2006 were accused of non­criminally related immigration
violations. However, the charge for aggravated felony convictions, INA §237(a)(2)(A)(iii), was
the second most frequent charge for both LOP and comparison groups (7 percent for LOP
participants versus 9 percent of persons in comparison groups).
Applications for Relief from Removal

LOP providers can enhance the effectiveness of their orientations by understanding trends and
patterns in the immigration charges that their participants face. However, an understanding of
any potential impact of the LOP on rates of applications for relief from removal must consider
these rates in relation to both representation rates and the specific charges. For example, were we
to see 99 out of 100 detainees charged with an offense that makes them ineligible for a particular
form of relief, we would not expect to see these individuals pursuing such relief if they were
appropriately oriented to eligibility requirements. This section of this report discusses the trends
and patterns we observed over time, nationally and at LOP sites, in order to contextualize our
evaluation of LOP’s impact.
LOP providers report to Vera only on possible forms of relief or defense that individual
orientation participants report they might pursue. Vera does not typically do much analysis of this
Vera Institute of Justice 37

data since it only tells us the forms of relief persons attending individual orientations might pursue.
Nevertheless, one noteworthy trend was observed by Vera program managers. In 2006, the LOP sites
reported seeing 129 persons at individual orientation sessions who were planning to pursue claims to
U.S. citizenship. Because LOP providers do not discuss the details of individual cases with
orientation session attendees, we do not know if those claims to citizenship are derivative, acquired,
or by birthright. Nor do we know the outcomes of these claims. We are simply able to report that 129
persons who attended LOP orientations in 2006 said that they planned to pursue claims to
citizenship.45
Figure 5, below, shows rates of application for relief among persons whose immigration
removal cases began while detained, from 2000 to 2005. The trend lines show that since 2000,
there has been a decline in the percentage of detained persons in removal proceedings pursuing
relief at both LOP and non­LOP sites.46 This trend is likely related to shifts in the detained
population and the types of relief those detained persons are statutorily eligible to receive.
However, the decline in relief application rates for cases that began at LOP sites has been sharper
than that of cases that began at comparison sites. At other immigration courts around the country
with comparable case volumes, relief application rates followed a different pattern, increasing
slightly around the time the LOP began. While these patterns do not identify or isolate what role,
if any, the LOP might have played in affecting these rates, they do show that immigration courts
with the LOP followed a different pattern than other immigration courts with a comparable
volume of cases. It is possible that the LOP was responsible for some of what we observed.
However, the comparative designs we are using make it just as plausible to argue that these
differences are the product of the very factors that led to the placement of the LOP in certain
locations in the first place—the concentration of certain types of detainees at certain detention
centers (such as recent entrants into the U.S. housed in the detention facilities in the southern
border region) or the low rates of legal representation at LOP sites. Further research should be
conducted to determine which of the rival interpretations is correct.
We observed no remarkable differences in the distribution of relief application types between
LOP and comparison groups, with the exception that immigration courts with the LOP received a
smaller percentage of asylum applications than courts with a comparable volume of cases.
Again, this is not surprising given that one of EOIR’s priorities for the programs at Denver, El
Paso, Eloy, Mira Loma, Port Isabel, and Seattle was that they be placed in sites with low
representation rates and serve a predominantly bilingual (English/Spanish) population. Locations
with large volumes of asylum cases tend to have a more linguistically diverse population of
detainees and higher rates of representation.

45

The number of claims to citizenship identified by the LOP providers increased to 322 in 2007. A table showing
the monthly breakdown of such claims by LOP site can be found in Appendix I.

46

These lines show rates of relief application; in other words, we are not showing the total numbers of persons who
pursued relief in the immigration courts, but the percentage of the total number of persons with active immigration
court cases whose cases involved relief applications. This means if the number of court cases changed from year to
year, but the same percentage of cases involved relief applications each year, we would see a straight line.
Vera Institute of Justice 38

Figure 5: Percentage of Relief Applications for Detained Cases at LOP and Comparison Hearing

Locations from 2000 to 2005.


As Figure 5 illustrates, overall rates of application for relief fluctuated between 20 percent
and 30 percent at LOP sites and other immigration court hearing locations from 2000 to 2005. In
other words, 70 to 80 percent of all cases initiated during this time involved no relief
applications. This point underscores a key component of the group orientation. LOP providers
use the group orientation to provide information to participants that enables them to make
informed decisions about their potential eligibility for relief (or defenses such as immigration
benefits, which are not heard in the immigration courts and not tracked in immigration court
records). Because such a high percentage of cases involves no relief from removal, LOP
providers spend a substantial amount of time in the group orientation discussing topics of
relevance to persons with no relief possibilities, such as ways to accelerate repatriation through
the removal process and consequences of unlawful reentry following removal.
Of note is the fact that while the total percentage of detained persons with immigration court
cases involving relief applications has decreased over time at LOP sites, the distribution of the
types of relief being pursued has not changed significantly. Though we do not show this
information here, we observed that from 2000 to 2005, at LOP sites, only about 10 percent of all
relief applications involved relief other than voluntary departure, both before and after the LOP.
Similarly, I­589 applications consistently comprised less than 5 percent of all relief applications
at LOP sites, before and after the LOP started. In other words, while a shrinking percentage of
detained persons with immigration court cases pursued relief, the distribution of relief
applications among those pursuing relief remained the same.47 These patterns held true in 2006.
47

Throughout this report we report on I­589 applications instead of naming these applications as “asylum.” When an
I­589 application is filed, court clerks cannot always tell if the applicant is seeking asylum, withholding, Convention
Against Torture (CAT), or some combination of these forms of relief. To avoid mislabeling, we are reporting on the
entire application as opposed to claims relating only to asylum, withholding, or CAT. Readers should not infer that
I­589 means asylum was pursued. At some LOP sites, providers report seeing participants who qualify for
withholding or CAT but do not meet statutory requirements for asylum, often because of criminal convictions or
other statutory bars.
Vera Institute of Justice 39

Figure 6, below, represents the rates of relief applications for all completed immigration
court cases that had an initial Master Calendar Hearing between January 1 and August 31, 2006.
This figure shows that LOP participants followed patterns of application for relief that were
consistent with rates in other immigration courts around the country. While a slightly higher
percentage of LOP participants’ cases involved no relief applications (69 percent versus 65
percent for comparison cases), LOP participants pursued I­589 applications at slightly lower
rates (4 percent versus 7 percent for comparison groups). Rates of application for voluntary
departure were 21 percent among LOP participants and 22 percent among comparison cases,
while all other applications for relief had a combined average rate of 6 percent among LOP
participants and 5 percent at comparison group sites.48
Figure 6: Relief Application Rates for Completed Cases that Began in Detention, January 1 –
August 31, 2006
LOP (N=7,528)

Comparison (N=30,728)

Relief Application Type
No Application

N

%

N

%

5,184

69%

19,978

65%

Voluntary Departure Only

1,563

21%

6,904

22%

I­589

318

4%

2,215

7%

Other Application Combinations

463

6%

1,631

5%

Nationwide Case Outcomes—Rates of Grants of Relief and Orders of Removal

When persons in removal proceedings do not file applications for relief from removal and have
not pursued any other defenses, they will almost never be granted permission to remain in the
United States. Not surprisingly, then, given the low rates of applications filed for relief from
removal shown in Figure 6, very few of the cases we analyzed resulted in grants of relief or other
decisions that resulted in permission to remain in the U.S. lawfully. In our analysis of 91,747
completed cases nationwide that began in detention in 2006, we observed that 87 percent of

48

Though we intended to analyze each type of relief application separately, rates of relief application for most forms
of relief were so low that we did not have enough cases to study. For example, of 44,054 concluded cases we studied
(cases that began in detention between January 1 and August 31, 2006), only 3.6 percent of all cases included
applications for 240A cancellation of removal for lawful permanent residents, 0.9 percent included applications for
240B cancellation of removal for non­residents, 0.9 percent included 245(i) applications, 0.5 percent included
212(c) applications, and all other application types not listed here (excluding asylum and voluntary departure)
occurred in less than 0.5 percent of all cases. As a result, in the analysis, we grouped cases into the four categories
shown in Figure 6: no relief application, voluntary departure only, I­589, and the category “other application,”
which combines all other forms of relief such as 240A and 240B and any combinations in which multiple
applications occurred. Where possible, we analyzed cases including these forms of relief separately, but given the
relatively few cases in which they occurred, it was often impossible to do so. As noted earlier, when reporting on
rates of relief application, we include voluntary departure since it is considered a form of relief by the immigration
courts despite the fact that it does not result in permission to remain in the United States. Many immigration
practitioners do not consider voluntary departure a form of relief. In some discussions we separately identify
voluntary departure to show the low rates of relief granted when voluntary departure is analyzed separately. In those
instances, we indicate that voluntary departure has been separated from the overall category of relief.
Vera Institute of Justice 40

completed cases resulted in orders of removal (Figure 7, fourth bar).49 The 13 percent of cases
that began in detention and did not result in removal in 2006 fell into the following groups: 3
percent resulted in grants of relief, immigration judges terminated 2 percent, 1 percent resulted in
administrative closure, and most of the remaining 7 percent were granted voluntary departure.
Figure 7 below shows nationwide outcomes (as percentages) for completed cases that began
in detention in 2006. The first bar shows that the overwhelming majority of persons whose
immigration removal proceedings begin while they are detained are not released from detention
(93 percent) and remain detained at the time the immigration judge issues a decision in the case.
Of all completed cases, 77 percent involved no application for relief (second bar). Inversely, 23
percent, less than 1 in 4, involved applications for relief from removal, including voluntary
departure. The third bar shows that 86 percent of these completed cases involved no legal
representation at any point in the case, prior to the appeals process, which is not reported on here.
Finally, as discussed above, low rates of release, low rates of relief application, and low rates of
representation combine to contribute to an overall national rate of removal of 87 percent in the
2006 completed case group. As discussed above, the inverse of this 87 percent includes 3 percent
that resulted in grants of relief and 2 percent that resulted in termination by immigration judges.50
The inverse group also included 7 percent that involved grants of voluntary departure and 1
percent that resulted in administrative closure.51
Figure 7: National Averages for Completed Cases Beginning in Detention in 2006

93%
77%

NO RELEASE

NO RELIEF
APPLICATION

86%

87%

NO REPRESENTATION

REMOVED

Note: Total number of cases is 91,747.

When interpreting findings about the impact of the LOP in the section that follows, it is
important to remember that the LOP is operating in the context of these outcomes nationwide
and that LOP services are structured according to the reality that the overwhelming majority of

49

This includes stipulated removals. In 2006, 22 percent of all cases that began in detention were coded by the
immigration courts as receiving orders of stipulated removals.

50

Termination occurs when the immigration judge determines that the government cannot sufficiently prove its case
that a person should be removed from the country.

51

Administrative closure takes a case off an immigration judge’s calendar but does not result in a decision on the
case.
Vera Institute of Justice 41

detained persons in removal proceedings do not pursue defenses or relief from removal in the
immigration courts, do not retain legal counsel, and do not receive grants of relief. It is in this
national context that the LOP strives to improve efficiencies by educating those small numbers
of detainees with viable defenses or claims to relief that they are eligible to pursue these defenses
or claims and helping them understand how to do so. Though our findings on the impact of the
LOP focus on subgroups of persons released from detention, retaining legal counsel, or pursuing
or granted relief, readers should keep in mind that these subgroups represent a very small
percentage of detained persons in removal proceedings in the immigration courts. Since not all
detained persons are in immigration court proceedings and we are only reporting on persons with
immigration court cases who were detained at the time their immigration court cases began in
2006, the percentage of all detained persons nationwide who are removed from the United States
is actually greater than 97 percent.52 Given that most detained persons in removal proceedings
will not be granted relief from removal that allows them to remain lawfully in the United States,
the LOP group orientation devotes substantial time to providing information that enables
detained persons to make informed and timely decisions about ways to accelerate repatriation
and that educates them about the penalties attached to re­entering the United States unlawfully.
Though modest when expressed as a percentage, the actual number of people with a valid
defense or claim to relief is substantial, and their legal claims significant.
The next section reports on our observations on the impact of the LOP up to this point. In
Section IV, we discuss recommendations and next steps based on our discussion of the process
and performance of LOP as well as the impact/outcomes of the LOP presented in this report.

52

ICE does not report publicly on the total numbers of immigration detainees in removal proceedings versus in
detention for other reasons. As a result, we are unable to calculate the exact percentage of detained persons who
were eventually removed/deported. Just as rates of removal orders for detained persons with immigration court cases
should not stand in for rates of removal orders for all detained persons nationwide, similarly, we cannot assume that
removal orders resulted in actual removals. ICE does report on the total numbers of removals each year, and while
most detained persons with a final order of removal are removed, a small percentage are not. Finally, we did not
analyze rates of removal orders for persons in removal proceedings who were never detained.
Vera Institute of Justice 42

III. Measuring the Impact of the Legal Orientation
Program
As noted at the beginning of this report, the primary objectives of the LOP are twofold: to
improve legal access for detained persons in removal proceedings by providing impartial,
accurate orientations to the immigration court process and providing detainees with information
to help them determine how to proceed in immigration court; and to improve efficiencies in the
immigration detention and immigration court process for detained persons in removal
proceedings. Efficiencies may be defined as the best possible allocation of resources (maximum
benefits for minimum costs) or as enhancements of systems and processes that enable them to
work more smoothly.
In addition to studying the LOP’s progress in meeting the program’s objectives, Vera
researchers considered the extent to which the LOP might also assist EOIR in meeting some of
the agency’s priorities as outlined in its 2005­2010 Strategic Plan. In particular, we focused on
three agency­wide objectives described in the EOIR Strategic Plan that intersect with the
objectives of the LOP. First, objective 1.1 of that plan states that EOIR “must eliminate case
backlogs by the end of fiscal year 2008” and “must render ‘expeditious decisions’ and continue
to reduce ‘frivolous’ applications.” Objective 1.2 of the plan commits the agency to “implement
improved caseload management practices,” which includes “studying failure to appear rates.”53
Finally, the plan notes that, “a longstanding area of concern is the large number of unrepresented
aliens in immigration proceedings.” In response, EOIR’s plan commits the agency to “encourage
pro bono representation,” noting that, “effective representation can add value to the adjudicative
process.”54 In the discussion that follows, we detail ways in which the LOP is meeting its stated
objectives and may be an efficient and cost­effective method of helping EOIR meet some of its
agency­wide objectives.
In order to gauge the impact of the Legal Orientation Program, Vera researchers analyzed
program and immigration court data, talked to program stakeholders, and reviewed literature
from other studies of the immigration courts and pro se education programs, specifically those
focused on legal rights information or carried out in prison or detention settings. As detailed at
the beginning of this report, EOIR requested that Vera analyze any potential impact of the LOP
on legal representation, case outcomes, case processing time, relief application rates, in absentia
rates, and stakeholder and detainee satisfaction with the program. This report presents key
findings from the research to date.
In Section II of the report, we presented observations from analyses of program service data
to determine if the program is functioning as intended and who the program is serving. We also
presented an analysis of administrative data from the immigration courts identifying trends and
patterns nationwide at immigration courts in the years immediately before and after the LOP
53

U.S. Department of Justice, Executive Office for Immigration Review, Fiscal Years 2005­2010 Strategic Plan,
September 2004.
54

Ibid.
Vera Institute of Justice 43

began. In this section of the report, we continue to discuss observations of trends and patterns
before and after the LOP began. We also present observations from analyses in which we
matched program service and immigration court administrative data in order to observe
•	 LOP participants’ cases in the immigration courts and to study differences or similarities
between LOP participants and other detained persons with active immigration court
cases;
•	 any differences among LOP participants based on level of LOP participation (group
orientation, individual orientation, workshop); and
•	 potential relationships between the LOP and other variables that may have an impact on
immigration court cases.
We supplement these observations with findings from qualitative interviews with LOP and
immigration court stakeholders.

Analyzing Trends in Immigration Court Data
As described previously, Vera built a program service database (LOPster) that allows LOP
providers to track and report to Vera researchers data on all persons who participate in LOP
services. As the LOP contractor, Vera is provided with confidential access to immigration court
records for the purposes of tracking LOP participants’ immigration court outcomes. As a result,
Vera researchers are uniquely able to track which cases in the immigration courts correspond to
LOP participants. Because of the way we collect data, we are also able to identify the levels of
intensity of LOP services received by various participants (group orientation, individual
orientation, self help workshop, pro bono referral). Vera reports to EOIR monthly program
statistics, in aggregate and stripped of any individually identifying information. When monthly
data are received by Vera, researchers organize and store these data in a relational database that
allows us to easily merge these data files with those received from the immigration courts (see
Appendix II).
When determining which cases to include in our analysis of the LOP’s impact, we first
conducted a series of analyses that studied changes in the immigration courts in the years
immediately before and after the implementation of the LOP, from 2000 to 2005. These analyses
did not merge LOP data with EOIR data. Rather, we studied only immigration court data in order
to observe trends and patterns nationwide at the Batavia, Denver, Eloy, El Paso, Mira Loma, Port
Isabel, and Seattle immigration courts and at 33 other immigration courts with comparable
numbers of cases.55 In this analysis, we plotted immigration court data in the years immediately
55

In this analysis, we omitted immigration court hearing locations with small caseloads, primarily because case time
and variables such as rates of application and representation could not adequately be compared to hearing locations
that see as many as several thousand cases per year. We also omitted all hearing locations that are used only for
juvenile cases, as well as hearing locations used exclusively for televideo hearings or the Institutional Hearing
Program. As a result, we were left with 33 hearing locations in addition to the six LOP sites. Finally, in some of our
Vera Institute of Justice 44

before and after the start of the LOP to see if we could identify any “big­picture” trends that
might point to differences in representation rates, types of relief sought, case outcomes, and case
processing times for persons whose cases began at LOP sites. This historical and comparative
analysis allowed us to examine rates (comparative magnitudes—i.e., one line on a graph is
higher than another) and trends (comparative directions—i.e., one line on a graph is moving up
and the other is moving down).
At the time we began the pre­ and post­LOP analysis of immigration court data, Vera was
only beginning to collect and organize LOP participants’ Alien identification numbers (A­
numbers), unique identifiers that allow us to track individual cases across multiple datasets.
Since we were not yet able to match LOP participants’ A­numbers with court data and, thus,
identify which immigration court cases corresponded to LOP participants, we did not study data
at an individual level. As a result, the before and after analysis does not allow us to make
conclusive statements about the impact of the LOP. Even without being able to make definitive
statements about the LOP, the historical analysis helped us narrow our focus and provided a
comparative, national context in which to situate further analysis. This also helped us develop
questions for LOP stakeholders about differences in immigration court procedures and their
explanations for the patterns and trends we observed. The historical analysis was only a first (but
important) step that, combined with stakeholder interviews and input from experts in the field,
helped us refine the focus when conducting analysis of individual case outcomes.

Tracking LOP Participants in the Courts
After analyzing macro­level immigration court trends, we matched program data collected by
LOP providers with immigration court data. This enabled us to track case outcomes and make
more conclusive statements about how LOP participants fared in the immigration court process
in comparison with detained individuals who did not receive LOP services. Because some of the
six new LOP sites began offering services in September 2006, we were concerned that including
those sites as LOP sites might leave too few non­LOP cases for meaningful analysis. Our further
analysis of the cases showed that more than 85 percent of the cases with an initial Master
Calendar Hearing in the first eight months of 2006 had been completed at the time of our initial
examination of the data and that the number of cases in this sample was sufficiently great for
meaningful analysis. We thus determined that we would focus our comparisons on the eight­
month period of time from January 1 to August 31, 2006.

analysis of LOP sites we included Batavia. Batavia began operations in mid­2003 like some of the other original
sites, but ceased operations for one year from 2005­2006. Because Batavia was operating in 2005, which was the
end date for our analysis, we included it in the pre­ and post­LOP study. Immigration court hearing locations
included in the pre­ and post­ LOP analysis included courts identified with the following hearing location codes:
FLO, ELXC, CCA, SDE, TUC, KRO, HOD, SPD, YOR, BDC, SAD, LVG, KAN, ATD, SFD, DAD, NEW, OMA,
SAJ, BLM, WAS, CHD, NOO, HAS, ELZ, BOS, PDT, NYD, JAM, MEM, HAR, ELP.
Vera Institute of Justice 45

Comparisons Between LOP Participants and Other Detained Persons
When comparing LOP participants to other groups of detained persons with immigration court
cases, we created various subgroups according to characteristics that our analysis showed might
impact an immigration court case, including the following information tracked in immigration
court records:
•	 Type of relief from removal sought in the immigration courts;
•	 Custody status (detained or released);
•	 Legal representation status (represented at some point in the immigration court case
versus never represented);
•	 Charges brought by ICE on the notice to appear (criminal versus non­criminal

violations);

•	 Nationality and/or language used in immigration court by the detained person;
•	 Immigration judge and/or immigration court (used for internal purposes to check for
administrative and procedural variation that might distort findings); and
•	 Level of LOP service (no participation in the LOP versus participation in the LOP;
intensity of LOP service).
For each analysis, we divided subgroups by at least representation, custody status, and type of
relief from removal sought in the immigration courts. This enabled us to avoid comparisons
between persons pursuing radically different forms of relief, such as voluntary departure and
asylum/withholding/CAT; it also ensured we did not conflate represented cases with those that
were heard pro se. We did not, however, group by nationality, language, or immigration charge
for every analysis we ran at this stage.

Qualitative Interviews
In order to contextualize quantitative observations, we integrated qualitative research into our
evaluation. Using multiple methods of analysis or “triangulation” of methods ensures that even
when researchers cannot observe any measurable program impact in one form of analysis, they
have methodological checks and balances that provide a broader context for understanding
results. Some outcomes are simply impractical to measure through quantitative data, and time
constraints combined with the challenges accompanying research with detained, transient, and
multilingual populations ruled out methods such as large sample surveys. Moreover, very few
studies have produced validated research instruments that can be used to measure the success of
one­time legal interventions, particularly among incarcerated programs. We therefore opted to
conduct qualitative interviews that, while not necessarily generalizable, complement,
contextualize, and give greater depth to quantitative findings. We will draw on these interviews

Vera Institute of Justice 46

as we work with LOP program managers to create and test additional research instruments to
measure the success of legal interventions such as the LOP.
Qualitative interviews can also help us understand, for example, how terms like “efficiency”
and “success” are defined by different LOP stakeholders. Some detainees might evaluate success
by whether they felt empowered to speak in immigration court or not, regardless of the case
outcome, or they might believe the most important outcome is receiving a lower bond at a bond
redetermination hearing, even if they are unable to afford the new bond amount. These are not
the same standards of success defined by LOP providers, immigration court managers,
immigration judges, immigration enforcement, or detention center personnel. Our research takes
into consideration the importance of the LOP to each set of stakeholders in order to show the
potential relevance of quantitative results—such as shorter case processing times—for each of
these groups of stakeholders.
In the pages that follow, we present key findings on the LOP’s impact and discuss their
implications.

Key Findings
Below we present key findings on the relationship between the LOP and case time, in absentia
removal orders, rates of representation, case outcomes/grant rates, and efficiencies in the
immigration court and detention systems.
Case Time

A key concern for many LOP stakeholders (immigration courts, detention facility staff, ICE,
detained persons, LOP providers) is ensuring that cases are completed in as timely a manner as
possible. Timely case completions are important to different stakeholders for different reasons.
For example, while individual detained persons may seek to spend as few days in detention as
possible, ICE must ensure it has sufficient available beds to house persons subject to mandatory
detention under the law. For their part, the immigration courts seek to ensure that there are no
undue delays in scheduling hearings. Yet, in spite of these different motivations, there is near
universal consensus among stakeholders that a key indicator of success for the LOP is its ability
to ensure that people do not spend unnecessary time in detention. This is not to say that all
stakeholders support an accelerated court process. However, they do concur that those persons
seeking repatriation through the removal process should be able to access removal quickly, and
those pursuing relief from removal should not encounter undue delays in the immigration court
process.
Detained LOP participants have shorter average case times
As mentioned in the introduction to this report, previous analyses of the LOP have suggested that
its participants have shorter case times in the immigration courts. Our analysis similarly found
that immigration court cases for detained LOP participants were completed in fewer average
Vera Institute of Justice 47

days than national averages and comparison group cases.56 As Figure 8 shows, the combined
average time for LOP participants whose cases were completed while they were in detention was
27 days (column 3), versus 40 days for comparison groups (column 5), a difference of 13 fewer
days for LOP participants.57 Figure 8 also shows the average (mean) number of days for each
category of relief application. As columns 1, 3, and 5 indicate, lower average case times for LOP
participants were consistent across every type of relief application.
Figure 8: Mean Case Time by Relief Application Type for National Averages, LOP Participants,
and Comparison Cases, January 1 – August 31, 2006
Relief Application Type

All Cases
Total Number
Mean Days
of Cases

Mean
Days

LOP
Total Number
of Cases

Comparison
Mean
Total Number
Days
of Cases

No Application

22

29,465

15

5,184

24

19,978

9,591

18

1,563

19

6,904

Voluntary Departure

19

Other Application Combinations

163

754

135

137

170

525

I­589

163

2,650

151

318

165

2,215

212C Only

127

156

107

24

134

116

240A Only

120

1,272

92

265

131

879

240B Only

207

166

178

37

207

111

Total

35

44,054

27

7,528

40

30,728

Note: These overall means are not controlling for custody or representation status. See Figures 9 and 10.

Overall, cases that did not include applications for relief were completed in 15 days for LOP
participants, versus 24 days for comparison groups—a difference of nine days. Similarly, cases
that did include relief applications were also completed in fewer days for LOP participants. For
instance, for LOP participants pursuing 240A relief, case times were 39 days faster than for
comparison groups. Below, we discuss the relationship between the LOP and these case time
differences. We also describe potential cost savings to the federal government created by a
reduction in case processing times.
The LOP is associated with faster case time
Vera’s analysis of case time before and after the start of the LOP found that case processing
times (from initial Master Calendar Hearing to final case completion pre­appeal) have decreased
for cases concluded in detention at all hearing locations across the country. That analysis shows,
however, that case processing times have decreased more for cases that began at LOP sites. But

56

This held true even when we excluded stipulated removal cases and controlled for representation status.

57

We defined case processing time, case time, and case completion time as the time the case spent in the
immigration courts before any appeals. We defined the initial Master Calendar Hearing as the start of the case and
the date of the case decision/completion issued by the immigration judge, before appeal, as the end date. We do not
presume that appeal information is not relevant, but in our analysis we wanted to first look only at time spent before
immigration judges pre­appeal so as to avoid comparing cases on appeal with those that did not involve an appeal.
Vera Institute of Justice 48

is the LOP responsible for this difference, or are there other factors that have led LOP
participants’ cases to move more quickly through the courts? We now examine this issue.
Although LOP participants had shorter case processing times overall in 2006 (see Figure 8),
we cannot conclude that the overall differences shown are a result of the LOP. However, when
we analyzed case processing times in a different way, looking at LOP participation in
conjunction with type of relief application sought, representation status, and custody status at the
time of completion, we did find that LOP participation was associated with shorter case time.
Figure 9 below shows that when we isolated unrepresented cases concluded in detention, we
observed a similar trend of shorter case time for LOP participants, with the exception of cases
involving applications for voluntary departure.
Figure 9: Mean Case Time for Unrepresented Cases by Relief Application Type for Detained LOP
Participants and Comparison Cases, January 1 – August 31, 2006
LOP

Comparison

Relief Application Type

Mean Days

Total Number
of Cases

Mean
Days

Total Number
of Cases

Days fewer for LOP

No Application

5

4590

10

16082

5

Voluntary Departure

7

1326

5

5464

2 more

I­589

108

180

132

632

24

Other Application Combinations

87

278

99

410

12

We know that the LOP sites were selected because they shared common features, and unless
some features(s) other than these have led the six courts we studied to process cases faster than
all the other courts in the country—which we are unable to imagine—we can reasonably assume
that there is a possible relationship between these differences in case times and the LOP. As we
see in Figure 10 below, the pattern of shorter case times for detained LOP participants held true
for cases with no applications for relief and applications other than I­589 even when those cases
involved legal representation.
Figure 10: Mean Case Time for Represented Cases by Relief Application Type for Detained LOP
Participants and Comparison Cases, January 1 – August 31, 2006
Relief Application Type

LOP
Total Number
Mean Days
of Cases

Comparison
Mean
Total Number
Days
of Cases

Days fewer for LOP

No Application

32

309

46

1702

14

Voluntary Departure

30

160

30

793

0

I­589

128

73

133

966

5

Other Application Combinations

90

120

104

719

14

Figure 11 below shows the differences in case time for represented LOP participants and
comparison groups that are presented in Figures 9 and 10. The first columns for each category in
Figure 11 show the average case time by rates of relief application for LOP participants with
legal representation, while the third columns in each category present the average case time by

Vera Institute of Justice 49

rates of relief application for unrepresented LOP participants. Similarly, the second columns in
each category show average case times by rates of relief application for comparison groups with
representation, while the fourth (and last) columns in each category show average case times by
rates of relief application for unrepresented comparison groups.
Figure 11: Mean Case Times for Detained Cases by Representation Status and Relief Application
Type, January 1 – August 31, 2006
LOP w/ Representation

Comparison w Representation

LOP w/o Representation

Comparison w/o Representation
128133

132

108

104
90

99
87

46
32

30 30
5 10

No Application

7

5

Voluntary
Departure

I­589

Other
Application
Combinations

We conducted an additional analysis to see whether LOP participants’ cases simply move
through the system faster because perhaps they are in immigration courts with faster­moving
dockets. At the suggestion of EOIR staff, we examined the relationship between case processing
time and the time between the date a case was entered into the immigration court database and
the date of the initial Master Calendar Hearing. Our thought was that if the shorter case
processing times for LOP participants were simply a product of fast immigration court dockets,
there would be a connection between quick scheduling (time from case input to first Master
Calendar Hearing) and fast case processing times. We did not find such a pattern, leaving us
reasonably sure that the LOP—or at least some other unidentified variable unique to the several
thousand LOP cases we studied—had a shortening effect on case processing times.58
Implications of reductions in case time
As the use of detention—
and bed space in many of the facilities hosting the LOP—has expanded, the program has
continued to serve more and more people each year. However, the expansion of detention has
outpaced the expansion of funding for the Legal Orientation Program, so the number of people
REDUCTION IN IMMIGRATION COURT TIME COULD LEAD TO COST SAVINGS.

58

Because of data quality concerns and time constraints, we were unable to test if LOP participants also had fewer
adjournments (court appearances) than comparison groups, but in the recommendations at the end of this report, we
discuss the importance of this additional analysis to any argument that a reduction in case time leads to cost savings.
Vera Institute of Justice 50

receiving LOP services represents a shrinking percentage of the overall detained immigration
court population each year. Many court stakeholders we interviewed—and indeed many system
stakeholders more broadly—have expressed an interest in learning more about how reductions in
case time might translate into resource savings for the immigration court and detention
systems.59 While we have not yet developed the economic models necessary to attach a monetary
value to the time savings observed and have not yet studied the ways in which the immigration
courts might save costs by reducing case processing time, fewer days in immigration court would
logically and could actually result in savings of time and financial resources for both EOIR and,
presumably, ICE. The faster cases are completed, the sooner detainees can be released or
removed, reducing their time in administrative custody and creating additional detention space.
To illustrate this point we have conducted a few simple calculations using an average
detention bed cost of $97 a day.60 For the purposes of this analysis, we are showing the savings
to the detention system if court days and detention days were the same, which they almost
certainly are not (i.e., a savings of five days in court also leads to a savings of five days in
detention, even if the total number of detention days is always greater than court days). We are
simply illustrating a point about the possible cost savings if every day less in court equaled a day
less in detention. If ICE were to make available accurate data on the time that elapses between
immigration court decisions and removal, we could determine what the potential cost savings to
the detention system would have been in 2006. Even if a reduction in court days only led to a
partial reduction in detention days, there would still be cost savings, the savings would just be
less. If detention days were associated with court days so that five fewer court days meant five
fewer detention days, or 10 fewer court days meant 10 fewer detention days, in 2006 LOP
participants’ cases would have cost $3.2 million less in detention bed days. After the cost of $1
million invested in 2006 in the LOP sites included in this study, approximately $2.2 million
would have been saved in detention costs. Below we describe how we calculated this.
Using the numbers shown in Figures 9 and 10, we multiplied the number of LOP
participants’ cases by the average reduction in detention days for LOP participants’ cases. For
example, 4,590 unrepresented LOP participants did not file relief applications. Those 4,590 spent
an average of five fewer days in court than comparison group cases, or (4,590 cases) x (5 days) =
22,950 fewer total bed days. If each bed day cost the federal government $97, and five fewer
days in court meant five fewer days in detention, the savings caused by the reduction in court
days for LOP participants’ cases would be (22,950 bed days) x ($97 per bed day) = $2,226,150
less spent on bed days for unrepresented LOP participants without relief applications. Of course,
59

Meeting case completion goals and reducing case backlogs are a key management goal outlined in EOIR’s 2005­
2010 Strategic Plan. See U.S. Department of Justice, Executive Office for Immigration Review, Fiscal Years 2005­
2010 Strategic Plan, September 2004.
60

In recent testimony before Congress, ICE Director Julie L. Myers stated, as paraphrased in a news account, “the
cost of deporting the estimated 12 million unauthorized foreigners in the U.S. would be $94 billion, based on
holding each person for 32 days in jail at $97 a day, transporting them home at $1,000 each, and covering the cost of
ICE personnel. Myers acknowledged that the cost estimate was approximate since, for example, many foreigners
might decide to go home on their own to avoid a month in jail.” Migration News, October 2007.
Vera Institute of Justice 51

this is just a hypothetical argument. It might be that court time has no impact on detention bed
days at all, such that reducing court time does nothing to change the total time in detention. Or, it
might be that only some of the shorter court cases are associated with a corresponding reduction
in detention time. This might depend, in large part, on the amount of time ICE requires to
process travel documents necessary for repatriation to most countries. Using the same formula of
cases multiplied by average days difference, we calculated the following potential cost
differences:
Unrepresented LOP Participants
No Application
Voluntary Departure Only
I­589
Other Applications

N
Average Days Difference
4,590
5 fewer
1,326
2 more
180
24 fewer
278
12 fewer

Cost Difference
$2,226,150 less
$257,244 more
$419,040 less
$323,592 less

Represented LOP Participants
No Application
Voluntary Departure Only
I­589
Other Applications

N
309
160
73
120

Cost Difference
$419,622 less
no difference
$35,405 less
$162,960 less

Average Days Difference
14 fewer
0 difference
5 fewer
14 fewer

In total, actual differences in case time in 2006 may have saved $3,329,525 in detention costs
(the sum of the two “cost difference” columns above).61 Even after we subtract the $1,000,000
invested in the LOP in 2006, there is still a net savings of $2,329,525.
Could the LOP further reduce detention costs if it were expanded? Even if the LOP only
reduced bed days by five average days for 20,000 persons a year at a cost of $97 a bed day, the
cost reductions to the federal government would be almost $10 million a year (20,000 persons) x
(5 days savings) = (100,000 bed days), (100,000 bed days) x ($97 a day) = $9,700,000 total
potential savings. If bed days could be reduced by 10 days for the same 20,000 persons or by five
days for 40,000 persons, the cost reductions would rise to $19.4 million a year. Even if the cost
of the LOP needed to be doubled from $2 million to $4 million or tripled to $6 million to attain
these results, there could still be a substantial reduction. As we recommend in our conclusion,
this hypothetical analysis shows that if the federal government is interested in increasing cost
reductions as a result of legal access programs, one way to do so would be for ICE and EOIR to
invest as many resources as possible into working together to determine the best formula for
ensuring persons in detention spend the fewest number of days in custody as possible.
Because there is a long­term detention bed shortage—and an estimated more than 12 million
foreign nationals in the United States unlawfully according to ICE Director Julie L. Myers—the
reductions associated with the LOP might not mean actual monetary savings via cost reductions
61

In computing the sum of the two cost difference columns, the $257,244 greater cost of unrepresented applicants
for voluntary departure is subtracted rather than added.
Vera Institute of Justice 52

but, rather, cost diversion.62 That is to say, instead of investing several million dollars in new
detention beds above and beyond the current cost of detention, following the formula above,
there would now be an additional 100,000 bed days available in existing facilities. In fact, then,
the actual cost savings or diversion might be much greater than the several million dollars
calculated here because instead of building new facilities, ICE could populate vacant beds in
existing facilities. Or, the money saved could be devoted to the construction of new beds so that
a greater volume of persons could be detained each year. No matter how we calculate the costs,
if fewer days in court led to fewer days in detention for LOP participants, the federal government
would see reduced costs.
ICE could experience even greater possible cost reductions if it could find a way to ensure
that released LOP participants would return to immigration court. While the government does
not expend money for the living costs of released persons in removal proceedings, we know that
released persons do not always continue to appear for their immigration court proceedings and
may, in fact, end up producing greater costs, as ICE is then tasked with locating and
apprehending absconders. However, as we discuss below, detained persons with access to legal
information have lower in absentia removal rates.63 This means that ICE may be able to further
reduce costs and utilize existing detention bed space by increasing the numbers of persons
released from detention during their removal proceedings. Before discussing that point, we
address below a few additional concerns about faster case processing time for LOP participants.
While reductions in
case time have the potential positive effect of reducing detention costs substantially, there may
be downsides to shortening the number of days that an immigration court case lasts. In our
interviews with legal representatives working with detained persons, we noted that attorneys are
concerned that when cases move too quickly through the immigration courts, detainees may have
fewer opportunities to pursue relief. Immigration attorneys we interviewed were also worried
that fast cases might be associated with lower grant rates.
Overall, we did not see many differences in grant rates between LOP and comparison groups
when we looked at case processing times for different application types, representation, and
custody status. However, we did find a weak relationship between case processing time and case
outcome for cases with I­589 applications, suggesting that cases with I­589 applications that took
longer to complete received slightly more grants.64 We are carrying out more analyses to see if
we can better gauge what amount of case time might be related to positive grant rates. For
example, this preliminary finding might simply suggest that cases that take more than a certain
REDUCTIONS IN CASE TIME MAY REDUCE POSSIBILITIES FOR GRANTS OF RELIEF.

62

Migration News, October 2007.

63

Vera Institute of Justice, Testing Community Supervision for the INS: An Evaluation of the Appearance Assistance
Program. Final Report to the Immigration and Naturalization Service (New York: Vera Institute of Justice, 2000).
64

We did see much higher rates of removal for cases that had only one hearing, which is likely a product of the fact
that those respondents did not pursue representation or relief. Indeed, many LOP providers encourage those
detainees who want to expedite their removal to request removal at the earliest possible Master Calendar Hearing.
Vera Institute of Justice 53

number of days stand a better chance of being granted, but after that threshold, there may be no
difference. For example, a case may need more than 50 days to have the best chances of success,
but there may be no difference between 50 days and longer case times.
We also noticed that, contrary to the patterns we observed for all other types of relief, cases
in which only voluntary departure was sought took longer for unrepresented detained LOP
participants than for comparison groups. This difference may be a result of the fact that our
analysis showed there to be a slightly greater chance of receiving a grant of voluntary departure
when a case (1) takes slightly longer, and (2) is decided at the Merits Hearing level. In this
instance, if the LOP is lengthening cases with voluntary departure as the only application, the
LOP is, in fact, helping persons achieve the outcomes they seek—grants of voluntary departure.
The added cost(s) then may be “worth” it, using a non­monetary standard to gauge the level of
benefit(s).
Even though there may be relationships between case processing time and grant rates for
certain relief applications such as I­589 and voluntary departure, we are not concluding that
grants are solely the result of longer case times. More likely there are many other factors at play,
such as the merits of the case, legal representation, and whether the case was determined at a
Master Calendar or Merits Hearing.
Released LOPparticipants have longer average case times
Although detained LOP participants’ cases take fewer days on average, their cases follow the
opposite pattern when they are released, at least in the few cases of released LOP participants. In
fact, with only one exception, the few released LOP participants’ cases we analyzed took more
time than comparison group cases following release.65 Figures 12 and 13 below illustrate that
just as there may be a relationship between LOP and shorter case time for detained cases, so,
too, may there be a relationship between LOP participation and longer case processing times for
cases of persons released from detention, at least for those cases involving no relief applications
and cases with I­589 applications. The same patterns of longer case processing time for released
LOP participants’ cases holds true whether released cases are represented or unrepresented.
Interestingly, case time is one of the few outcomes we looked at for which the overall patterns
for LOP participants did not change substantially with representation.

65

One possible explanation suggested by some LOP providers and stakeholders is that LOP cases take longer upon
release because LOP participants have been informed of the value of obtaining representation for complicated cases
and that participants are requesting more time to find representation upon release. Our findings suggest this
explanation may be particularly salient for I­589 cases, for which we see dramatic differences in representation rates
between detained and released LOP participants.
Vera Institute of Justice 54

Figure 12: Mean Case Time for Unrepresented Cases by Relief Application Type for Released

LOP Participants and Comparison Cases, January 1 – August 31, 2006

Relief Application Type

LOP
Total Number
Mean Days
of Cases

Comparison
Mean
Total Number
Days
of Cases

Days more for LOP

No Application

146

192

101

1370

45

Voluntary Departure

164

36

130

155

34

I­589

292

19

207

133

85

Other Application Combinations

274

27

249

100

25

Figure 13: Mean Case Time for Represented Cases by Relief Application Type for Released LOP

Participants and Comparison Cases, January 1 – August 31, 2006

Relief Application Type

LOP
Total Number
Mean Days
of Cases

Comparison
Mean
Total Number
Days
of Cases

Days more for LOP

No Application

146

93

141

824

5

Voluntary Departure

165

41

121

492

44

I­589

297

46

263

484

34

Other Application Combinations

254

38

256

402

2 fewer

More research is needed to understand what may be causing these small numbers of released
cases to take longer for LOP participants than their counterparts in comparison groups and to see
if there are any negative implications of longer case times for released persons. We did find that
rates of in absentia removal orders are not influenced by case time (i.e., so long as a released
person appears for the first hearing after release, long case times do not seem to yield higher
rates of in absentia removals). If a lengthy case for a released person is no more likely to lead to
an in absentia order of removal (or a failure to appear) than a speedy case, there may not be any
additional resource burden caused by the fact that these cases take slightly longer. These cases
would need to be studied to understand what factors might lead them to take longer. If it is
because now­released persons who have attended the LOP are motivated to seek representation
and therefore request additional adjournments following release, these longer case processing
times may in fact be a positive finding.
In Absentia Removal Orders

Because of the small numbers of LOP participants released from detention, our study has not
included a comprehensive analysis of the factors contributing to in absentia removal orders.
However, reducing in absentia rates remains a main area of concern for both EOIR, which
identifies this as a goal in its 2005­2010 Strategic Plan, as well as ICE, which is charged with the
resource­intensive task of apprehending persons who fail to appear for their hearings.66 Ensuring
that released persons appear for court is also a broader public policy and public safety concern
that is frequently mentioned in studies and reports about the detention and removal processes.
66

U.S. Department of Justice, Executive Office for Immigration Review, Fiscal Years 2005­2010 Strategic Plan,
September 2004.
Vera Institute of Justice 55

Moreover, with better methods for ensuring that released persons continue to appear in
immigration court, more persons could be released.67 Releasing more persons in removal
proceedings from detention could potentially save the detention system substantial amounts of
money and result in better use of existing bed space. This could additionally save the government
from needing to build more detention capacity, reducing costs further.
Our data and findings from other studies suggest that representation and access to legal
information through programs like the LOP helps reduce in absentia removal orders.68 Of course,
the overwhelming majority of all persons whose immigration court cases begin in detention have
their cases decided in detention and at the same hearing location at which they began.69 Very few
persons were released from detention in 2006. After we excluded stipulated removal cases, only
11 percent of all detained persons were eventually released, almost all on bond.70 Therefore,
when discussing in absentia removal orders, we are talking about only a fraction of the 11
percent of cases that are concluded upon a respondent’s release from custody, a very small
number.71
Unrepresented LOP participants receive fewer in absentia removal orders
Rates of removal in absentia vary greatly when we examine subgroups by representation status,
charges, and type of relief from removal being sought. However, there is an overall pattern of
unrepresented, released LOP participants having an average of 7 percent fewer in absentia

67

In absentia rates vary across immigration courts even when the volume of cases for released persons is the same.
If we hold constant other variables while taking into consideration the released hearing location, further tests may
show that the location to which the case is transferred following release has more of an effect on in absentia removal
rates than other variables, or at least works with them to increase the chances of an order of removal in absentia.
This could lead us to predict even greater chances of success for LOP participants, who currently have their released
hearings at some of the immigration courts with the highest overall in absentia removal rates nationwide. More
analysis of change of venue patterns and access to legal information may help explain these differences further as
we continue to analyze these patterns in subsequent phases of research.
68

Nonetheless, legal representation is still more strongly associated with reduced in absentia orders than are LOP
services. In other words, participating in the LOP, while important, may have less influence over improved
appearance rates than does having legal representation.
69

In fact, we found that 71 percent of all completed cases received a decision from an immigration judge at the
initial Master Calendar Hearing, and 94 percent of those cases resulted in orders of removal. In other words, the
overwhelming majority of all persons in detention do not pursue relief and are ordered removed quickly, often at the
initial Master Calendar Hearing.
70

One hundred seventy five people in the comparison group were coded as having been released on electronic
monitoring/ISAP. Only 130 people (including LOP, comparison cases, and some in the excluded groups) were
released on their own recognizance. Notably, those cases coded as electronic monitoring were ordered removed in
absentia 7 percent of the time. They were left in the comparison group since omitting them had little effect on the
overall outcomes.
71

Of the 44,054 cases we studied over an eight­month period, 4,834 involved persons released from detention, and
of the released persons 37 percent or 1,786 received in absentia removal orders in 2006. This represents 4 percent of
the total group we studied.

Vera Institute of Justice 56

removal orders.72 As Figure 14 below shows, for unrepresented I­589 applicants, differences
were the most dramatic—13 percent fewer in absentia orders for LOP participants pursuing I­
589 relief (21 percent for LOP versus 34 percent for comparison groups). On the other hand,
there was no real difference in rates of in absentia removal orders for those persons who did not
pursue relief from removal. Not surprisingly, as Figure 14 illustrates, those persons who did not
identify a form of relief to pursue or did not otherwise seek relief were the least likely to
continue to appear in court and the most likely to be ordered removed in absentia. These
individuals accounted for three quarters (75 and 76 percent) of all in absentia orders.
Figure 14: Rates of In Absentia Removal Orders for All Unrepresented Released Cases
LOP

Relief Application Type

Comparison

Difference

N

Percent

N

Percent

No Application

192

75%

1485

76%

no difference when no
application

All Applications (including
VD)

81

12%

423

17%

5% fewer absentia orders for
LOP cases with applications

I­589

19

21%

136

34%

13% fewer absentia orders for
LOP cases with applications

Total (overall absentia)

273

56%

1908

63%

7% fewer absentia orders for
LOP cases

(N= 4,834)

Fewer in absentia orders with more legal information
PARTICIPANTS IN INTENSIVE LOP SERVICES RECEIVED IN ABSENTIA ORDERS AT LOWER RATES.

Figure 14 shows that the overall rate of orders in absentia for unrepresented released persons
who did not pursue relief was virtually identical for LOP participants (75 percent) and
comparison groups (76 percent). Yet, when we isolate LOP participants who attended more than
group orientations, the rate of orders in absentia for unrepresented persons not pursuing relief
drops from 75 percent to 67 percent (not shown here).
While we cannot know for sure if the LOP is responsible for lower rates of orders of removal
in absentia, Vera’s evaluation of the Appearance Assistance Project (AAP) found that
participation in a supervised release program that facilitated access to legal and social services
did help ensure that released persons continued to appear in immigration court.73 In that study,
groups categorized as “undocumented workers” apprehended in worksite enforcement actions in
the New York City metropolitan region received many fewer in absentia orders when they
participated in the AAP. While comparison group (non­AAP) “undocumented workers” were
72

We are not showing in absentia rates by representation because it is impossible to identify which cases had
representation for bond only and which had representation at the point of release. However, cases that had
representation at any point had average rates of removal in absentia below 25 percent as opposed to 56 percent
overall for LOP participants and 63 percent for comparison groups.
73

Vera Institute of Justice, Testing Community Supervision for the INS: An Evaluation of the Appearance Assistance
Program. Final Report to the Immigration and Naturalization Service (New York: Vera Institute of Justice, 2000).
Vera Institute of Justice 57

ordered removed in absentia 41 percent of the time, undocumented workers participating in AAP
“intensive supervision” received in absentia orders only 12 percent of the time.74 People in both
groups were enrolled in the study without regard for the merits of their immigration court cases;
instead, Vera staff considered factors such as the strength of their ties to the community and the
immigration offenses with which they were charged. Though the AAP did not factor relief
possibilities or merit into the screening tool used to enroll participants in the program,
presumably the undocumented workers had few possibilities for relief aside from cancellation of
removal.
The AAP study concluded that lower rates of orders in absentia for undocumented workers in
the study was a product of the supervised release they received. In fact, the AAP’s one­on­one
supervision sessions offered legal orientations based on the Florence Project’s rights presentation
model as well as orientations to social services, such as mental health counseling.75 We found
similar results in our analysis of the LOP. LOP participants who received more intensive
services, such as individual orientations or self­help workshops, were ordered removed in
absentia at lower rates, even when they did not pursue relief from removal.
LOP PARTICIPANTS WHO RECEIVE INTENSIVE SERVICES AND PURSUE RELIEF HAVE LOW RATES OF IN

When we controlled for
both level of LOP service and application type, we found even more dramatic differences in rates
of orders in absentia. When we looked only at LOP participants’ cases, we observed that rates of
in absentia orders for LOP participants who only attended group orientations and did not pursue
relief from removal were 74 percent. On the other hand, rates of orders in absentia for all other
LOP participants (those who attended more than group sessions and/or those who attended only a
group session but pursued relief from removal) were only 24 percent. This leads us to conclude
that outcomes of LOP participants can approximate the low in absentia removal rates for cases
ABSENTIA ORDERS, APPROXIMATING THE RATES OF REPRESENTED CASES.

74

The AAP provided supervised release from detention at two levels, intensive and regular. Intensive participants
were persons initially detained by the INS and then released to the AAP; they had to report regularly to AAP
supervision officers in person and by phone. Program staff monitored each participant and re­evaluated the risk of
non­compliance or flight. Regular participants were noncitizens apprehended by the INS and then released on
recognizance; they entered the program voluntarily. People in both types of supervision received information about
immigration proceedings and the consequences of noncompliance, reminders of court hearings, and referrals to legal
representatives and other services.
75

Vera’s AAP intensive supervision program should not be confused with ICE’s current intensive supervision
program, or ISAP. The AAP intensive supervision relied on graduated sanctions without electronic monitoring. The
current ISAP program, on the other hand, relies heavily on electronic monitoring. In addition, the AAP included a
strong emphasis on legal orientations and social services, which are not a central component of ISAP. The AAP
relied on theories of compliance which posit that people will choose to “do the right thing” if presented with
accurate information about processes that are fair and transparent—in contrast with electronic monitoring, which
induces compliance solely through control. While compliance through control may be more effective in the short­
run, theorists who write about procedural justice have noted that such methods may have the opposite of the
intended effect in the long run. In the immigration context, electronic monitoring may lead people who would
otherwise comply with the removal process to develop antagonistic attitudes. See Tom R. Tyler, Readings in
Procedural Justice (Burlington, VT: Ahsgate, 2005).
Vera Institute of Justice 58

with representation when respondents receive more intensive LOP services, just as we found
with the 2000 AAP study.
While the
combined total in absentia rate for cases we studied was 62 percent for unrepresented persons
(56 percent for LOP and 63 percent for comparison groups, as indicated in Figure 14),
represented persons were ordered removed in absentia only 17 percent of the time—more than
three times less than represented persons (not shown here). Rates of removal in absentia were
even lower for persons pursuing certain types of relief from removal. Figure 14 shows that
unrepresented I­589 applicants who participated in the LOP were ordered removed in absentia 21
percent of the time. Our analysis also revealed (not shown here) that rates of orders of removal in
absentia for represented I­589 applicants were only one third that figure, or only 7 percent.
Unrepresented I­589 applicants in the comparison group received three times as many in absentia
removal orders as the represented group (34 percent versus 10 percent). While representation
appears to be more successful than the LOP in reducing in absentia removal orders for released
individuals, in the absence of counsel for all detainees, LOP may be the next best method—and a
more cost­effective approach—for reducing the rates of orders of removal in absentia.
CASES WITH REPRESENTATION RECEIVED IN ABSENTIA ORDERS AT MUCH LOWER RATES.

Representation

In our analysis of completed cases that began in detention in 2006, the nationwide representation
rate was 14 percent; the rate was even lower for cases that began and ended in detention.
Historically, representation rates at many LOP sites have been lower than the national average.
As noted above, one criterion for selection of the original LOP sites was that they had low rates
of representation, thereby ensuring that the program would provide the assistance to those most
in need.76 Indeed, at the detention centers hosting LOPs, the LOP generally represents detainees’
only opportunity to talk to someone well­versed in the laws that govern immigration court and
the removal process.
We know from our research and other studies that cases with representation have higher rates
of relief applications filed, fewer orders in absentia, and higher rates of grants of relief.77
Although LOP providers encourage individuals who are not able to obtain representation to
develop the skills necessary to appear pro se, they also make pro bono referrals when possible
and encourage detained LOP participants to pursue representation upon release. While
76

In our analysis, “representation” signifies that there was a name (of either an attorney or an accredited
representative) in the “attorney” field of the court database. The “attorney” field is filled in whenever an E[EOIR]­
28 form is filed with the court. In our analysis, representation means there was an E­28 filed at some point in the
case, before any appeals to the Board of Immigration Appeals.
77

Vera Institute of Justice, Testing Community Supervision for the INS: An Evaluation of the Appearance Assistance
Program. Final Report to the Immigration and Naturalization Service (New York: Vera Institute of Justice, 2000);
Ramj­Nogales, Jaya, Andrew Schoenholtz, and Philip G Schrag, “Refugee Roulette: Disparities in Asylum
Adjudication,” Stanford Law Review 60 (2008). TRAC. Immigration Judges.
http://trac.syr.edu/immigration/reports/160/
Vera Institute of Justice 59

representation rates remain comparatively low at LOP sites, it appears that the LOP is increasing
representation rates for individuals with relief possibilities and decreasing representation rates
for individuals with no possibilities for relief.
Since the LOP began, there has been a steep decline in representation rates for persons with
initial Master Calendar Hearings at LOP sites, particularly for those persons who do not pursue
relief applications. At the same time, cases that begin at comparison sites have not experienced a
significant change in representation rates. As Figure 15 shows, before the LOP began, detainees
not pursuing relief at LOP sites (indicated by the solid line) had higher rates of representation
than detainees not pursuing relief at non­LOP sites (indicated by the broken line). After the
introduction of the LOP, this trend reversed.
Figure 15: Representation Rates for Detained Cases without Relief Applications, 2000 ­ 2005

As Figure 15 shows, rates of representation for detainees who do not pursue relief are greater
at non­LOP sites than at LOP sites. We believe this could be a positive trend. When attorneys are
not available, LOP providers aim to help prepare detainees who are not pursuing relief to appear
in immigration court pro se. In addition, the LOP group orientation briefly explains what an
individual should expect from a legal representative. Some LOP providers—particularly
ProBAR, in Port Isabel, Texas—believe that fewer program participants are hiring unqualified
representatives, or “notarios,” than in the past. LOP providers believe that diverting persons
away from unqualified attorneys or individuals who do not act in the client’s best interest is a
positive outcome of the LOP.
The decreasing rates of representation for those who do not pursue relief at LOP sites may
also reflect a decrease in the use of representation solely for bond hearings. ProBAR and other
LOP providers told us that the LOP may be helping people make more informed determinations
about bond—in particular by choosing not to request bond hearings when they are statutorily
ineligible. Of course, it is also possible that there are simply fewer legal representatives available
than in the past at LOP sites—and that without legal representation, many potential relief
Vera Institute of Justice 60

applications go undetected. Nonetheless, most LOP providers reported observing fewer detainees
seeking representation in voluntary departure or removal cases.
When we isolate representation rates for those who pursue forms of relief other than
voluntary departure, we see a trend in the opposite direction.78 Just as representation rates have
decreased for cases without relief applications at LOP sites, there have been large increases in
the numbers of represented cases with applications for forms of relief other than voluntary
departure at LOP sites. When we look only at detained persons applying for forms of relief other
than voluntary departure, there is an increase in representation rates at LOP sites since 2002—or
about six months before the program began (Figure 16). This trend line continued to move up
after the start of the LOP—though more gradually—while representation for cases with relief
applications at comparison sites held steady. This finding points to the fact that the LOP may be
contributing to the modest but sustained increase in representation for individuals identified by
the LOP as potentially eligible for relief or those individuals who decide to pursue relief
regardless of their eligibility, bringing their rates of representation closer to national averages.
Thus, the LOP may be providing detainees with the information to make better decisions about
when to pursue paid legal representation in removal proceedings.
Figure 16: Representation Rates for Detained Cases with Relief Applications Other than
Voluntary Departure at LOP and Comparison Hearing Locations from 2000 to 2005

Although representation rates have increased in cases that involve relief applications at LOP
sites, detained individuals continue to be represented at very low rates overall: in our sample, 87
percent of cases that started and ended in detention in 2006 had no representation at the time of
the final decision by the immigration judge. As Figure 17 illustrates, representation rates also
continue to vary by type of relief application and access to LOP services. While detained I­589
applicants who received LOP services had much higher rates of representation (60 percent – 29
percent = 31 percent) than some LOP participants pursuing other forms of relief, detained I­589
78

Figure 15 and other figures from our historical analysis represent outcomes for cases that began and ended in
detention.
Vera Institute of Justice 61

applicants in the comparison groups had twice the representation rates (60 percent). Since LOP
sites were selected in part because of their low rates of representation, it is not entirely surprising
to find higher rates in the comparison groups.
Figure 17: Rates of Representation by Application and Custody Status
Detained
Relief Application Type

Released

LOP

Comparison

LOP

Comparison

No Application

6%

10%

33%

38%

Voluntary Departure

11%

13%

53%

76%

I­589

29%

60%

71%

78%

Other Application Combinations

30%

64%

59%

80%

13%

Total

55%

If it were not for the historical data showing that rates of representation have slowly
increased for cases with relief applications at LOP sites, we might conclude from Figure 17 that
LOP reduces rates of representation. But as Figure 15 indicates, representation rates at LOP sites
have declined only for cases with no relief applications, while representation rates for all other
applications have increased. In addition, as Figure 20 illustrates, the LOP may be helping I­589
applicants realize the importance of obtaining representation following release from detention.
Figure 18: Percent Represented for I­589 Cases (Detained and Released)

LOP

Comparison

78%

71%
60%

29%

Detained

Released

Vera Institute of Justice 62

As Figures 17 and 18 both indicate, LOP participants pursuing I­589 applications have dramatic
increases in representation—from 29 percent to 71 percent—after release from detention.
Comparison cases also show an increase in representation—albeit less dramatic—when released
from detention. We see a similar doubling of representation rates for released LOP participants
pursuing applications for types of relief other than voluntary departure. More research will be
needed to determine what is responsible for this increase in representation rates: one explanation
is that people are moving to large cities with abundant legal resources after being released from
detention; another is that LOP participants are working harder to secure representation after they
are released.
Case outcomes
Given the low overall rate of granting relief for people whose immigration court cases begin
while they are detained (3 percent overall), it is not surprising that we observed few differences
in case outcomes between LOP participants and comparison groups. However, when we looked
at case outcomes by the level of LOP service, we found evidence suggesting that when
unrepresented LOP participants received “intensive” LOP services (defined as any services
beyond the group orientation), they had case outcomes that moved closer to those in cases with
representation, as described below.79
Intensive LOP participants received grants of relief at higher rates
UNREPRESENTED DETAINED LOP PARTICIPANTS HAD HIGHER OVERALL GRANT RATES WHEN THEY

Grant rates for unrepresented LOP participants who
received more intensive levels of service were five times those of participants who received
group orientations alone (4.1 percent versus 0.77 percent) and three times the grant rate of
comparison groups (4.1 percent versus 1.4 percent). However, it is difficult to draw conclusions
about the significance of these figures, as case decisions by immigration judges almost always
involve judicial discretion.
RECEIVED MORE INTENSIVE SERVICES.

UNREPRESENTED I­589 APPLICANTS HAD HIGHER GRANT RATES WHEN THEY PARTICIPATED IN

LOP participants who received more intensive services had I­589 grant
rates of 9.4 percent versus 2.4 percent for those LOP participants who attended group
orientations alone. This statistic may illustrate the impact of the LOP better than overall grant
rates: presumably, most unrepresented detainees who pursue I­589 applications—regardless of
INTENSIVE LOP SERVICES.

79

While almost all LOP participants receive group orientations, only 38 percent of the group orientation attendees in
the group we studied received more intensive levels of service, in combinations of group orientations, individual
orientations, and self­help workshops. To test the theory that more intensive levels of individualized service might
be associated with different case outcomes, we created two categories of LOP participants: (1) group orientation
only; and (2) more than group orientation.

Vera Institute of Justice 63

whether they are LOP participants—need specialized information on how to prepare for a merits
hearing.80
UNREPRESENTED DETAINED LOP PARTICIPANTS WERE MORE LIKELY TO HAVE THEIR CASES

LOP
participants who did not file applications for relief were three times more likely to have their
cases terminated when they received intensive LOP services (3.1 percent versus 0.9 percent).
Their termination rates were also higher than those of unrepresented comparison groups (3.1
percent versus 2.2 percent). While there are a variety of factors that can influence the decision to
terminate proceedings, people who lack familiarity with immigration court processes may not
understand how or when to contest information presented by ICE during the proceeding. They
may not even know that it is possible to ask an immigration judge to terminate a case. The LOP
providers and immigration judges we interviewed reported that decisions to terminate are often
issued by immigration judges when a respondent successfully contests the removability charges
asserted by ICE or when an application for citizenship or legal resident status is approved by
USCIS. While not conclusive, the data point to the possibility that intensive LOP services are
helping detained people learn the skills they need to advocate for termination of their cases.
Readers should be cautious about drawing conclusions about the causality of these
observations because LOP participants are not required to participate in individual orientations;
rather, they self­select for these services. Also, in some high­volume LOP sites, participants who
have determined they have no relief or benefits to pursue are not prioritized for individual
orientations. As a result, we might expect that detainees who have self­selected for help have the
potential to learn the skills necessary for successful pro se representation. However, as is the case
with legal representation, we cannot tell if better outcomes are the result of people with relief
possibilities self­selecting for individual orientations, or if individual orientations might be
responsible for helping people effectively argue for termination of their cases. It is most likely a
combination of both.
TERMINATED BY IMMIGRATION JUDGES IF THEY PARTICIPATED IN INTENSIVE LOP SERVICES.

Unrepresented LOP participants have higher voluntary departure grant rates
For most cases, we were not able to identify a relationship between participation in the LOP and
case outcomes—perhaps because so few detained persons pursue relief and/or are granted relief
nationwide. However, our analysis shows that persons who apply for voluntary departure after
participating in the LOP are more likely to be granted voluntary departure than comparison
groups (44 percent for LOP participants versus 27 percent of those in comparison groups).
This contrast in success rates for applications of voluntary departure is even greater when we
compare subgroups. For unrepresented persons whose cases were concluded while detained,
immigration judges granted voluntary departure applications 37 percent of the time to LOP
participants versus 13 percent of the time for unrepresented comparison groups. Additionally, 41
80

We have not yet performed this particular analysis controlling for notice to appear charges and nationality, which
may change the distributions we are reporting here.
Vera Institute of Justice 64

percent of Mexicans who participated in LOP services were granted voluntary departure versus
only 14 percent of Mexicans in comparison groups. LOP participants generally were more than
twice as likely to be granted the voluntary departure they sought, while Mexican LOP
participants were three times as likely to be granted voluntary departure.
There appears to be an unusually strong connection between seeking and being granted
voluntary departure for persons whose initial Master Calendar Hearings occurred while detained
in Port Isabel, El Paso, or Seattle. In fact, our analysis shows that over the past few years,
detained persons in Port Isabel have had about a 90 percent chance of being granted voluntary
departure when they sought this form of relief. This relationship suggests that LOP providers at
these sites may be particularly effective in helping detainees to determine whether they should
pursue this form of relief.
However, it is also possible that local procedures may be contributing to high voluntary
departure grant rates. Under voluntary departure, persons must demonstrate their ability to pay
for transportation back to their countries of origin. At Port Isabel, El Paso, and Seattle, these
transportation costs were negligible for Mexican nationals. Since Port Isabel and El Paso are not
far from the Mexican border, transportation costs are negligible. In Seattle, local operating
procedures have enabled similar low­cost transport arrangements.
Case outcomes discussion
THE LOP IS NOT A SUBSTITUTE FOR FULL REPRESENTATION.

Although participation in the LOP
and representation both lead to higher relief application rates, higher grant rates, and fewer
orders in absentia, the latter is more effective at producing these results. This is especially true
for noncitizens who are released from detention. Even for detained persons with no applications
for relief from removal, representation appears to help reduce the chances of removal by working
to divert some cases otherwise likely to result in removal towards case terminations by
immigration judges. While some of the stakeholders we interviewed remarked that these
difference in outcomes between represented and unrepresented individuals may simply be a
function of strong cases and good representatives finding each other, our data show that LOP
participants who received intensive services but were unrepresented—including a number of
participants with only voluntary departure applications—have higher grant rates and result in
fewer orders of removal and fewer in absentia orders than unrepresented individuals who did not
participate in the non­LOP participants.
It follows that the LOP is no substitute for representation—even for people who are not
pursuing relief applications. Still, in some instances, access to intensive LOP services and pro se
assistance can approximate the outcomes associated with legal representation. In others,
obtaining access to legal information that can help a respondent appear pro se may be more
efficient than obtaining full representation, particularly when the respondent simply wants to
return to his or her nation of citizenship as quickly as possible. Indeed, it is in cases where there
is no application for relief and cases that involve only voluntary departure that we see some of
the clearest indications of the LOP’s impact. The LOP providers we interviewed almost

Vera Institute of Justice 65

unanimously agreed that providing detained persons with the information they needed to proceed
with their immigration court cases was a principal part of their job, and that this had the
collateral benefit of helping to make the immigration courts more efficient.
The immigration judges we spoke with believe strongly in due process, and want people who
are statutorily eligible for relief to have access to resources like the LOP. They also want more
people to appear in immigration court with well­prepared applications and documents—and,
when necessary, representation. Many expressed frustration that the pro se model was not
meeting the needs of detainees as well as representation would, although they also recognized
that the LOP helps make the immigration detention and removal process more efficient.
More specifically, many judges said that detainees need additional assistance in filling out
applications, especially lengthy and detailed I­589 applications. They also observed that
detainees need help in writing suitable and effective narratives about their cases and in
explaining any discrepancies between oral testimony and written documents.81
Most of the judges and LOP providers we interviewed believe that there is a limit to what the
LOP can offer in pro se cases that involve legal research or complicated legal arguments. In our
view, more research is needed to determine whether this belief is accurate—and if so, whether
there are other legal service models that might provide assistance in such cases.
Efficiencies in the Immigration Court and Detention Systems

The Legal Orientation Program depends on the cooperation of local ICE and detention facility
personnel who volunteer to host the program in their facilities. In our qualitative interviews,
detention facility staff repeatedly commented on the value of the program.
Detention facility staff believe the LOP reduces detainee anxiety and behavior problems and
makes detention “safer and more humane”
Several detention facility administrators observed that by providing detainees with accurate
information about immigration court processes, the LOP has reduced anxiety levels among
detainees, with the result that there are fewer behavior problems. In more than one location,
detention facility staff said that the LOP had done so much to improve the quality of life for
detainees that if it didn’t exist, they would find a way to replicate it. “Why would I object to the
program?” one officer said. “It just tells them about their rights—it doesn’t give them any more
rights.”
Others echoed the sentiment that detainees should be provided with information about their
rights. One warden told us that, in his view, the LOP helped make detention “safer and more
humane.” Other detention facility staff said that by providing detainees with legal information
and a sense that their rights had not been violated, the LOP has led to reductions in the number of
81

While LOP providers may be able to do more to help applicants prepare, limitations placed by EOIR on program
activities prevent LOP providers from filling out application forms for participants. We found that many
immigration judges were not aware of the fact that their agency and not the individual LOP providers had made this
decision.
Vera Institute of Justice 66

violent disturbances and even the use of segregation. Of course, these are subjective views
expressed by a small number of people; it is difficult to confirm such impressions without access
to detention center data on the use of segregation. Still, the fact that detention facility staff
perceive safety benefits as a result of the LOP is itself an important finding.
In fact, even when we asked stakeholders, in confidential interviews, to be frank about any
complaints they might have about the LOP, we heard relatively little. Some remarked that the
program occasionally created confusion for new officers and that, because it required the
assignment of additional personnel, it could be a burden. Others questioned whether the LOP did
much more than “make detainees feel good.” On the other hand, some of the stakeholders we
contacted declined to be interviewed—perhaps because they have negative opinions about the
program.
Immigration Judges at LOP sites believe the LOP helps the immigration courts run smoothly
While immigration judges often told us they wished the LOP could do more for detained people,
many reported that LOP participants are better prepared to answer routine questions, know to
verbalize responses instead of nodding their heads, ask fewer questions about court processes,
and are more likely to pursue relief only when they are statutorily eligible, thus helping reduce
the caseload of the immigration courts. In fact, many immigration judges told us their greatest
critique of the program is that it does not offer more assistance with application preparation and
direct representation.82
Some immigration judges did note that many detained people who have taken part in the
LOP continue to appear fearful and confused in court. According to many judges, detainees were
particularly confused when it came to answering questions about reserving the right to appeal. In
one instance, a judge expressed frustration with a detainee’s confusion after learning that the
detainee had attended the LOP, but this appeared to be an anomaly. Almost all the judges we
spoke to told us that the LOP reduced confusion but did not eliminate it entirely. Court
proceedings generally occur extremely quickly, and for many people, immigration court is their
first experience with a court in the United States. Most LOP providers reported that detainees
often want more time to make decisions and process the information they learn in the LOP
orientation; many also said it was unclear to them whether a one­time intervention is really
enough to help detainees understand more than “bits and pieces” of the immigration court and
removal process.
Many immigration judges and ICE or detention personnel have told us (informally as well as
in formal interviews) that legal orientations need to be offered by independent, non­
governmental third parties if detainees are to trust them. These stakeholders noted that detainees
are more willing to accept unpleasant information from independent, nonprofit attorneys than
82

The function of the immigration courts is not to make immigration policy but to interpret administrative and case
law to adjudicate cases initiated by the Department of Homeland Security against noncitizens charged with
removability from the United States. If a respondent is eligible for relief from removal under the law, it is the
immigration judge’s duty to inform the respondent of this right and evaluate the merits of the case without bias.
When a case is prepared well and both sides are represented by counsel, a judge is better able to evaluate the merits.
Vera Institute of Justice 67

from detention facility staff. One ICE employee told us that, while he believes he provides
detainees with the same information as LOP attorneys, he is not perceived as a credible source.
Echoing this sentiment, several judges said that detainees are more willing to accept
disappointing information (with the result that they are less likely to seek adjournments) from
independent nonprofit staff than from immigration judges. Detainees made similar observations,
pointing out that while they were hardly eager to hear bad news about their chances of staying in
the United States, they appreciated being told the truth and believed that LOP providers were
giving them accurate and reliable information. In a few sites with very low overall representation
rates—or in one case, when there was no telephone access to the outside world for several
weeks—detainees who took part in the LOP reported that the program helped them make sense
of the conflicting information they heard from their fellow detainees. Some were quite
exasperated by the time they arrived at the LOP presentation, having been told that nobody
would help them. LOP providers struggle against this perception, as they are often able to do
little more than provide group and individual orientations and must explain to detainees that they
do not have the resources to provide representation.
Finally, while immigration judges were, on the whole, extremely supportive of the program,
they were concerned with what they viewed as a scarcity of immigration court resources—for
expanded LOP services as well as for their own work. Several immigration judges said there was
a need for more judicial appointments to help reduce individual caseloads and decrease
adjournment time. Judges also commented on the need for additional immigration court clerks to
aid in reviewing case law or drafting decisions. These comments suggest that some immigration
judges believe that EOIR is currently underfunded, with the result that they may be reluctant to
support the expansion of the LOP if they believe it will be at EOIR’s expense.

Vera Institute of Justice 68

IV: Recommendations and Next Steps
This evaluation points to several areas that require further attention from Vera program staff and
LOP providers.
1.	 Pro bono referrals. Most LOP providers reported an average of four pro bono referrals
per month in 2006. However, not all of the detainees who were referred to pro bono
representatives actually obtained representation; the number of successful referrals is
much smaller than the 257 reported referrals. Pro bono referrals are not the primary focus
of the LOP, but Vera and EOIR should nonetheless work with LOP providers to identify
challenges and develop steps for improving access to pro bono counsel. Vera should also
work to ensure that all LOP providers accurately report pro bono development efforts.
Vera should then document whether pro bono representatives are interested in taking
available cases and determine whether LOP providers have the resources they need to
make referrals and ensure the placement of cases. Vera should also examine whether
there are other factors that prevent referrals from being made. Finally, Vera should
consult with EOIR and LOP providers to consider what plan of action, if any, might be
taken to improve access to counsel for detainees.
2.	 Determine priorities for participation of persons who do not have active
immigration court cases. Vera should work with each LOP site to determine needs of
individuals who have signed stipulated removal orders before attending LOP sessions and
to prioritize developing materials or presentations on stipulated removal procedures in
LOP sites, particularly in those sites at which stipulated removals account for as many as
50 percent of immigration court cases. Similarly, Vera and EOIR should work together
with LOP providers to consider any legal access materials that might be relevant for
detainees not in removal proceedings, including those subject to expedited removal,
reinstatement of removal, post­removal hearing review, or with prior orders of removal.
3.	 Detainees not seen by the LOP. Vera should work with each LOP provider to
document the reasons why detainees do not receive LOP services. Then, building on what
is learned, Vera and the LOP providers should work to ensure that as many detainees as
possible are able to participate in the LOP. Vera and EOIR might also share their findings
with detention facility staff and encourage them to take care of any logistical
arrangements that are necessary to ensure full access to the program for anyone who is
interested.

Vera Institute of Justice 69

Recommendations and Next Steps: LOP Impact
While not conclusive, the impact findings we described above point to a number of areas where
additional research and program improvements are necessary, as well as several areas that might
be of interest to ICE or EOIR as those agencies continue to explore ways of ensuring access to
legal information and continued compliance with immigration court appearance requirements.
While some of the next steps we recommend are relatively short­term activities, others may
require a significant commitment of time and resources.
Case Time

Recommendation: The finding that participation in the LOP reduces case times by an average
of 13 days points to two important next steps. First, researchers should conduct further tests to
determine whether the LOP is responsible for the reduction in case processing times we are
seeing. At the same time, a cost assessment should be carried out to determine the impact of
shorter case times. Fully assessing the cost savings created by the LOP would require access to
ICE data, as it would be necessary to test the theory that court days and detention days are
correlated. EOIR should work with Vera to develop these two research activities as part of a
broader effort to assess the LOP’s potential to help EOIR achieve goals related to reducing case
completion times (as detailed in its 2005­2010 Strategic Plan) and in order to obtain definitive
information concerning LOP’s impact on the detention system.
Recommendation: Reduction in case time may impact possibilities for grants of relief. Vera
should continue to develop predictive models in order to understand the relationship between the
speed with which a case moves through the immigration court system and the chances of relief
being granted in that case.
Recommendation: Released LOP participants have longer average case times. More analysis is
needed to determine whether there are any negative effects associated with the longer case times
of released LOP participants. Additional study is also needed to determine what factors are
behind longer case times for released LOP participants. Finally, ICE data should be considered
alongside EOIR data to determine whether there is a cost to either agency associated with longer
case times for those released persons who appear at all of their hearings and comply with final
orders of removal.
In Absentia Removal Orders

Recommendation: We concluded that the low in absentia removal rates of LOP participants that
receive intensive LOP services approximate those of individuals who obtain legal representation.
This finding supports research conducted by Vera’s Appearance Assistance Program in 2000,
which also found that rates of in absentia removal rates were lower for individuals who had
access to information about the immigration court process and the consequences of failing to
Vera Institute of Justice 70

appear in court. Together, these findings suggest that EOIR and ICE will want to consider
methods for improving access to legal information as they expand alternative­to­detention
programs. In addition, these agencies might consider that studies in the field of procedural justice
have found that people are more likely to accept and comply with judicial outcomes when they
have access to accurate legal information and believe that the process is transparent and fair.
Increased representation for released persons seems to be the best way to ensure continued
immigration court appearances and avoid in absentia removal orders.
Representation

Recommendation: While the LOP has succeeded in improving access to legal information for
detained persons and appears to be assisting detainees in making informed decisions about when
to seek paid counsel, it is not clear that the LOP is working as effectively as it might to help
detainees obtain pro bono representation and to ensure access to legal representation for indigent
detainees. LOP providers face numerous challenges in this regard, including limited financial
resources and remote detention facilities with large populations. Additional work should be
carried out, with additional funding, to explore innovative ways of increasing legal
representation rates at LOP sites.
Additionally, given the limited availability of representation, researchers, LOP program
managers, and providers should work together to study which cases are most in need of
representation and which can proceed pro se. As part of that process they might determine (for
example) that legal representation is always preferable to pro se representation for certain types
of relief.
Case Outcomes

Recommendation: Vera should work with EOIR to determine how intensive services might be
used to maximize the impact of the LOP for all participants. More specifically, EOIR should
document what costs would be associated with expanding intensive LOP services to all
interested detainees at existing LOP sites. Related to this recommendation, LOP providers
should, as described above, assist researchers and LOP program managers in determining which
types of pro se activities should be enhanced and which may simply be unable to produce
desirable outcomes for unrepresented persons. Vera should also work with LOP providers to
document detainee learning in LOP program services. This process might include enhanced court
observations using standardized instruments to record how detainees perform in immigration
court after participating in the LOP.
Recommendation: The LOP is not a substitute for full representation. The government should
also consider funding direct representation, perhaps beginning with a pilot program that focuses
on a particular population (children or adults with mental illness, for example), or that tests the
impact of full representation in one immigration court.

Vera Institute of Justice 71

Efficiencies in the Immigration Court and Detention Systems

Recommendation: A study should be carried out to test the theory, stated by several detention
facility staff, that the LOP reduces detainee anxiety and behavior problems and leads to a
reduction in the use of segregated housing. Such a study could provide additional data about cost
savings or other impacts associated with the LOP. Additional research should be conducted,
using standardized observation tools, to test the notion that LOP participants understand only
“bits and pieces” of the immigration court process.

Vera Institute of Justice 72

Appendix I: Five­Year Life of Program Statistics (2003­
2008)
This report analyzes statistics through December 31, 2006. However, because report readers may have an
interest in more current numbers, we have included a table showing all program services from early 2003
when the program began through February 2008. We are additionally including a table that shows
potential claims to United States citizenship reported to Vera by LOP providers in 2007.
Figure 19: LOP Services, 2003 – February 2008
Average
Number of
Group
Presentations
Per Month

Number of
Group
Presentations

Average
Number of
Group
Presentation
Participants
Per Month

Group
Presentation
Participants

Average
Number of
Participants
Per Group
Presentation

Number of
Individual
Orientations

Average
Number of
Individual
Orientations
Per Month

Site

Court
Hearing
Location

Time Period

Denver

WSI

Jun 22, 2003 to Feb 29, 2008

825

15

10838

193

13

4858

86

El Paso

EPD

Jun 2, 2004 to Feb 29, 2008

618

14

13395

298

22

2587

58

Eloy

EAZ

Mar 7, 2003 to Feb 29, 2008

1258

22

19258

333

15

13345

231

Mira Loma

LAN

May 27, 2003 to Feb 29, 2008

585

10

16503

289

28

4258

75

Port Isabel

PIS

Feb 18, 2003 to Feb 29, 2008

1178

20

32568

540

28

4407

73

Seattle

AIR

Mar 17, 2003 to Feb 29, 2008

1122

19

11256

189

10

6144

103

Feb 21, 2003 to May 18, 2005

365

14

1621

60

4

908

34

Dec 1, 2006 to Feb 29, 2008

201

13

1627

109

8

505

34
57

Batavia

BTV

Houston

HOD

Dec 1, 2006 to Feb 29, 2008

214

14

3011

201

14

848

Newark

NEW

Dec 1, 2006 to Feb 29, 2008

146

10

1225

82

8

409

27

Laredo

LAR

Dec 1, 2006 to Feb 29, 2008

51

3

509

34

10

150

10

San Antonio

SAD

Dec 1, 2006 to Feb 29, 2008

207

14

2424

162

12

704

47

San Pedro

SPD

Dec 1, 2006 to Feb 29, 2008

110

7

854

57

8

610

41

York

YOR

Dec 1, 2006 to Feb 29, 2008

213

14

2008

134

9

503

34

San Diego

CCA

Feb 1, 2008 to Feb 29, 2008

13

13

160

160

12

64

64

17

40300

Total

7106

117257

# per year for all sites

1579

26057

8956

# per month for all sites

132

2171

746

# per workday for all sites

6

99

34

Figure 20: Potential Claims to U.S. Citizenship among LOP Participants, January 1 – December

31, 2007

Site

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec
1

# of Unique Individual
Orientation Attendees
Identified w/ Potential
Relief of U.S. Citizenship
15

Denver

1

2

0

1

1

1

3

1

0

1

3

El Paso

1

0

0

1

3

3

1

5

2

5

3

0

24

Eloy

4

8

8

4

4

2

5

3

5

9

12

6

70

Mira Loma

0

1

1

0

1

3

5

0

1

1

1

0

14

Port Isabel
Seattle

13
3

8
4

1
0

3
0

6
1

8
0

4
0

2
3

1
1

6
1

8
4

6
0

66
17

Batavia

3

1

1

1

0

0

0

0

0

1

0

0

7

Houston

1

2

2

3

2

1

6

3

4

4

5

5

38

0

1

0

0

Newark

0

0

0

0

0

0

0

0

Laredo

0

0

0

0

3

0

2

3

San Antonio

2

2

3

0

0

0

0

4

0

1

San Pedro
York

3
1

4
1

3
4

4
1

1
3

1
1

1
4

3
4

2
1

2
0

32

33

23

18

25

20

31

31

17

32

Total

1
8

2

2

16

2

0

24
22

40

20

322

Note: Sites below the black bar began services in late 2006.

Vera Institute of Justice 73

Appendix II: Data Organization and Analysis
In this appendix we describe steps taken in organizing and analyzing court data and how we
managed and re­organized data obtained from the Executive Office for Immigration Review
(EOIR).

Sources of Data
EOIR

EOIR maintains an administrative database that captures case­level information from its
immigration courts and the Board of Immigration Appeals (BIA). At the time of our study, the
immigration courts were in the process of replacing the Automated Nationwide System for
Immigration Review (ANSIR) with Case Access System for EOIR (CASE). CASE is a new
web­based system that will eventually replace ANSIR in all hearing locations. While both
systems capture the same basic data elements, CASE will record a more updated and diverse set
of elements and is intended to integrate databases for immigration courts and the BIA more
effectively than ANSIR. There are also slight differences in functionality between the two
systems that mean researchers conducting analysis with data extracted during the transition for
ANSIR to CASE may encounter challenges to working with EOIR data, which we describe in
detail throughout this appendix.
EOIR’s Office of Planning and Technology (OPAT) provided Vera with a list of hearing
locations that switched to CASE as of June 6, 2007. Knowing which sites were using each
system enabled us to better identify and correct idiosyncrasies and errors unique to each
database. We provide the list below for those readers not familiar with the dates of CASE
implementation.

Vera Institute of Justice 74

Figure 21: Rollout of Case Access System for EOIR (CASE)

Court Location
Arlington
Baltimore
Bloomington/Saint Paul
Buffalo/Batavia
Cleveland
Dallas
Eloy
El Paso
Guam/Honolulu
Harlington/Port Isobel
Hartford
Headquarters
Houston/Houston SPC
Imperial/El Centro
Lancaster
Las Vegas
Los Angeles
New Orleans
Oakdale
Philadelphia
Phoenix/Florence
Portland
Salt Lake City
San Antonio
San Diego/East Mesa
San Francisco
San Pedro
Seattle
Tuscan
Ulster/Fishkill
Varick
York

Date of Implementation
March, 2005
July, 2006
February, 2007
September, 2006
September, 2006
April, 2007
April, 2006
March, 2007
January, 2007
May, 2007
October, 2006
December, 2006
April, 2007
February, 2007
August, 2006
August, 2006
March, 2007
November, 2006
November, 2006
December, 2006
January, 2007
April, 2006
August, 2006
May, 2007
February, 2007
May, 2007
June, 2006
January, 2006
January, 2007
October, 2006
September, 2006
December, 2006

Note: All unlisted courts were still using ANSIR as of June 6, 2007

Obtaining Data from EOIR

The Vera Institute’s previous work as an INS contractor on the Appearance Assistance Program,
and our work on earlier phases of this Performance and Outcome Measurement Program
(POMP) meant Vera researchers were already familiar with data elements collected by the
immigration courts and the general reliability of different variables. During the first year of
research, Vera staff carried out a historical analysis of court data that looked at aggregated data.
In the second year, we analyzed individual­level data. Methods used during the first year of
research are described in the report submitted by the Vera Institute to EOIR in mid­2006.
For the second year of research described in this report, we requested data from EOIR that
was provided to us in three separate Microsoft Access tables labeled Bond, Proceeding, and
Appeal. The three tables list data by the case­processing proceeding level, meaning that data
Vera Institute of Justice 75

elements in the tables can be linked by either the unique proceeding number
(IDNPROCEEDING), the unique case identification number assigned by the immigration courts
(IDNCASE), or by the Alien identification number (A­number). The variables Vera requested
and received are listed below organized according to the variable names assigned by EOIR.
Figure 22: EOIR Administrative Data Received
BOND
IDNPROCEEDING
IDNCASE
Alien Number
Bond Hearing Request Date
Initial Bond Amount
Bond Hearing Date
Hearing Location Code
New Bond Amount
Bond Completion Date
Decision

APPEAL
IDNPROCEEDING
IDNCASE
Alien Number
Appeal Type
Date Appeal Filed
Filed By
Attorney E­27 Date
Attorney ID
Date of BIA Decision
BIA Decision
Lead Alien Number

PROCEEDING
IDNPROCEEDING
IDNCASE
Alien Number
NTA Date
Generation
Court Input Date
Base City Code
Base City Name
Hearing Location Code
Immigration Judge Code
Immigration Judge
Hearing Date
Initial Telephonic
Number of Charges
Charge 1
Charge 2
Charge 3
Charge 4
Charge 5
Charge 6
Asylum Application Received Date
Asylum Decision
Asylum Withholding Decision
212c Application Received
212c Application Decision
245 Application Received
245 Application Decision
VD Application Received
VD Application Decision
VD Number of Days
WD Application Received
WD Application Decision
Suspend Application Received
Suspend Application Decision
EOIR 42a Application Received
EOIR 42a Application Decision

PROCEEDING (cont.)
EOIR 42b Application Received
EOIR 42b Application Decision
Other Application Received 1
Other Application Decision 1
Other Application Received 2
Other Application Decision 2
Adjournment Date 1
Adjournment 1 Calendar Type
Adjournment Reason 1
Adjournment Telephonic 1
Adjournment Date 2
Adjournment 2 Calendar Type
Adjournment Reason 2
Adjournment Telephonic 2
Adjournment Date 3
Adjournment 3 Calendar Type
Adjournment Reason 3
Adjournment Telephonic 3
Adjournment Date 4
Adjournment 4 Calendar Type
Adjournment Reason 4
Adjournment Telephonic 4
IJ Decision Code
Proceeding Completion Code
Proceeding Completion Date
Custody
Case Type
Nationality
Language
Absentia
Decision Type
Alien Attorney Name
E­28 Date
Case ID
Date of Entry
Case Completion Date

After Vera received electronic files containing the requested data, we imported the data
contained in three tables into Microsoft SQL Server, a relational database that allows for easy
manipulation of large datasets. SQL allows researchers to easily connect data from disparate
datasets and databases without having to merge these data into single flat files. This results in
significant time savings and other efficiencies when manipulating large data sets.

Vera Institute of Justice 76

Vera’s Legal Orientation Program (LOP) Data Repository

The Vera Institute developed a Microsoft Access database called LOPster to collect
programmatic data at each subcontractor site. As described in the body of this report, LOP
providers use LOPster to record important information regarding LOP attendees and the level
and intensity of services provided. Providers report the following information to Vera each
month: alien number, first name, last name, sex, session type (group/individual/workshop),
session date and time, presenter, workshop subject if applicable, and whether or not the person
was in expedited removal. The sites also report more detailed demographic information about
participants who receive individual orientations, as well as information about types of relief from
removal or benefit applications before USCIS the detainee may have chosen to pursue. When we
designed LOPster, we built it to extract this detailed data in anonymous form, by a unique
number assigned automatically by the database and which replaces the A­number. We did this to
ensure client confidentiality. As a result, information like potential forms of relief from removal
identified by LOP providers cannot currently be matched with administrative data provided by
EOIR. LOP providers collect data such as potential forms of relief purely for the purposes of
program management and performance measurement—information which is reported each
month to EOIR in aggregate form. LOP sites submit data on a monthly basis to Vera. After
LOPster data submitted by the sites is checked for quality and accuracy by Vera researchers, they
import the data into a central repository located in Microsoft SQL Server that allows us to
aggregate the data and eventually match LOPster data with information in EOIR’s administrative
data.

Data Quality Check and Clean Up
The dataset provided to Vera by EOIR included all immigration court proceedings with a hearing
date between January 1 and December 31, 2006, amounting to 279,325 unique cases. We
included in our analysis adults with initial Master Calendar Hearings in calendar year 2006 who
were coded by the immigration courts as detained or released at some point during the first
proceeding. In order to correctly select these cases from the 279,325 provided by EOIR, we
analyzed and re­organized the data according to the steps described below.
First, we used SQL to run reports on the frequencies and distributions of the values within
each variable (e.g., how many people of each nationality were in the immigration court system in
a given month) to determine the quality and reliability of data. We looked for any illogical or
anomalous data patterns and then checked to see if these patterns reflected larger problems
within the database. With the help of immigration court administrators, OPAT staff, and other
EOIR personnel, we made decisions about when to exclude information or when to assign new
values (re­code) to data we received.
Creation of the final analytical dataset involved a number of steps in which the 279,325
unique cases provided to us by EOIR, covering January 1 to December 31, 2006, were deleted or
added through a sequence of deletion steps that have been numbered for ease of reference and
Vera Institute of Justice 77

discussion. These steps are depicted in the Figure 23 below. Each step listed in the flowchart
corresponds to a deletion or addition procedure detailed below that is also identified by number.
Generation

All EOIR records contain a variable known as the “generation.” The first proceeding in an
immigration court case will always correspond to a generation value of 99, while each
proceeding thereafter is numbered in descending order. Thus, the second proceeding in a case
will always correspond to a generation value of 98, the third proceeding to a value of 97, and so
on. In order to limit the dataset only to those with an initial Master Calendar Hearing in calendar
year 2006, we excluded all cases that did not contain a proceeding with a generation value of 99,
which meant that the initial Master Calendar began prior to 2006. Using this exclusion rule, we
omitted 32,835 cases from the dataset, or about 12 percent of the total (step 1).
Figure 23

Flowchart of Data Cleaning
ORIGINAL DATASET
279,325 unique cases
with hearing dates
between January 1 to
December 31, 2006 (data
received from EOIR)

DELETION STEPS
1) Deleted 32,835 cases with
initial Master Calendar hearings
before 2006
2) Deleted 134,635 cases that
were coded as never detained or
non­detained
3) Deleted 10,865 cases coded
with juvenile case IDs
4) Deleted 6,744 cases
identified as unaccompanied
children (UAC)
5) Deleted 6,143 cases that
were found to be detained
children’s cases
6) Deleted 20,749 cases where
the alien number was not the
same as the lead alien number
7) Deleted 94 cases due to data
entry issues
Note: These numbers are not
mutually exclusive

WORKING DATASET
103, 118 cases with initial
Master Calendar hearings
in calendar year 2006 and
custody statuses of either
detained or released

DELETION STEPS
8) Deleted 37,906 cases with
initial Master Calendar hearings
between September 1 and
December 31, 2006
9) Deleted 95 cases at new LOP
sites
10) Deleted 14 cases that had
negative case times
11) Deleted 1,309 cases that
had LOP on or after the last
hearing
12) Deleted 874 cases with a
type other than removal
13) Deleted 22,406 cases that
were Stipulated Removal
14) Deleted 1,720 cases
identified as an Institutional
Hearing Program
15) Deleted 68 cases that were
“re­detained”
Note: These numbers are not
mutually exclusive

ANALYTICAL DATASET
44,054 cases that began
between January 1, 2006 and
August 31, 2006

SUBSET 1: 7,528 cases
participated in LOP services
SUBSET 2: 30,728 cases began
at non­LOP sites and were
included in comparison groups

SUBSET 3: 5,798 cases that
did not participate in LOP
services but began at LOP sites

Vera Institute of Justice 78

Custody Status

The EOIR data we received included cases with custody statuses of detained, released, and non­
detained/never detained. We excluded all cases coded as never detained or non­detained. This
ensured that we were comparing LOP participants only to other persons whose cases began in
detention, which was indicated by a custody status of either detained or released. Ideally we
would have liked to know each respondent’s custody status at the time of the initial Master
Calendar Hearing, but EOIR tracks information about custody status and representation at the
proceeding level not the hearing level, making it impossible for us to discern the exact date of the
change in custody status. Some cases also had a custody status value that was indiscernible, e.g.,
�, which we excluded. In total, we omitted 134,635 cases that were neither coded as detained
nor released (step 2).
Children

The EOIR data we received also included records for both adults and children. We did several
things in order to ensure we identified as many children as possible.
First, we determined which values in the case ID variable identified juveniles. These values
were: J (juvenile code), J1 (juvenile has turned 18 while in proceeding), UJ (unaccompanied
juvenile), ND (NACARA dependents), and �U (an errant value that we assume to be a juvenile
since no other code uses the letter U). We deleted all cases coded with juvenile case IDs. In total,
10,865 cases were omitted on this basis (step 3).83
Typically, immigration court administrators track information on children by using flags on
the physical case files (e.g., a red sticker on the front of the file folder) and/ or entering a “J”
code as the case identifier in their administrative database. However, this is an unreliable method
for identifying all children in the immigration court system, in part because the case
identification variable in the administrative database can only contain one value and can be
overwritten, meaning a database user might overwrite the J code with a different unique
identifier.
We thus requested data from the Office of Refugee Resettlement (ORR) that identified all
unaccompanied alien children (UAC) in the care and custody of the Division of Unaccompanied
Children’s Services (DUCS) in calendar year 2006. By matching the A­numbers of the UAC to
the A­numbers in the EOIR data, we were able to identify many more children’s cases than was
possible relying on the J code alone. This step required reformatting ORR data to match the data
format used by EOIR, described below.
The Department of Homeland Security (DHS) now issues 9­digit A­numbers. However,
EOIR’s administrative databases only allow for 8­digit A­numbers. When entering A­numbers
into EOIR records, immigration court staff have to remove the left­most digit of all 9­digit
numbers and add it to the front of the last name. For example, if “John Immigrant” had an A­
number 211­999­999, the number would be entered as 11­999­999 and the last name as
83

Exclusions are not mutually exclusive. The same case may be excluded for more than one reason.
Vera Institute of Justice 79

“2Immigrant.” That same A­number was stored in the backend EOIR administrative database as
119999990. All numbers in the EOIR dataset we received were similarly formatted as 9­digit
numbers with a 0 as the right­most digit.
We formatted the A­numbers provided by ORR to match the format provided by EOIR. We
did so by removing the left­most digit from the ORR data A­number and placing a 0 to the right
of the remaining number. This allowed us to match the A­numbers of 6,744 UAC in the
immigration courts (step 4). Of these 6,744 UAC identified through our matching, 587 were not
identified as children in EOIR’s records, and 6,157 were identified with one of the “J” codes.
However, only 3,429 (58 percent) of these cases were properly identified as UAC.
Finally, we investigated which hearing locations handle UAC dockets. With help from OPAT
and the court administrators, we identified 14 UAC hearing locations. We then examined the
detained and released populations whose initial Master Calendar Hearings occurred at one of
these sites. When court administrators indicated that certain hearing location codes were only
assigned to detained children’s cases, we eliminated all cases at those hearing locations. As part
of our data cleanup, we excluded a total 6,143 cases (step 5) as follows: 376 persons with an
initial Master Calendar Hearing location code of CHI; 43 persons with a first hearing location
code of NYC; 1,324 persons with a first hearing location code of PHO; 1,793 persons with a first
hearing location code of SAD; 269 detained persons with a first hearing location code of ELP; 36
detained persons with a first hearing location code of LOS; 1,752 released persons with a first
hearing location code of HLG; 545 detained persons appearing before Immigration Judge
Margaret Burkhart with a first hearing location code of HLG; and 5 detained persons with a first
hearing location code of SND. It is possible some of these cases were for adults, but by
eliminating all of these cases, we increased our certainty that no children were included in the
analysis.
Finally, one other clue as to whether or not an alien in our dataset was a juvenile was the
filing of certain relief applications. We excluded one case where an alien applied for relief
application code CI which indicates a child without parents under the Haitian Refugee
Immigration Fairness Act (HRFA). We also excluded one case where an alien applied for CIII,
which indicates a child abandoned under the HRFA.
Dependents

We then eliminated all cases where the A­number was not the same as the lead A­number. When
the lead A­number does not match the case A­number, it is possible one case is dependent on the
other, which could skew the results of our analysis of the impact of the LOP on individual cases.
There were 20,749 cases that had lead A­numbers that did not match the case A­number (step 6).
These cases were omitted from the dataset.
Other Excluded Cases

The remaining excluded cases appear to be the result of data entry errors. One case was
associated with two different A­numbers. Ninety cases had case completion dates in calendar
Vera Institute of Justice 80

year 2005, despite the fact that the initial Master Calendar hearing occurred in 2006. Three cases
included hearing dates before 2006, which we explain below. In total, 94 cases were excluded for
data entry issues (step 7).
After excluding data according to the criteria above, we filtered the data to include only those
cases with initial Master Calendar Hearings in calendar year 2006 and custody statuses of either
detained or released. This yielded a total of 103,118 cases, our working dataset, which was the
basis for creating the final analytical dataset defined by case start dates between January 1,
2006 and August 31, 2006.

Data Organization
Case Information: Proceedings and Hearings

For the purposes of this research, Vera defines a case as the sum of all the proceedings involving
a single respondent before the immigration courts. This means in our analysis a single case may
contain numerous proceedings and numerous applications for relief that have been initiated and
decided in the time between an initial Master Calendar Hearing and the final decision issued by
the immigration judge in the last proceeding in the case. In many reports authored by EOIR,
cases are evaluated and reported at the proceeding level. Vera researchers determined that for the
purposes of our study, it would be confusing to report on proceedings as opposed to what we
defined as “cases.” This is because in EOIR’s case management system, each case—from initial
Master Calendar Hearing to final case decision—may be comprised of several “proceedings,”
which are distinguished in EOIR’s data by “generation” numbers that descend from “99.” When
a respondent’s detention status or hearing venue changes, the “case” before the immigration
courts is typically transferred from one hearing location or immigration judge to another. When
this occurs, the first proceeding in the case is closed, and a new proceeding is opened in the
immigration court records. However, the case has not been concluded and reopened. For
example, if an asylum seeker is detained at the initial Master Calendar Hearing and is later
released and granted a motion to change of venue to a different hearing location, the pending
asylum application remains active as it moves from one immigration court and judge to another,
but a new “proceeding” is opened in the records. Thus, the number of days in each “proceeding”
does not reflect how long a “case” was active in the immigration courts. Many studies of
immigration court data conflate proceeding and case when they should not.
Additionally, because we wanted to measure any potential impact of the LOP on the
immigration courts using the most consistent measures, we made a decision not to include in
case processing time any days that might have accrued after the immigration judge’s decision
was issued. When either party (ICE or the respondent) reserves the right to appeal, the case is not
completed until the appeal deadline has passed with no appeal filings, until a decision has been
issued on the appeal by the Board of Immigration Appeals, or, if a case is remanded to an
immigration judge, until a decision has been issued. There are other scenarios that might also
prevent a case from being immediately completed (or closed) after the immigration judge issues
Vera Institute of Justice 81

a decision. Because of all these reasons, our definition of “case processing time” may not match
definitions used by EOIR or other researchers. However, we believe our definition allows us to
most accurately assess time for our purpose, which is to see if LOP is associated with any
reduction or increase in the number of days a matter remains before the immigration courts.
Most of the 103,118 cases in our working dataset had only one proceeding, while 9 percent
of the cases had more than one proceeding. We re­organized this information to produce one
record for each case. In order to organize our data, we first needed to identify the date of each
respondent’s initial Master Calendar Hearing date. This task was not as straightforward as we
expected. The initial Master Calendar Hearing date is supposed to be recorded in the field
labeled “hearing date” in ANSIR and CASE. However, when we began to analyze the data, we
found that some adjournment dates were earlier than the “initial” hearing date. We discovered
from conversations with court administrators and OPAT staff that this information is captured
differently in CASE and ANSIR.
In ANSIR, the initial Master Calendar Hearing date is always entered into the hearing date
field. If there is a request to change the initial hearing date prior to the actual hearing date, the
initial date is overwritten with the new date. However, once the initial Master Calendar Hearing
is scheduled in CASE, it cannot be overwritten. Users must thus enter a new “initial” hearing
date in the first adjournment date field, with an adjournment reason “99.” If the date changes
again, the “real” initial hearing date would be placed in the second adjournment date field,
because overwriting is prohibited by CASE. Conceivably, the date in the second adjournment
date field may actually be the initial Master Calendar Hearing date, and it may be earlier than the
dates found in the hearing date and first adjournment date fields.
Therefore, for cases entered into CASE, the initial Master Calendar Hearing date may be
found in the hearing date field or any of the four adjournment date fields. After consultation with
immigration court staff, we determined that we should select the earliest date found in the first
proceeding of each case and name this as the initial Master Calendar Hearing.
The final and/or latest hearing date was easier to identify in each case because it was the final
date of the most recent proceeding. Unfortunately, we could not determine the number of
hearings between the initial Master Calendar Hearing and the last hearing date in this analysis for
a few reasons. First, we were not provided with a hearing or adjournment count that states how
many hearings or adjournments occurred in a case. Second, while the initial Master Calendar
Hearing date and corresponding data is never overwritten in the immigration court’s
administrative database, the database allows a maximum of four adjournment dates per
proceeding, after which the dates can be overwritten. In the data we received, if more than four
adjournments occurred in a single proceeding, we would not be able to reliably discern the true
number of adjournments. However, the overwhelming majority of cases in our analysis were
decided at the initial Master Calendar Hearing, making this an issue irrelevant to the majority of
cases we studied.
In ANSIR, hearing dates are displayed vertically, with the initial hearing date always shown
on top. Then, four adjournment dates are arranged chronologically from top to bottom.

Vera Institute of Justice 82

Figure 24: Screen Display of Hearing Dates in ANSIR
9/30/2006 <­­ hearing date (fixed ­ will always occupy same space on screen
10/30/2006 <­­ adjournment date 1 (no variable label)
11/30/2006 <­­ adjournment date 2 (no variable label)
12/30/2006 <­­ adjournment date 3 (no variable label)
1/15/2007 <­­ adjournment date 4 (no variable label)

In ANSIR, only four adjournment dates can be seen at once. If, for example, there are seven
adjournments in all, the four adjournment dates on the screen will be collectively moved to
another page and the remaining three adjournment dates will appear chronologically from top to
bottom. The initial hearing date remains the same (Figure 25).
Figure 25: Screen Display of Adjournment Dates 5­7 in ANSIR
9/30/2006 <­­ hearing date (fixed ­ will always occupy same space on screen

2/10/2007 <­­ adjournment date 5 (no variable label)

3/14/2007 <­­ adjournment date 6 (no variable label)

4/14/2007 <­­ adjournment date 7 (no variable label)

­­­­­­­­­­­­

<­­ (no adjournment date yet/ever)


The first set of adjournments are moved to another page.
10/30/2006 <­­ adjournment date 1 (no variable label)
11/30/2006 <­­ adjournment date 2 (no variable label)
12/30/2006 <­­ adjournment date 3 (no variable label)
1/15/2007 <­­ adjournment date 4 (no variable label)

The first set of adjournments that are moved can be called up by the user at any point. A
possible indicator of a previous set of adjournment dates is the amount of time scheduled
between the initial Master Calendar Hearing and the second hearing date for those cases with
three adjournment dates and no fourth. CASE, however, shows all hearing dates on one single
screen, so it remains a possibility that going forward, we may be able to generate a count of all
hearings.
Through the EOIR data, we can also identify the immigration judge, custody status, and
whether or not the respondent had any representation, by proceeding level. Therefore, we know
the immigration judge, custody status, and representation status at the time of the first proceeding
(but not the first hearing date). We also know the immigration judge, custody status, and
representation status for the proceeding with the final/most recent hearing date.
Representation status
EOIR records the date the E[EOIR]­28 (notice of legal representative) was filed with the
immigration court and the name of the legal representative. In ANSIR, there is room for one date
Vera Institute of Justice 83

and one name only. If there is a change in representation within the same proceeding, the E­28
date and name of legal representative will be overwritten. We would not be able to see the
change and are therefore unable to determine at what point in the case the respondent initially
obtained counsel. EOIR also instructs users to enter the record for the most recent E­28 date,
even if it is for the same attorney. For example, some attorneys will attach an E­28 with every
document they submit to the court. Each time they do so, the most recent E­28 date will be
recorded. We cannot tell with absolute certainty if the listed E­28 date is the initial filing date for
representation. In CASE, there can be multiple attorney records, but the problem is that there is
uncertainty about how the data is held for both ANSIR and CASE users. If there is an ongoing
case that is carried over from ANSIR to CASE, immigration court staff we consulted were not
sure how the data is displayed or stored in the EOIR administrative database.
Because of the limitations above and the immediate aims of our study, we focused on
whether or not the respondent had representation for a case by creating a yes/no variable for
representation. Any respondent with a populated E­28 date or representative name in the EOIR
data would receive a 1 value (meaning yes), and all others would receive a 0 value (meaning no).
At this point, we are not interested in looking at individual attorneys. These methodological
decisions have meant we are unable in this analysis to distinguish between representation for the
bond hearing only and representation in the removal proceeding.
Charges

Charges on the notice to appear issued by ICE and recorded in the EOIR data are attached to
each proceeding, like many of the variables discussed earlier. The user can enter up to six
charges per proceeding. The charges often carry over from one proceeding to the next, but
charges can either be added or dropped during each proceeding. Multiples of the same charge in
a proceeding (in ANSIR) mean that these same charges have different grounds. This typically
occurs with criminal charges but can occasionally happen with other charges. However, CASE
will not allow multiple entries of the same charge and will produce an error message.
The charges are populated fairly reliably. The charges in CASE are selected from a drop­
down menu, reducing user input error. In ANSIR, the charges are manually entered, but there is
no possibility of entering gibberish into that field because the text is controlled, meaning the
database will reject information that is not entered according to administrative rules. However,
there is always the possibility of entering the wrong charge in either ANSIR or CASE.
Because charges can be contested and dropped or ICE can add charges at each hearing, we
created a list of all unique charges that the respondent faced in a case. This way, we retain the
possibility of grouping persons based on certain charges, although the groups may not
necessarily be mutually exclusive.
Relief applications
Relief applications, too, are attached to proceeding­level data and can change by proceeding.
However, information on all relief applications, regardless of when they were filed with/accepted

Vera Institute of Justice 84

by the immigration court, are attached to the case itself. Thus, we accumulate all applications and
application decisions for each case without regard to proceeding.
We reorganized the data so that each relief application is its own variable, simply recording
whether or not a particular application was filed, with a 1 value for yes and a 0 value for no. The
only set of applications we did not separate is asylum, asylum withholding, and withholding
under the Convention Against Torture (WCAT), all of which are applied for on the same I­589
form. Upon receipt by the immigration court, both asylum and asylum withholding are recorded
as filed, but practices for recording this data are inconsistent because most court clerks will not
be able to determine from looking at an application which of these forms of relief is being
sought. WCAT, too, is applied for using the same application, but court clerks are expected to
record this application in the “other application type” variable. Given concerns expressed by
court staff about the reliability of these categories, we grouped the three applications together
into the I­589 category. Grouping the three applications also increased the size of the subset of
cases that included I­589 applications and gave us a larger group to analyze.
Unfortunately, we could not use the date found in the variable “asylum application receive
date” because of reliability concerns communicated by immigration court staff. The date of this
variable changes with each proceeding because of a programming error that prevents the I­589
date from following the case when it is transferred between locations. The clerk at the new
proceeding location will create a new date for that I­589 application, typically matching the input
data.84
Aside from the I­589 applications, the EOIR court data provides yes/no variables for the
following main types of relief applications: 212c, 245(i), voluntary departure, cancellation,
EOIR42A, and EOIR42B. A value of X for these application variables indicates receipt by the
immigration court, while a blank value or a null value means that the court did not receive an
application for that particular form of relief.
Many other, less frequently occurring relief applications are recorded using unique codes in
the two variables labeled “other application type.” We pulled each distinct application from the
two variables for “other application type” and produced yes/no variables for them. A total of
27,700 applications were filed by detained and released adults who had their initial Master
Calendar Hearing in 2006.
Case decision, completion dates
EOIR also tracks the decision for each of the applications. For our working dataset, we took the
decision for the application from the most recent proceeding in the event that there is more than
one of the same application submitted. There are three types of decisions that are found in the
original proceedings related to a respondent’s case. One decision, as discussed earlier, is
associated with an application received by the immigration court.

84

EOIR data lock in one date as the asylum application date. We did not have access to that variable.

Vera Institute of Justice 85

Another decision is the case decision handed down by the immigration judge. Cases typically
have one decision issued by the immigration judge, but, because some cases may have multiple
proceedings due to appeals or administrative opening of previously closed cases, we used the
first decision issued by the immigration judge as the case outcome in this analysis.85
The third type of decision is the one issued by the Department of Homeland Security (DHS),
referred to as an “other completion” in EOIR records. This decision is typically an administrative
order for a transfer or change of venue but can also be a termination of the case or administrative
closure or other decision to suspend the case at DHS’s request. We understand that a decision in
this field does not necessarily mean that a case is closed. It can be reopened at any time as the
case “sits on the shelf” rather than concludes.
The date that we used as the case completion date is found in the original data provided by
EOIR as the “proceeding completion date.” This is the date that accompanies the immigration
judge’s first decision in a case. Typically, this date is the same as the date of the last or latest
hearing date in a particular proceeding, but this is not always so. If the proceeding completion
date occurs after the last hearing date, there may be a written decision because the judge reserved
decision at the conclusion of the last hearing and then issued one in writing thereafter. It is
possible for a judge to issue a written decision on the same day as the last hearing, but we
learned from immigration court administrators that this is not common. Usually, when there is a
proceeding completion date on the same date as the last hearing date, the decision is an oral one.
Oddly, there are some proceeding completion dates that occur prior to the last hearing date.
However, this is only possible where data were collected using CASE. In ANSIR, system edits
do not allow the proceeding completion date to be earlier than the last hearing date. One would
have to delete or change the last hearing date to be no later than the proceeding completion date
in ANSIR. In CASE, because the user cannot change or overwrite the entries in the hearing date
field or adjournment date fields, a proceeding completion date can be entered that is earlier than
the last hearing date. For our purposes, we confirmed that this occurs at CASE sites at the time of
the first hearing date. If so, then the case time is zero days. If not, then the case was excluded.
Another completion date that exists in the original data is the “case completion date,” but we
largely ignored this variable in favor of the proceeding completion date due to its mixed
reliability and the way we constructed our working dataset. If either side, the respondent or ICE,
reserved the right to appeal, then there should be no case completion date for a case in which the
appeal period has not elapsed. Theoretically, if the parties who reserved the right to appeal do not
file an appeal within 30 days of the immigration judge’s decision, the case is closed. But this
does not often happen, according to immigration court administrators. We confirmed this by
checking to see if those with a case completion date also had an appeal filed, and they did. Also,
since the case completion date does not necessarily correspond to the first decision issued by the
immigration judge, we did not find this variable as relevant as the proceeding completion date.
85

The decision code found in the proceeding will always represent the immigration judge’s decision, regardless of
appeal. If there is a motion to reopen a case or there is a motion to re­calendar or the BIA or circuit court remands
the case, a new proceeding will be opened for that same case.
Vera Institute of Justice 86

However, we did use the case completion date to confirm that all of the cases with a case
completion date also had a decision code issued by the immigration judge. We discovered that
there were 2,100 of 103,118 cases with a case completion date but with no immigration judge
decision. However, the other completion decision variable was populated. We included these in
our working dataset.
Orders of removal in absentia
If the yes/no in absentia variable that accompanies the final decision in the case is yes, then we
coded our in absentia variable with a 1. Otherwise, the variable received a value of 0.
Nationality and language
The only detailed demographic information collected by EOIR is language and nationality. The
immigration courts enter the nationality information found on the notice to appear. This
information is not always accurate, as it is sometimes changed for a respondent by case or by
proceeding. We used the nationality from the most recent proceeding in the most recent case as
the respondent’s nationality.
Language, too, in some cases changed by case or proceeding, and it did so more frequently
than nationality. Depending on the court, the recorded language may actually be the language of
the interpreter for either the respondent or the respondent’s witness, as opposed to the
respondent’s native or most fluent language. We used the language from the most recent
proceeding in the most recent case as the respondent’s language.
LOP Services

From the data we collected from LOP providers and stored in Vera’s LOPster repository, we
matched the A­numbers reported by the LOP sites to the EOIR data. In order to do so, we had to
format the A­numbers found in LOPster to match the A­numbers found in the EOIR data. As
with the data collected by ORR, we had to remove the left­most digit of a nine­digit number and
attach a zero as the right­most digit. Because some persons had unknown A­numbers or no A­
numbers or poorly reported A­numbers, we excluded those (most records without A­numbers did
not make it into our LOPster repository in the first place).
We then created variables in our working dataset that would track the site at which the
detainee received LOP services, the level of service—group (yes/no), individual (yes/no), and
workshop (yes/no)—and the corresponding earliest dates of service. If the person received LOP
services at a second site, we also created variables to capture that second site and the additional
services.
Working Vera Dataset

After organizing the data according to the steps described above, we were left with a working
dataset including records on 103,118 unique cases that were either “detained” or “released”
during the first proceeding.
Vera Institute of Justice 87

Total Original Sample

Before we excluded any cases, the total number of cases in the original data file used for our
analysis (every court case that was coded as “detained” or “released” with an initial Master
Calendar Hearing between January 1 and December 31, 2006) is 279,325.
Cases Excluded During Analysis

After our first round of exclusions, during the data clean­up phase, we eliminated cases
enumerated throughout this appendix—largely because they were for persons who were “never
detained” according to court records, though also to exclude juveniles’ cases and problematic
records. We were left with 103,118 cases, what we have termed the working dataset.
LOP Cases

In total, there were 15,747 individuals in the 12­month matched dataset who received LOP
services. At the six LOP sites studied in our research, 15,022 received LOP services, and 725
received services at the six LOP sites that began operations in 2006.
•	 86 percent (N=13,537) participated in the LOP on or before the initial Master Calendar
Hearing, including 1,174 cases in which the initial and last hearing occurred on the same
date.
•	 6 percent (N=901) participated in the LOP between the initial Master Calendar and last
hearing.
•	 8 percent (N=1,309) participated in the LOP on or after the last hearing date (if the last
hearing date is different from the first hearing date). Of these persons, 1,259 participated
in the LOP after the last hearing date, and 50 participated in the LOP on the date of the
last hearing.
Comparison Group Cases

After 15,474 cases were identified as receiving LOP, 87,371 cases remained in our original non­
LOP group. 18,927 of these cases were scheduled at immigration courts at LOP sites, but these
respondents did not participate in LOP services. They were grouped as “LOPsites_noLOP.” We
analyzed these cases as a separate group but generally did not report on them in our analysis as
we cannot explain why these individuals did not receive the LOP. This left 68,444 cases from the
original 103,118 in the non­LOP group.
We then excluded a number of cases along several lines, often because their unique
characteristics made them inappropriate to compare to other cases in removal proceedings.

Vera Institute of Justice 88

Excluded Cases

Initial Master Calendar Hearing occurred between September 1 and December 31, 2006 (37,906
cases)
Contracts with the six new LOP sites began in September 2006, but we officially started data
collection from these sites on December 1, 2006. We decided not to include cases with an initial
Master Calendar Hearing in the months from September to December 2006 so that we could use
the cases at these new LOP sites in our comparison group from January 1 to August 31, 2006.
Without these cases, we would have had a much smaller comparison group. In all, 37,906 cases
were deleted because their initial Master Calendar Hearings fell between September 1 and
December 31, 2006 (step 8).
LOP new sites (95 cases)
Even after excluding cases whose initial Master Calendar Hearing occurred after September 1,
2006, there were still 95 respondents who received LOP services at one of the six new LOP sites
before September 1, 2006 (step 9). They were excluded as well.
Negative case time (14 cases)
There are 3,295 cases that initially appeared to have a negative value for case processing time.
Most of these negative case times were corrected, and 14 others were excluded (step 10),
according to the following rules:
•	 If the initial Master Calendar Hearing was entered into the EOIR database using CASE
rather than ANSIR, then the court completion date was changed to be the same as the last
hearing date, resulting in a case processing time of zero days. Unlike ANSIR, CASE
allows the clerk to enter a completion date that may be earlier than a previously
scheduled future hearing date. The future hearing date cannot be modified or deleted
from the system. Thus, we make the modification so that the date sequence appears
logical.
•	 Any case with negative time that occurred at a local court using CASE but prior to the
month in which CASE was implemented was excluded.
•	 If the case decision was “terminated,” the court completion date was changed to be the
same as the last hearing date.
•	 Two cases with two proceedings were also excluded because of obvious data entry errors.
LOP on or after the last hearing (1,309 cases)
There were 1,309 persons who received LOP services on or after the latest hearing date who
were excluded from our matched comparisons, though we do report on this subgroup separately
(step 11).

Vera Institute of Justice 89

Cases of a type other than removal (874 cases)
We eliminated all 874 cases not coded as removal cases, though we do report on some of these
subgroups separately (step 12). We removed cases that were referred to the courts after being in
an expedited removal process, including cases coded as asylum/withholding only, credible
fear/reasonable fear review, and claimed status review. We also removed a handful of cases
coded as continued detention review, exclusion, NACARA adjustment, and rescission. We report
separately on the cases that began in expedited removal.
Stipulated removal (22,406 cases)
We coded 22,406 of the detained/released cases as stipulated removals in calendar year 2006
(step 13). When we report national numbers and averages, we include these cases, but in our
analysis of case outcomes, we have excluded stipulated removals in order to have more
consistent comparisons.
Institutional Hearing Program (1,720 cases excluded)
We removed cases identified as Institutional Hearing Program cases from our analysis (step 14).
Custody from “released” to “detained” (68 cases excluded)
Very few individuals are re­detained after being released from custody. Of the 103,118 cases in
the analysis, only 68 cases were re­detained. Because these cases likely represent unique
situations, we did not feel it was appropriate to include them in our comparisons of other cases
that were completed in detention, so they were deleted (step 15).
Electronic Monitoring
We did not exclude cases coded as electronic monitoring or Intensive Supervision Program
(ISAP) from our overall analysis, but we did exclude them from certain analyses, such as in
absentia rates, because of the different conditions to which this released population is subjected.
Total Excluded Cases (59,064 of 103,118)

Because some cases fell into more than one category of excluded case, the total number of cases
excluded from our analysis does not equal the sum of the numbers listed above. In total, we
excluded 59,064 cases from our original sample, leaving 44,054 completed cases that had initial
Master Calendar Hearings between January 1 and August 31, 2006.
In total, we ended up excluding slightly more than half of the original LOP and comparison
group cases.
The 44,054 cases that began between January 1, 2006 and August 31, 2006 were distributed
into the following subsets:

Vera Institute of Justice 90

•	 Subset 1: 7,528 (17 percent) participated in LOP services.
•	 Subset 2: 30,728 (70 percent) began at non­LOP sites and were included in the
comparison group.
•	 Subset 3: 5,798 (13 percent) did not participate in LOP services but began at LOP sites
and were included in the analysis to look for any trends. They generally were excluded
from our report because we cannot account for contamination or the unique conditions of
these cases that may have prevented them from receiving LOP services.

Vera Institute of Justice 91