FREEDOM DENIED
iii

Table of Contents
AUTHORS & ACKNOWLEDGMENTS

P.5

EXECUTIVE SUMMARY

P.15

CONTEXTUALIZING THE CULTURE OF DETENTION

P.53

1. Publicly Available Data Provide Little Insight into the
Causes and Contours of the Federal Bail Crisis.
2. The COVID-19 Pandemic Did Not Change the Federal
System’s Culture of Detention.
3. A Judge’s Decision to Jail Someone Pretrial Has
Damaging and Enduring Effects.
4. Legal Standards at Federal Bond Hearings

FINDINGS & RECOMMENDATIONS
1

Judges Must Follow the Correct
Legal Standard at the Initial
Appearance Hearing and Stop
Jailing People Unlawfully.
TAKEAWAYS
THE LAW
THE PRACTICE
THE SOLUTION

P.84
P.87
P.93
P.107

2

P.81

Judges Must Stop Unlawfully
Jailing Poor People Without
Lawyers at the Initial
Appearance Hearing.
TAKEAWAYS
THE LAW
THE PRACTICE
THE SOLUTION

P.82

P.116
P.118
P.128
P.141

P.110

3

Judges Must Follow the Correct
Legal Standard in Presumptionof-Detention Cases to Reduce
Racial Disparities and High
Federal Jailing Rates.
TAKEAWAYS
THE LAW
THE PRACTICE
THE SOLUTION

P.148
P.151
P.158
P.168

4

Judges Must Stop Unlawfully
Jailing People for Poverty
Through Excessive
Financial Conditions.
TAKEAWAYS
THE LAW
THE PRACTICE
THE SOLUTION

P.172
P.174
P.180
P.190

P.170

P.146
CONCLUSION
APPENDIX A: BACKGROUND & METHODOLOGY
APPENDIX B: DISTRICT-SPECIFIC DATA COMPARISONS
ENDNOTES

P.193
P.199
P.217
P.241

4

AUTHORS &
ACKNOWLEDGMENTS
5

Authors
Lead Author: Professor Alison Siegler, Founding Director of the Federal Criminal
Justice Clinic and Clinical Professor of Law, University of Chicago Law School,
alisonsiegler@uchicago.edu.
Federal Criminal Justice Clinic Student Authors:
Report Team Leaders and Primary Authors:
• FCJC students (2021–2022): Clare Downing, Stephen Ferro, Angela Chang,
and Jaden Lessnick
• Project Manager (2021–2022): Jacqueline Lewittes
• FCJC students (2020–2021): Sam Bonafede and Kate Harris
• Project Manager (2021): Noadia Steinmetz-Silber
Additional Authors:
• FCJC students (2021–2022): Alessandro Clark-Ansani, Krysta Kilinski,
Mikaila Smith, and Paige Petrashko
• FCJC students (Summer 2022): Emma Stapleton and Sebastian Torrero
• FCJC students (2020–2021): Michael Belko and Adam (Psi) Simon
About the Authors: The Federal Criminal Justice Clinic (FCJC) is the
nation’s first legal clinic devoted to representing indigent clients charged with
federal felonies, pursuing impact litigation in federal court, and engaging
in systemic reform of the federal criminal system to combat racial and
socioeconomic inequities.
Additional Collaborators in the Federal Bailwatching Project
This Report grew out of a data-gathering effort called the Federal Bailwatching
Project, which was led by Professor Siegler and her students and supported by
the collaboration of many others. More details on the project and the roles of the
specific collaborators listed below can be found in Appendix A: Background &
Methodology.
• Technical Advisor: Professor Sonja Starr, Julius Kreeger Professor of Law &
Criminology, University of Chicago Law School
6

• Primary Advisor: Professor Erica Zunkel, Associate Director of the Federal
Criminal Justice Clinic and Clinical Professor of Law, University of
Chicago Law School
• Advisor: Professor Judith Miller, Faculty Member in the Federal Criminal
Justice Clinic and Clinical Professor of Law, University of Chicago Law School
• Data Managers:
• FCJC student (2021–2022): Angela Chang
• Coase-Sandor Institute for Law and Economics (2021–2022): Research
Specialist Dylan Baker and Research Professional Morgen Miller
• Jacqueline Lewittes (2022)
• Project Managers:
• FCJC intern (Summer 2022): Jacqueline Lewittes
• FCJC student (Summer/Fall 2021): Krysta Kilinski
• FCJC intern (Summer 2021): Michael Jeung
• FCJC intern (Spring 2021): Noadia Steinmetz-Silber
• Law Faculty:
• Courtwatching Site Managers:
Baltimore Site Manager: Professor Maneka Sinha, Director of the
University of Maryland Carey School of Law’s Criminal Defense Clinic
Boston Site Manager: Aliza Bloom, former Visiting Scholar, Boston
University School of Law’s Criminal Law Clinical Programs (now Faculty
Fellow, Criminal Law and Procedure, New England Law School)
Salt Lake City Site Managers: Professor Jon Powell, Director of Campbell
Law School’s Restorative Justice Clinic; and Assistant Monica Veno
• Additional Faculty Collaborators:
Harvard Law School’s Criminal Justice Institute Clinic: Professor
Dehlia Umunna, Faculty Deputy Director; Senior Clinical Instructor
Lia Monahon, Managing Attorney; and former Clinical Instructor Sadiq
Reza (now Visiting Professor of Law, Boston University School of Law)
Boston University School of Law’s Criminal Law Clinical Programs:
Professor David Rossman, Director; and Professor Wendy Kaplan
• Student Courtwatching Site Managers, Miami: FCJC students Sam Bonafede,
Kate Harris, and Stephen Ferro
• Interview Team Leaders: FCJC students Krysta Kilinski and
Alessandro Clark-Ansani
• Interview Team: Jennifer Chang and Sarah Tanner

7

• Technical Research Assistants: Jacob Jameson and Dhruv Sinha
• Chief Federal Public Defenders:
Baltimore: Chief Federal Public Defender James Wyda
Boston: Chief Federal Public Defender Miriam Conrad (ret.), Assistant
Federal Public Defender Jane Peachy, and CJA Attorney Jessica Hedges
Miami: Chief Federal Public Defender Michael Caruso
Salt Lake City: Chief Federal Public Defender Scott Wilson and Assistant
Federal Public Defender Robert Steele
• Qualtrics Experts (2020–2021): FCJC student Desiree Mitchell and Kaushal
Addanki, Research Application Developer with the Center for Decision
Research, University of Chicago Booth School of Business
• Fundraising Assistance: FCJC alum Kristin Greer Love
• Courtwatching Teams:
• Baltimore Site, University of Maryland Carey School of Law students:
Summer Akhtar, Katie Aumiller, Rose Cowan, Cynthia Faris, Rachel
Greenwald, Elizabeth Harris, Meghan Howie, Kathryn Meader, Veronica
Mina, Daniel Mooney, Emilia Lucy Petrillo, Avery Potts, Rachel Rabinowitz,
and Kelsey Robinson
• Boston Site:
Boston University School of Law’s Criminal Law Clinical Program
students: Sophia Caprio, Michelle Daker, Nicole Dill, Andrew Gourley,
Christina Morris, Alexander Phillips, Zachary Reid, Curtis St. Pierre, and
Jacob Tarjick
Harvard Law School’s CJI students: Taylor Dodson, Maddy Gates, Maggie
McGovern, Julia More, Erin Wetzel, and Gordon (Riley) Worrell
• Miami Site, FCJC students: Michael Belko, Sam Bonafede, Clare Downing,
Stephen Ferro, Kate Harris, Kathleen Larkin, Mikaila Smith, and Adam (Psi)
Simon; Project Manager: Noadia Steinmetz-Silber; Interns: Alessandro
Clark-Ansani and Molly Norris
• Salt Lake City Site, Campbell Law School’s Restorative Justice Clinic
students: Aryana Ainolhayat, Ryan Cooke, John Davis, Haley Haynes,
Ashley Holton, and Grace Massarelli
• Counsel at Initial Appearance Team: Haim (Ariel) Aiash, Jonathan Concepcion,
E.J. Morera, Adam (Psi) Simon, and Luz (Or) Tur-Sinai Gozal
• FCJC Summer Interns:
• 2022: Ricky Holder, Mabel Lewis, and Ethan Ostrow
• 2021: Jacqueline Lewittes, Miles Ogihara, Esi Koomson, Gabriela Meschino,
Molly Norris, and Shiv Puri
8

Acknowledgments
This Report is the culmination of the tireless commitment, inspiring passion,
zealous advocacy, and unwavering generosity of many over the last two years.
Words are insufficient to communicate the depth of our gratitude, but we
endeavor to acknowledge the support of those who made this Report possible.
We received a number of philanthropic contributions from individuals and
organizations across the country. Their generosity fueled all aspects of our work,
from quantitative analysis to data visualizations to the entire visual composition
of this Report. This support and encouragement is extremely meaningful and
impactful, and this Report would not exist without their engagement. We would
like to recognize the Astor Street Foundation, as well as its President/Treasurer
Sarah Olson Zimmerman, Vice President Lynde B. Uihlein, and Secretary John
U. Olson; Pegah Sadr; the Yagan Family Foundation; Leah Harp; Paul Mandell;
Michelle Williams; and Mitch Paradise.
We could not have conducted this project without the pivotal work done by the
authors and collaborators named herein. Every contributor was essential, and
many selflessly devoted hundreds—if not thousands—of hours of their precious
time to this complex and multifaceted project.
We are deeply indebted to our Technical Advisor, Professor Sonja Starr, and our
Primary Advisor, Professor Erica Zunkel. Their contributions and insights cannot
be overstated. Professor Starr and Professor Zunkel provided an irreplaceable
foundation for the data-gathering structure of the Federal Bailwatching Project
that underlies this Report. They were integral in creating and conceptualizing
this project, participating in countless early meetings and co-creating initial
versions of the project’s forms, protocols, online surveys, and trainings. This
project would not have gotten off the ground without their involvement and
effort. Professor Starr and Professor Zunkel remained invaluable thoughtpartners throughout the project, providing ongoing editorial commentary as
we honed our procedures and materials. They also contributed meaningfully to
early fundraising efforts.

9

Professor Starr oversaw many of the empirical and technical aspects of the
Federal Bailwatching Project. Her quantitative deftness was essential: she spent
myriad hours cleaning the data and conducting statistical analyses and provided
indispensable support with the interpretation of our results. In addition, she
hired and supervised the Technical Research Assistants and was extraordinarily
generous in devoting her own faculty resources to compensate them.
Professor Zunkel had the brilliant idea of using courtwatching—an efficacious
tool in state-level bail reform—to conduct an in-depth examination of the
federal bail system and co-directed the Chicago Pilot Study (2018–2019) that
served as the model for this broader project. Professor Zunkel also played a
central role in the FCJC’s legislative advocacy around the Bail Reform Act, was
instrumental in setting up this project’s first courtwatching site, and provided
outstanding feedback on this Report throughout its concept and drafting phases.
In addition, Professor Zunkel’s personal support was indispensable to completing
this enormous project.
We benefited tremendously from our Advisor, Professor Judith Miller. We are
deeply indebted for her guidance and support throughout this project. This
Report is informed by her in-depth research into numerous legal issues, including
the presumption of detention and the right to counsel at Initial Appearance. The
insights and feedback that she provided throughout the concept and drafting
phases improved this Report immeasurably.
The FCJC students who worked on this project are the unmistakable
embodiment of this Andy Warhol observation: “They always say time changes
things, but you actually have to change them yourself.” Words cannot begin to
express our gratitude to the FCJC Report Team Leaders and Primary Authors
(Clare Downing, Stephen Ferro, Jaden Lessnick, Sam Bonafede, and Kate Harris),
Project Managers (Jacqueline Lewittes, Noadia Steinmetz-Silber, Krysta Kilinski,
and Michael Jeung), Courtwatching Site Managers (Sam Bonafede, Kate Harris,
and Stephen Ferro), Interview Team Leaders (Krysta Kilinski and Alessandro
Clark-Ansani), and to Angela Chang for serving as both a Data Manager and as an
FCJC Report Team Leader & Primary Author. The whole is greater than the sum
of its individual parts; together, they were the backbone of this prodigious project
and devoted themselves to it entirely. While still students, they exemplified the
most impressive and commendable traits of the very best lawyers. They went
above and beyond at every turn, working nights, weekends, and vacations to
ensure the project’s success. Their selflessness is unmatched: several continued
to contribute to this project after their graduation while studying for the bar.
We are forever indebted to them for their unwavering devotion to our collective
mission, their tireless work ethic, and their remarkable commitment to bringing
the vision for this Report to fruition.

10

This project also benefited tremendously from the colossal contributions made
by the Law Faculty Site Managers—Professor Maneka Sinha, then-Visiting
Scholar Aliza Bloom, Professor Jon Powell, and Assistant Monica Veno; Law
Faculty Collaborators—Professor Dehlia Umunna, Senior Clinical Instructor Lia
Monahon, then-Clinical Instructor Sadiq Reza, Professor David Rossman, and
Professor Wendy Kaplan; Chief Federal Public Defenders James Wyda, Miriam
Conrad, Michael Caruso, Scott Wilson; First Assistant Federal Public Defender
Robert Steele; and their staff. Their active and enthusiastic partnership enabled
us to expand the courtwatching component of our study to 4 federal courthouses.
This Report centers around the quantitative data their support enabled us
to collect. The Chief Federal Public Defenders committed themselves wholeheartedly to getting the courtwatching off the ground, sustaining the project,
and troubleshooting issues as they arose. Our faculty partners pivoted from their
own work to dedicate themselves and their students to this joint project. We are
deeply appreciative of their willingness to venture into new territory, supervise
their students throughout the complex and intensive courtwatching phase, and
devote their staff’s energy and resources as well.
Special appreciation goes to Boston Site Manager Aliza Bloom, who was a true
partner in creating and executing the entire courtwatching project—she wrote,
edited, and contributed to countless iterations of every form and protocol,
conducted multiple trainings of students in her clinic and beyond, and advised
students across the project on courtwatching and coding techniques to ensure
the accuracy of our data.
Many thanks as well to the courts whose public access policies enabled us to
observe bond hearings via Zoom in their districts: U.S. Chief District Court Judge
James Bredar of the District of Maryland, that district’s federal magistrate bench,
and court staff; the federal magistrate judges in the District of Massachusetts;
the federal magistrate judges in the Southern District of Florida; and the federal
magistrate judges in the District of Utah. We are also grateful to the many
current and former judges who consulted with us at various points in our study,
including Judge Gabriel Fuentes, Judge Nancy Gertner, Judge Jonathan Hawley,
and Judge Evelyn Furse.
We greatly appreciate the 48 federal magistrate judges, Chief Federal Defenders,
and Assistant Federal Defenders who participated in an interview with our team.
To ensure candor, we promised interviewees that we would de-identify them,
would not list their names in this Report, and would not attribute quotes to
them directly. We therefore thank them here, collectively and individually, for
giving so generously of their time, insight, and wisdom. Despite their different
backgrounds, experiences, and perspectives, the interviewees were united by
their commitment to the fairness and transparency of the federal criminal
system. We are especially humbled by the frankness every stakeholder brought to
11

these interviews. Their interviews were essential to our ability to provide a truly
national perspective on the federal bail process.
A special acknowledgment for the Coase-Sandor Institute and our Data Managers
Morgen Miller and Dylan Baker, who graciously dedicated their data expertise
to this project for over a year. Thanks, too, to Technical Research Assistants
Jacob Jameson and Dhruv Sinha, who worked with Professor Starr and alongside
Coase-Sandor. Together, the data team spent countless hours writing code to
clean, merge, and analyze our massive data set, often working late into the night
to ensure the validity of our results. Their boundless patience in addressing our
around-the-clock requests is commendable.
We also express our sincerest thanks to the many law students from a diverse
array of legal clinics who willingly gave their time and brainpower to courtwatching, observing hearing after hearing, and meticulously coding their results
into complex online forms. Their work lies at the heart of this Report, which
would not have been possible without them.
Thanks to the entire team at Voilà:—Francis Gagnon, Timour Screve, Celia
Albano, and Jonathan R. Miranda—for creatively and industriously crafting a
visual identity for this Report and shepherding it from concept to reality.
We greatly appreciate the commitment of the Law School’s Communications
office, Associate Dean for Communications Marsha Nagorsky, Manager of
Digital Communications Quinn Baron, and Director of Information Technology David Gulbransen. Many thanks for their phenomenal website design
and production support.
The FCJC would not be able to push for systemic change without the institutional support of the University of Chicago Law School, Dean Thomas Miles, and
Professor Jeff Leslie, the Director of Clinical and Experiential Learning. Their
generous support enabled us to print and bind our Executive Summary.
Professor Siegler personally thanks Professor Zunkel and Professor Miller for
being zealous advocates, inspirational professors, dedicated change-makers, and
exceptional partners in the FCJC.
Professor Siegler also thanks the following for their unsparing support, help,
and encouragement throughout this project: Patricia Richman, David Patton,
Milagros Cisneros, Eric Vos, and the federal defender community; Premal Dharia;
Judge Ben Hernandez-Stern; FCJC alums Andrew Grindrod, Amanda Penabad,
Matthew Deates, and Professor Stephanie Holmes Didwania; Brandon Buskey,
Andrea Woods, Jennesa Calvo-Friedman, and the ACLU Criminal Law Reform
Project; Frank Draper, Craig Crawford, and the Training Division of the Defender
Services Office of the Administrative Office of the U.S. Courts; Mark Sherman,

12

Jim Chance, Cassandra Snyder, Jefri Wood, and The Federal Judicial Center; the
bail reform community, including Kanya Bennett, Twyla Carter, Sharlyn Grace,
Robin Steinberg, Alec Karakatsanis, Rachel Foran, and Dr. Rachael Bedard;
Kyle O’Dowd and the members of the Pretrial Release Working Group; Rachel
Rossi; Bill Hicks; Diane Testa; Steve Marenberg and Alison Whalen; Ali Flaum;
Robin Walker Sterling; Heather Lewis Donnell; Alaina Harkness; Damon
Hemmerdinger; and Kara Hailey. I am also grateful to my children, Zetta and
Mira, and my parents, Anna and Mark Siegler.
And finally, Professor Siegler thanks her PD mentors: Carol Brook, John Murphy,
Abbe Smith, John Copacino, Jonathan (Rap) Rapping, Steve Bright, and Glenn
Falk. Your brilliance and dedication have guided me for decades. You inspired me
to speak out in the face of injustice and to believe that system-change is possible.
Design and visualizations by Voilà: | chezVoila.com

13

14

EXECUTIVE SUMMARY
15

Over thirty years ago, the Supreme Court held that people
charged with federal crimes should only rarely be locked in
jail while awaiting trial: “In our society, liberty is the norm,
and detention prior to trial or without trial is the carefully
limited exception.”1 Given that everyone charged with a crime
is presumed innocent under the law, federal judges should
endeavor to uphold the Court’s commitment to pretrial liberty.
This Report reveals a fractured and freewheeling federal pretrial
detention system that has strayed far from the norm of pretrial
liberty.2 This Report is the first broad national investigation of
federal pretrial detention, an often overlooked, yet highly consequential, stage of the federal criminal process. Our Clinic undertook
an in-depth study of federal bond practices, in which courtwatchers
gathered data from hundreds of pretrial hearings. Based on our
empirical courtwatching data and interviews with nearly 50 stakeholders,3 we conclude that a “culture of detention” pervades the
federal courts, with habit and courtroom custom overriding the
written law.4 As one federal judge told us, “nobody’s . . . looking at
what’s happening [in these pretrial hearings], where the Constitution
is playing out day to day for people.”
Our Report aims to identify why the federal system has
abandoned the norm of liberty, to illuminate the resulting federal
jailing crisis, and to address how the federal judiciary can rectify
that crisis. This Report also fills a gaping hole in the available
public data about the federal pretrial detention process and identifies troubling racial disparities in both pretrial detention practices
and outcomes.
Federal pretrial jailing rates have been skyrocketing for decades.
Jailing is now the norm rather than the exception, despite data
demonstrating that releasing more people pretrial does not endanger
society or undermine the administration of justice. Federal bond
practices should be unitary and consistent, since the federal bail
statute—the Bail Reform Act of 1984 (the BRA)—is the law of the
land and governs nationwide.5 Yet this study exposes a very different
reality than that envisioned by the Supreme Court, one in which
federal judges regularly deviate from and even violate the law, and
on-the-ground practices vary widely from district to district.

16

This Report was researched and written by Professor Alison Siegler
and students and interns in the Federal Criminal Justice Clinic at
the University of Chicago Law School (FCJC).6 Over the course of
two years, our team conducted an extensive courtwatching study in
which we observed over 600 bail hearings across 4 federal district
courts: The District of Massachusetts in the First Circuit, the
District of Maryland in the Fourth Circuit, the District of Utah in the
Tenth Circuit, and the Southern District of Florida in the Eleventh
Circuit. With the help of faculty and clinic students from 4 other
law schools, we gathered, coded, and analyzed data about federal
pretrial detention, and mined the docketing system for additional
information. We also generated qualitative data by interviewing 48
federal magistrate judges and federal public defenders from 36 federal
district courts across 11 federal circuits.7
In this Report, we document the federal bail crisis on a national
scale.8 Professor Siegler previously testified before Congress: “The
federal pretrial detention system is in crisis . . . but its problems have
been largely overlooked.”9 We use both quantitative and qualitative
data to bring attention to this disturbing reality. Our Report highlights a troubling divergence between the written bail law and on-theground practices across the country, as well as racial disparities in
pretrial detention practices.10 The legal violations that we identify in
this Report are surely unintentional. Federal judges respect the law
and do their best to follow it. But based on our research, we conclude
that federal courts have allowed misguided and entrenched practice
norms to overshadow the law.
To rectify the situation, judges must adhere more closely to the laws
governing the pretrial process and take decisive steps to shift the
culture from one that prioritizes detention to one that prioritizes
release. This Report seeks to encourage that culture shift by:
• Describing our courtwatching data, which reveal the myriad ways
in which judges detain federal arrestees in contravention of the
legal standards in the BRA, and clarifying those legal standards;
• Furnishing qualitative evidence that our findings are replicated
beyond the 4 districts where we engaged in courtwatching;
• Highlighting the racial disparities that result from
judges’ detention and release decisions and prosecutors’ requests
for pretrial detention;

17

• Illuminating the individual and societal harms of jailing; and
• Providing a set of concrete recommendations and best practices
for judges to rectify the crisis.
Although the existing evidence shows that the federal bail
system is in crisis, it does not show why or how that crisis is
occurring. Our data provide insight into federal pretrial detention
practices that cannot be evaluated via publicly available information published by the Administrative Office of the United States
Courts (AO) or the Bureau of Justice Statistics (BJS). The AO and
BJS provide zero data about basic and fundamental aspects of
the federal pretrial detention system. The AO’s national public
data often effaces information about racial disparities and provides
little insight into the drivers of mass pretrial jailing. Although the
AO retains vast quantities of data,11 it sharply curbs public access to
information pertaining to race,12 the length of pretrial detention,13
and the miniscule rate at which people released on bond reoffend
or flee.14 The AO also fails to disaggregate data into the two distinct
stages of the federal pretrial process—the Initial Appearance and the
Detention Hearing—frustrating researchers’ abilities to understand
how pretrial detention plays out in practice.15 And neither the AO nor
BJS publicly tracks the rate at which indigent individuals go unrepresented by counsel during the Initial Appearance hearing, let alone the
race and citizenship status of people locked in jail without lawyers
at that hearing.

18

Our Report aims to identify
why the federal system
has abandoned the norm
of liberty, to illuminate the
resulting federal jailing
crisis, and to address how
the federal judiciary can
rectify that crisis.

19

1

Pretrial Detention
Is Now the Norm,
Not the Exception.

The BRA prioritizes pretrial release, placing the burden
on prosecutors to establish that a person who is
presumed innocent should be locked in jail pending
trial rather than released into the community. Under
the statute, there is a presumption of release for most
arrestees.16 The BRA’s preference for pretrial release is
further evinced in § 3142(j), which mandates: “Nothing
in this section shall be construed as modifying or
limiting the presumption of innocence.”17
In United States v. Salerno,18 the Supreme Court deemed the
BRA constitutional based on provisions that—on paper—
protect the rights of the accused. Because the statute
contained procedural and substantive “safeguards”
for arrestees, the Court found that “the provisions for
pretrial detention in the BRA” protect pretrial liberty and
render pretrial detention “[the] carefully limited exception.”19 Appellate courts agree that “[t]he default position
of the law . . . is that a defendant should be released
pending trial.”20
This Report illustrates that many of the safeguards implemented
by Congress and trumpeted by Salerno are not honored in practice.
Since the BRA was enacted in 1984, the rate at which people charged
with federal crimes are locked in jail pending trial has been on the
rise. In 1983, less than 24% of people charged with federal crimes
were detained pending trial.21 The year after the BRA was enacted,
the federal system’s pretrial detention rate increased to 29% (with
19% of arrestees held without bail and an additional 10% held on
20

This Report illustrates that
many of the safeguards
implemented by Congress
and trumpeted by Salerno are
not honored in practice.

21

financial conditions they could not meet).22 Pretrial detention rates
proceeded to soar; by 2019, people charged with federal crimes were
detained at a rate of 75%.23 See Figure 1. Even during the COVID-19
pandemic, pretrial jailing rates have remained extremely high.24 In
addition to demonstrating how far federal practice has strayed from
the presumption of innocence and the statutory presumption of
release, these exorbitant rates of pretrial detention have staggering
consequences. Every person “detained” pending trial is removed from
the community and locked in a jail cell, while every person released
returns home.25

Figure 1: Federal Pretrial Detention Rates Have
Skyrocketed Since the BRA Was Enacted (1983–2019).

100% pretrial
detention

74.8%

23.8% ,

1983 1985

1990

1995

2000

2005

2010

2015

2019

These rising jailing rates cannot be explained by prosecutors charging
individuals with more serious offenses. A recent study by the federal
courts found that, over the past ten years, the federal detention rate
has increased across all offense types, “even [after] adjusting for
the changing composition of the federal defendant population.”26
Differences in charging practices between state and federal systems
likewise do not explain the ballooning federal detention rates. The
states see a markedly higher rate of violent crime than the federal
system.27 Yet the states detain just 38% of people in felony cases and
45% of people charged with violent felonies, both a far cry from the
75% pre-pandemic federal detention rate.28
22

Not only have pretrial detention rates risen in recent decades, but
over the past 40 years, the length of time people spend in federal jail
awaiting trial has increased nearly sevenfold.29 See Figure 2. Today, an
individual who is detained pretrial spends, on average, nearly a year
in jail.30 Contrast this with the less than 2 months that the average
person who was detained pretrial spent in a federal jail cell in 1985,
the year after the BRA was enacted.31

Figure 2: The Average Length of Federal Pretrial Detention
Has Ballooned Since the BRA Was Enacted (1983–2021).

Average 1 year
length of - - - - - - - - - - - - - - - - - - - - - pretrial
detention

9 months

6 months

3months
1985

High detention rates and lengthy jail terms impose exorbitant and
unnecessary costs on taxpayers. In 2021, it cost $35,758 to put a
single person in jail for a year, a figure more than 8 times higher
than the $4,340 it cost to supervise that same person on bond in
the community.32 Based on the number of people detained pretrial
each year and the average length of their detention, we estimate
that taxpayers spend more than one billion dollars per year to pay
for federal pretrial jailing.33 The total cost of mass detention is a
substantial portion of annual allocations for all federal carceral
facilities,34 yet such high expenditures are not necessary to ensure
appearance at trial and community safety.
23

The fiscal cost of federal pretrial detention pales in comparison
to the human costs. The burden of federal pretrial detention overwhelmingly falls on poor people of color. Nationally, 81% of those
charged with federal crimes are non-white.35 In our study, 87% of the
arrestees whose cases we observed were people of color. See Figure 3.
The federal bail crisis thus exacerbates racial disparities in the
system writ large.36 On the economic front, 90% of people charged
with a federal crime do not have the money to hire their own lawyer,37
a clear indicator that most people facing pretrial detention are poor.

Figure 3: People of Color Are Disproportionately
Charged with Crimes in the Federal System.

0%

20%

40%

60%

80%

100%

I

National
Level

II
+@519

Other I Unknown

Federal Bailwatching
Project
White

Data prove that locking away so many human beings is not
necessary to promote the two goals at the heart of the BRA:
ensuring that people released on bond appear in court and do
not commit additional crimes.38 Releasing more people does not
lead to increased rates of flight or crime. In fact, the rates at which
people on federal pretrial release either fail to appear for court or are
rearrested for new crimes are extraordinarily low across the board,
with both sitting at approximately 1–2%.39 See Figure 4. Those rates
have remained vanishingly low over time, from the 1980s through
today, regardless of any changes in the federal criminal population
or the types of crimes charged.40

24

Figure 4: Even When Release Rates Increase,
Arrestees Almost Never Flee or Recidivate.

64% Release rates

20%

1.1% Failure to appear rates
1.1% Rearrest rates

1.6%
0.8%
I

I

Average for 10 districts
with lowest release
rates

Average for 10 districts
with highest release
rates

Strikingly, rates of nonappearance and rearrest are just as low in the
federal courts with the highest pretrial release rates as they are in the
districts with the lowest release rates.41 See Figure 4. And those rates
remained similarly low even as judges released slightly more people
during the pandemic.42 Moreover, these low rearrest rates certainly
overestimate recidivism, because they capture those people who were
arrested for any type of offense while on pretrial release (even a misdemeanor or driving violation).43 There is no public information about
conviction rates on pretrial release, but they are necessarily even lower
than rearrest rates.44
This evidence proves that federal judges could release far more people
pending trial without making their communities any less safe or
risking non-appearance. In fact, the federal rates are far lower than
the approximately 10% failure-to-appear and rearrest rates in what
25

are considered to be “high-performing” state-level courts, such as the
District of Columbia (D.C.) system.45 For example, the D.C. Pretrial
Services Agency reported that, over the past 5 years, D.C.’s failure-toappear rate was 10%46 and its rearrest rate was 12%,47 and the agency
set a “strategic goal”48 of maintaining these rates. A recent study of
bail reform in Harris County, Texas, similarly found an approximately
10% failure rate for both measures and concluded that this “did not
substantially impede the resolution of cases.”49 That same study
further “obtain[ed] unambiguous results that clearly show that the
increase in release rates [after reform] . . . was not associated with an
increase in future crime.”50
Given the lack of publicity around federal bail practices, it is entirely
possible that judges who jail people pretrial are neither aware
of these low failure rates nor relying on them in their detention
decisions. Instead, judges respond to institutional pressures and
misplaced fears that contribute to the culture of detention. Some
federal magistrate judges over-detain in response to the classic
“Willie Horton problem”—the fear that a released arrestee may
commit a new crime51—despite data showing that this is a statistical
improbability in the federal system. Other federal magistrate judges
may over-detain out of the more personal fear of losing their jobs,
since they serve limited terms at the discretion of the district court.52
High pretrial detention rates also fly in the face of overwhelming
evidence demonstrating that pretrial jailing does not advance its
stated purpose of ensuring appearance and community safety.
Although federal judges may believe that detaining more
arrestees will ensure community safety, evidence shows that
pretrial detention is instead criminogenic, harming individuals
and imposing additional costs on society. A series of studies has
proven that even short-term detention increases the likelihood of
reoffending by more than 25%.53 These data cast significant doubt
on the notion that pretrial detention curbs criminal activity and
benefits society. Rather, as one United States District Court judge has
observed, “Mass detention creates mass incarceration.”54
Our system of pretrial detention places significant burdens on individuals, families, and society while providing little provable benefit.
When jailed pending trial, people can face physical threats, such as
violence or difficulties in accessing necessary healthcare. Detained
people also suffer personal costs, such as employment instability,
housing instability, and the lost custody of children—all at a higher
rate than those who are released before trial.55 In addition to the
26

criminogenic effects of pretrial detention itself, jailing an individual
pending trial increases the likelihood that they will be convicted,56
sentenced to longer terms of incarceration, and face mandatory
minimums,57 which in turn impede their reentry into society.58
This Report seeks to understand why some federal courts detain
people at higher rates than are necessary to ensure community
safety. The BRA itself is partly to blame, as many of its provisions are
vague, confusing, or overbroad, and have therefore failed to provide
adequate guidance to the federal courts. The statute’s murkiness
helps explain why courts have developed pretrial detention practices
that violate the spirit or the letter of the law, as well as practices
that diverge across districts, undermining the unitary nature of
the federal system. But judges are frequently tasked with applying
convoluted laws. The complexity of the statute does not justify
allowing courthouse custom to effectively override the language and
intent of the BRA.
Our primary explanation for the legal violations documented in this
Report is the phenomenon we have labeled the culture of detention.
When the BRA offers ambiguous guidance, judges and prosecutors
interpret its provisions in ways that favor detention, either through
inadvertence, risk aversion, or both. Even when the BRA contains
clear instructions, judges and prosecutors frequently ignore those
instructions in favor of longstanding district practices, substituting
courtroom habits for the plain text of the statute and overincarcerating people in the process. For example, one judge we interviewed
justified those deviations by saying: “Oh, that’s just the way we do it.”
Chief Federal Defenders repeatedly told us that when they object that
the courthouse culture does not align with the law, the response they
are met with is usually some variation of: “Well, we’ve always done it
that way.” One Defender even coined a phrase: “We’re up against the
‘this is the way we’ve always done it’ attitude.”
The federal judiciary can rectify the federal bail crisis by scrupulously enforcing the BRA’s substantive and procedural protections.
As Justice John Paul Stevens famously said: “It is confidence in the
men and women who administer the judicial system that is the true
backbone of the rule of law.”59

27

2

Findings

Our courtwatching study of 4 federal district courts
focuses on 343 cases in which we observed the entire
pretrial detention and release process.60 To supplement
our quantitative data, we interviewed judges and
federal public defenders in the same 4 districts,61 as
well as in 32 additional federal districts. A detailed
explanation of the project’s origins, contours, and
methodology can be found in Appendix A: Background
& Methodology.
Based on these data, our Report reveals serious defects at
each stage of the federal pretrial process.
We present 4 findings about the federal pretrial detention
system, illustrating in each instance that courtroom
practices deviate sharply from the written law and fuel a
culture of detention.

28

A

Finding 1: At the Initial Appearance Hearing,
Federal Judges Jail People Unlawfully.
Our data expose a severe misalignment between the BRA’s prescribed
Initial Appearance process and the practice that unfolds in federal
courthouses around the country. We observed a problematic feedback
loop play out during Initial Appearances: the prosecutor requests
pretrial detention for reasons not authorized by the law, the defense
attorney does not object, and the judge neither questions the prosecutor nor adheres to the statutory requirements, sometimes jailing
people unlawfully. See Figure 5. When judges rubber stamp prosecutorial detention requests that deviate from the legal standard, prosecutors continue disregarding the law and judges continue jailing
people improperly in a subset of cases—in an endless cycle. The
illegal detentions that result from this mutually-reinforcing process
ultimately lead to higher jailing rates at the Initial Appearance and
beyond, and fall disproportionately on people of color. For more detail,
see Findings & Recommendations—Federal Judges Must Follow the
Correct Legal Standard at the Initial Appearance Hearing and Stop
Jailing People Unlawfully.

Figure 5: Problematic Feedback Loop at Initial Appearance

The prosecutor
cites improper
grounds for
detention at
Initial
Appearance.

The defense
attorney does
not object.

A

'

•

)

The judge neither
questions the
prosecutor nor
adheres to the
statutory
requirements.

•

Some arrestees
are jailed
unlawfully,
without a valid
basis under
§ 3142(f).

)

--------~---------

\

~

Judges don't realize that they are jailing
arrestees unlawfully, the practice further
diverges from the law, and pretrial detention
rates rise.

29

The Law
Congress and the Supreme Court envisioned the BRA as having a
narrow “detention eligibility net” that authorizes pretrial jailing for a
small subset of those charged with federal crimes.62 The parameters
of that net are explicitly set forth in 18 U.S.C. § 3142(f). Section 3142(f)
was intended to serve as a gatekeeper for federal pretrial detention
during the Initial Appearance hearing, which is the first of two
potential bond hearings authorized by the BRA. At the Initial
Appearance, a prosecutor must establish that one of the factors listed
under § 3142(f) is met in order for the judge to even hold the second
bond hearing—the Detention Hearing—where the judge determines
whether the arrestee should be detained pending trial.63 However,
if none of the § 3142(f) factors is met, the judge must immediately
release the accused at the Initial Appearance, and is forbidden from
holding a Detention Hearing at all.
If a case involves a charge listed under § 3142(f)(1) and the prosecutor
requests detention during the Initial Appearance, the judge must
hold a Detention Hearing and may order the arrestee detained
pending that hearing.64 But for cases that do not involve such
enumerated charges—which we call “non-(f)(1) cases”—the judge
may hold a Detention Hearing and detain the arrestee pending
that hearing only if there is a serious risk that the arrestee will flee,
obstruct justice, or threaten, injure, or intimidate a prospective
witness or juror.65

30

Although federal judges
may believe that detaining
more arrestees will
ensure community safety,
evidence shows that
pretrial detention is instead
criminogenic, harming
individuals and imposing
additional costs on society.

31

Initial Appearance Standard
Federal Bailwatching Findings

• In 81% of Initial Appearances in our study where the prosecutor requested detention, the prosecutor asked the judge to
hold a Detention Hearing without citing any legal basis under
§ 3142(f). In some of these cases, prosecutors cited invalid bases for
requesting a Detention Hearing, such as danger to the community
or non-serious risk of flight. See Figure 9.

81%

• In over 99% of Initial Appearances where the prosecutor requested
detention without citing a valid basis under § 3142(f), judges
detained people without questioning prosecutors’ grounds for
detention. See Figure 9. This created a problematic feedback
loop in which a prosecutor’s request for detention at the Initial
Appearance almost always resulted in a judicial order of
detention, even when based on improper grounds. See Figure 5.

• In 12% of Initial Appearances where the prosecutor
was seeking detention, judges entered a detention
order even though no statutory basis for detention
existed under § 3142(f). These detention orders,
therefore, were flatly illegal under the BRA.
See Figure 9.

99%
32

99%

12%

Racial Disparities
at the Initial Appearance
White people
People of color

White people
Black and Latino

21%
44%

• Prosecutors sought detention in cases that did
not qualify for a Detention Hearing under § 3142(f)
(1) more than twice as frequently for non-white
arrestees. See Figure 10.

23%
35%

• Prosecutors similarly cited improper grounds for
detention more frequently against Black and Latino
arrestees. See Figure 11.

Citizenship Status Disparities
at the Initial Appearance
Citizens

68%

Noncitizens

85%
79%
78%

• Prosecutors requested detention nearly 20% more frequently
if the arrestee was identified as a noncitizen. At 79% of these
hearings, prosecutors failed to cite a valid basis for detention
under § 3142(f). See Figure 15.
• In noncitizen cases, no judge questioned the prosecutor’s
grounds for detention when they failed to cite a valid basis for
detention at the Initial Appearance, leading judges to detain
arrestees in 78% of these cases.

33

B

Finding 2: At the Initial Appearance Hearing,
Federal Judges Unlawfully Jail Poor People
Without Lawyers.
Our study uncovered an egregious access-to-counsel problem in the
federal system: judges in more than one-quarter of federal district
courts do not provide every arrestee with a lawyer to represent
them during the Initial Appearance. See Figure 6. In fact, 72% of the
36 districts where we interviewed or surveyed stakeholders deprive
at least some individuals of counsel at this first bail hearing.66
This finding is particularly concerning given that 90% of those
charged with a federal crime cannot afford a lawyer.67 Jailing people
without lawyers definitively violates federal law and may violate
the Constitution. These deprivations of counsel also contribute to
high pretrial detention rates and exacerbate racial disparities. In
our study, every arrestee who was deprived of a lawyer at the Initial
Appearance was jailed, and nearly all were Black or Latino. For more
detail, see Findings & Recommendations—Federal Judges Must
Stop Unlawfully Jailing Poor People Without Lawyers at the Initial
Appearance Hearing.

Figure 6: There is a Nationwide Access-toCounsel Crisis in the Federal System.

• • • • • • • • •
• • • • • • • • •

More than one-quarter of the
94 federal district courts
nationwide do not ensure
that every single arrestee is
represented by a lawyer at
the Initial Appearance .

•••••••••
•••••••••
•••••••••
••••••••
••••••••
••••••••••
••••••••••
••••••••••
34

The Law
Federal law requires judges to ensure that anyone who cannot afford
a lawyer is represented by court-appointed counsel during their
Initial Appearance hearing. Under the law, every individual accused
of a federal crime must be represented by counsel “at every stage of
the proceedings from his initial appearance,”68 rendering it unlawful
for a judge to fail to appoint lawyers to represent indigent arrestees at
the Initial Appearance. The right to counsel at the Initial Appearance
is further supported by the principles underlying the Fifth and Sixth
Amendments and public policy.

35

Access to Counsel at Initial Appearance
Federal Bailwatching Findings

• Our interviews and survey data revealed that more
than one-quarter of federal district courts fail to
appoint a lawyer for every arrestee at the Initial
Appearance, with at least 26 of the 94 federal districts
exhibiting this problem. See Figure 6.

28%

• In one district where we courtwatched, 11% of
arrestees went unrepresented for the entirety of
their Initial Appearances, with no lawyer by their side
to advocate for their liberty interests. See Figure 16.

11%

• In that district, every single individual who faced their Initial
Appearance without a lawyer was jailed after the hearing, a
100% detention rate. See Figure 17.

100%

• In all 4 districts in our courtwatching study, when
arrestees were forced to proceed without counsel for
some part of their Initial Appearance, there was a
notable increase in pretrial detention: across courtwatched districts, partially represented individuals
were detained 89% of the time, while fully represented individuals were detained 67% of the time.
See Figure 18.

100%
36

89%
67%

Racial Disparities
in the failure to provide counsel before
deprivations of liberty

92%

• 92% of the arrestees who were unrepresented at
their Initial Appearances were people of color.
See Figure 17.

• Judges unlawfully detained unrepresented individuals in violation
of § 3142(f) in some of the Initial Appearances we observed,
compounding the harm of not providing a lawyer.
• We also observed Initial Appearances where arrestees made
incriminating statements while judges questioned them
without a lawyer, jeopardizing the person’s Fifth Amendment
right against self-incrimination and their ability to fight the case
in the future.

37

C

Finding 3: Federal Judges Misapply the
Presumption of Detention.
Judges have the power and responsibility to limit the impact of
the statutory “presumption of detention” that sometimes applies
during the Detention Hearing, but our study finds that they
routinely apply the presumption incorrectly, giving it more weight
than the law allows and failing to assess whether the presumption
has been rebutted.
At the Detention Hearing—the second pretrial hearing in the federal
system—a rebuttable presumption of detention applies in certain
types of cases.69 The presumption does not mandate detention.
Instead, courts of appeals have set an easy-to-meet rebuttal standard.
Our study found that judges overwhelmingly fail to find the
presumption rebutted, a clear indication that they are not adhering
to the legal standard in presumption cases. This misuse of the
presumption of detention causes many more people to be detained
pending trial than necessary, and it results in more burdensome
conditions of release in the rare cases in which judges grant release.
Additionally, since people of color face charges triggering the
presumption more often than white arrestees, the misapplication of
the presumption exacerbates racial disparities in the federal criminal
system. Judges’ treatment of the presumption of detention is particularly important given the prevalence of presumption-triggering
charges; nationally, the presumption of detention applies to 93% of
all federal drug offenses.70 For more detail, see Findings & Recommendations—Federal Judges Must Follow the Correct Legal Standard in
Presumption-of-Detention Cases to Reduce Racial Disparities and
High Federal Jailing Rates.

38

The Law
The BRA creates a rebuttable presumption of detention for certain
offenses. Because the presumption imposes such a heavy cost on
individual liberty, courts have set a standard that should make it easy
for an arrestee to rebut the presumption.71 This low rebuttal standard
matches Congress’s original intent for the presumption, which was
to lock up only “the worst of the worst” offenders.72 In every Detention Hearing in a presumption case, the law requires a judge to:
(1) determine whether the presumption has been rebutted under the
legal standard articulated in case law; and (2) weigh the presumption
against all of the other pretrial release factors listed in § 3142(g),73
keeping the burden of proving that detention is warranted on the
prosecution at all times.

39

Presumption of Detention
Federal Bailwatching Findings

• In our study, arrestees facing a presumption of
detention were detained at a rate of 72%, which
exceeded the rate of detention among arrestees to
whom a presumption did not apply by nearly 20%.
See Figure 19.

72%

• Judges detained arrestees facing a presumption of detention in
87% of the cases in which the prosecutor explicitly invoked the
presumption during the Detention Hearing, compared to 76% of
the time when the prosecutor did not invoke the presumption.

87%

• In 95% of the contested Detention Hearings we observed where the
presumption of detention applied, judges either failed to mention
whether the presumption of detention was rebutted or concluded
that the presumption was not rebutted. See Figure 20.

• In 100% of Detention Hearings where the judge found that the
presumption had not been rebutted, the judge detained
the arrestee.

40

55%

76%

95%

100%

Detained when
presumption
applied
Detained when
presumption
did not apply

Racial Disparities
in presumption-of-detention cases

89%

• In the cases where a presumption of detention applied,
89% of the arrestees were people of color.

97%

• Among the Detention Hearings where prosecutors
invoked the presumption of detention, 97% of the
arrestees were people of color. See Figure 21.

100%

• Prosecutors erroneously invoked the presumption
of detention exclusively against Black or Latino
arrestees.

People of color
detained

73%

White people
detained

68%

• Judges detained people of color at higher rates
than white individuals: the detention rate in
presumption-of-detention cases involving people
of color was 73%, while the detention rate in
presumption-of-detention cases involving white
arrestees was just 68%. See Figure 22.

100%
41

D

Finding 4: Federal Judges Impose Excessive
Financial Conditions that Violate the Law
and Jail People for Poverty.
Our study shows that judges consistently impose inequitable
and burdensome financial conditions of release. Some courts jail
arrestees simply because they are too poor to pay for their release,
thereby violating the BRA. These practices contribute to high pretrial
detention rates and have a disproportionate racial impact, further
aggravating racial disparities in the federal system. For more detail,
see Findings & Recommendations—Federal Judges Must Stop
Unlawfully Jailing People for Poverty Through Excessive Financial
Conditions.
The Law
The BRA unequivocally states that judges “may not impose a
financial condition that results in the pretrial detention of the
person.”74 Congress intended this provision to end the practice of
imposing “pretrial detention through the arbitrary use of high money
bail.”75 Congress hoped the statute would end one of the primary evils
of cash bail systems: caging poor individuals by conditioning their
release on their ability to pay. Although the statute authorizes judges
to condition release on financial requirements,76 such conditions are
prohibited if they result in the accused being jailed for poverty.

42

Because many people
charged with federal crimes
are indigent, financial
conditions of release run a
serious risk of serving as
de facto detention orders,
an indisputable violation
of the BRA.

43

Financial Conditions
Federal Bailwatching Findings

• In 37% of cases in our study, judges imposed
financial conditions, such as personal surety bonds,
corporate surety bonds, and the upfront posting of
collateral. See Figure 23.

37%

• In 34% of all cases and 91% of cases where financial
conditions were imposed, judges required arrestees
to post a secured bond, reintroducing the evils of cash
bail systems that the BRA sought to avoid.
See Figure 23.

34%

• In one district where we courtwatched, arrestees
were detained in 40% of cases involving financial
conditions solely because they did not have the money
to pay for their release. See Figure 25. Judges in that
district regularly imposed federal bail bonds known as
corporate surety bonds (CSBs). In 92% of cases where
a CSB was imposed, the accused was locked in jail
because they were unable to obtain a bail bond.
Every single individual subjected to a CSB was a
person of color. See Figure 24.
• Across all 4 districts, arrestees did not have the
money to meet financial conditions in 36% of cases
where such conditions were imposed. See Figure 23.
In fact, 21% of all arrestees detained at the Initial
Appearance remained in jail after the Detention
Hearing because they could not meet financial
conditions of release. For these individuals, the
financial conditions acted as de facto detention orders,
in violation of the law.

44

40%
92%

36%
21%

Racial Disparities
in the imposition of financial conditions of release

95%

• Black and Latino arrestees were much more likely
to face financial conditions of release than white
arrestees; among arrestees on whom secured bonds
were imposed, 95% were people of color. See Figure 26.

95%
45

3

Recommendations

In Salerno, the Supreme Court described the accused’s
“strong interest in liberty” as “importan[t] and
fundamental.”77 Our study shows courtroom custom
overriding the legal standards that were supposed
to preserve that fundamental right. To comply with
the law, federal judges must start from the statutory
premise that pretrial release is the default and that
pretrial detention can be justified only if no conditions
of release will reasonably assure the accused’s
appearance in court and community safety. Judges can
mitigate the culture of detention by abiding by the BRA
and other federal laws. Adhering to the rule of law, as
we describe in the below recommendations, would also
mitigate racial and socioeconomic disparities in the
federal criminal system. To align the on-the-ground
practices with the law, we recommend that judges
do the following.

46

A

Recommendation 1: At the Initial
Appearance, judges must prevent unlawful
detentions by following the § 3142(f)
legal standard.
Judges violate federal law during Initial Appearance hearings when
they set or hold Detention Hearings without any legitimate basis
for detention under § 3142(f). Judges must understand the legal
standard that applies at the Initial Appearance and enforce the strict
limitations on pretrial detention provided in the written law.
First, judges must break the problematic feedback loop by requiring
prosecutors seeking detention at the Initial Appearance to cite a
specific § 3142(f) factor listed in the BRA. If a prosecutor presents
improper grounds for holding a Detention Hearing, such as “danger
to the community” or non-serious “risk of flight,” the judge must deny
the prosecutor’s request and promptly release the arrestee (unless
§ 3142(f)(1) independently authorizes a Detention Hearing). Second, in
cases that do not fall under § 3142(f)(1), judges must hold prosecutors
to their burden of proof under § 3142(f)(2). In such cases, a judge must
release the arrestee at the Initial Appearance unless the prosecutor
justifies their request for a Detention Hearing by presenting individualized facts and evidence establishing a “serious risk” of flight or
obstruction of justice. In all cases, judges should vigilantly adhere
to the statute and ensure that no arrestee is unlawfully jailed at the
Initial Appearance.
Our recommendations relating to the Initial Appearance are
discussed in Findings & Recommendations—The Solution: At the
Initial Appearance, Judges Must Prevent Unlawful Detentions by
Following the § 3142(f) Legal Standard.

47

B

Recommendation 2: At the Initial
Appearance, judges must follow the law
and appoint lawyers to actively represent
every indigent arrestee.
Judges violate federal law, the Federal Rules of Criminal Procedure,
and potentially the Sixth Amendment when they jail poor people
without lawyers or force someone to appear pro se opposite a federal
prosecutor during the Initial Appearance. To comport with the law
and safeguard the accused’s liberty interest and constitutional rights,
judges must provide every indigent arrestee with an appointed lawyer
to actively represent them throughout their entire Initial Appearance
hearing. It is not sufficient to have a defense lawyer on standby.
Rather, the law entitles every arrestee to a lawyer who is functioning
in an adversarial capacity and can vindicate their client’s rights under
the legal standard in § 3142(f). Judges should follow best practices by
appointing counsel before questioning people, therefore protecting
arrestees’ Fifth Amendment right against self-incrimination.
Our recommendations relating to the appointment of counsel at the
Initial Appearance are discussed in Findings & Recommendations—
The Solution: At the Initial Appearance, Judges Must Follow the Law
and Appoint Lawyers to Actively Represent Every Indigent Arrestee.

48

C

Recommendation 3: At the Detention
Hearing, judges must adhere to the low
standard for rebutting the presumption of
detention and never treat the presumption
as a mandate for detention.
Judges give the presumption of detention more weight than the
law authorizes, treating it as a heavy thumb on the scale in favor
of pretrial jailing. Courts must instead treat the presumption
as genuinely rebuttable and give it the low weight to which it is
entitled. To adhere to the law, during every Detention Hearing in a
presumption case, a judge should apply a two-step analysis.
First, the judge must determine whether the presumption is rebutted,
a question that turns not only on the defense presentation but also
on all of the evidence in the record (including the Pretrial Services
Report). In this rebuttal analysis, the judge should follow case law
and hold that an arrestee has rebutted the presumption as long as
there is “some evidence” that the arrestee will not flee or endanger
the community if released.78 The judge’s finding as to whether
the presumption of detention has been rebutted should be stated
on the record.
Second, regardless of whether the presumption has been rebutted,
the judge must weigh all of the factors listed in § 3142(g) to reach
the ultimate release or detention decision.79 That ultimate determination must adhere to the legal standard in § 3142(e): as in any
case, the judge must release an arrestee in a presumption case if the
prosecutor has not proven by clear and convincing evidence that
there are no conditions that would “reasonably assure” that person’s
appearance in court and the safety of the community. Throughout
the analysis, the judge should never shift the burden of proof to the
defense or treat the presumption as a mandate for detention.
Our recommendations relating to the presumption of detention
are discussed in Findings & Recommendations—The Solution: At
the Detention Hearing, Judges Must Adhere to the Low Standard
for Rebutting the Presumption of Detention and Never Treat the
Presumption as a Mandate for Detention.

49

D

Recommendation 4: At both pretrial
hearings, judges must stop imposing
financial conditions that result in detention
and tailor release to each arrestee’s
individual economic circumstances.
Judges violate federal law when they impose inequitable and
burdensome financial requirements for release, especially when they
jail people who are too poor to pay for their freedom. In such cases,
financial conditions of release function as de facto detention orders,
contravening the plain language and spirit of the BRA. Instead,
judges must recommit to making individualized release decisions
and thoroughly consider whether financial conditions—including
bail bonds, cash bonds, and “solvent surety” requirements—are truly
the least restrictive conditions available. Such individualized determinations are critical to aligning the practice with the law as written,
since the vast majority of people charged with federal crimes are poor.
Our recommendations relating to financial conditions of release are
discussed in Findings & Recommendations—The Solution: At Both
Pretrial Hearings, Judges Must Stop Imposing Financial Conditions
that Result in Detention and Tailor Release Conditions to Each
Arrestee’s Individual Economic Circumstances.

50

51

52

CONTEXTUALIZING
THE CULTURE OF
DETENTION
53

Although a robust bail reform movement has
been combatting cash bail in state courts for
years, skyrocketing federal pretrial detention
rates have gone largely unnoticed. This Report
reveals a widespread culture of detention
in the federal system, with courtroom
practices around the country deviating
markedly from the law.
Contrary to the purpose of the Bail Reform Act of 1984 (BRA) and the
Supreme Court’s holding in United States v. Salerno,80 pretrial jailing
is now the rule, and pretrial release is the rare exception. Publicly
available data establish that federal judges detain people pretrial at
much higher rates than are necessary to achieve the twin purposes of
the BRA—ensuring community safety and appearance in court. The
pervasive consequences of these high jailing rates extend far beyond
the term of detention itself, ultimately harming not only the accused,
but also their families, communities, and society. In short, the costs
of mass detention outweigh any societal benefit.
One of the key obstacles to ascertaining the full scope of the
federal bail crisis is the lack of national public data on federal
pretrial detention practices. The available data do not provide any
insight into why federal detention rates have been rising for decades,
despite evidence showing no corollary increase in community safety.
The existing data likewise provide no information about how the
pretrial detention process plays out in day-to-day practices, nor do
they discuss racial disparities in specific federal pretrial detention
practices. Without more granular data, it is impossible to identify the
causes of the federal bail crisis and pinpoint necessary reforms.
This Report fills that gap. Through courtwatching—observing federal
pretrial hearings—and interviewing stakeholders, we gathered an
abundance of information about the interplay between judges, prosecutors, defense lawyers, and arrestees in courts across the country.
Our courtwatching centered on federal courts in Baltimore, Boston,
Miami, and Salt Lake City. We expanded the scope of our study to
the national level by conducting interviews that illuminate pretrial
detention practices in many other federal courts. Our quantitative
and qualitative research shows that the substantive limitations on
pretrial detention established by Congress and the Supreme Court
are not respected in practice.
54

This section provides context for the Findings & Recommendations
that follow. This section first describes the lack of information
about federal bail practices. It then establishes that the COVID-19
pandemic did not alleviate or exacerbate the federal bail crisis in a
way that would limit the import of this Report’s findings. The many
interviews we conducted made clear that the systemic problems
identified within this Report are extensive and long-term obstacles
that existed well before the pandemic. The weight of our findings is
thus not in any way minimized by the fact that we gathered our data
during the pandemic. This section next discusses the devastating
consequences of pretrial jailing, including its detrimental impact
on people’s physical and mental wellbeing, its adverse effects on an
arrestee’s ultimate case outcome, and the enduring harm pretrial
jailing wreaks on arrestees and their families. Finally, this section
clarifies the statutorily defined standards that govern federal pretrial
detention to provide the legal backdrop for our groundbreaking
findings and urgent recommendations.
We hope that this Report leads federal judges and other stakeholders to adhere to the letter of the law and resist the institutional pressures and implicit biases that created and sustain the
culture of detention. By following the law, judges can safeguard
the liberty interests of people accused of federal crimes and
mitigate the dire consequences of federal jailing detailed below.

Key to read the charts
Each square represents an individual
hearing observed during our study.

! subsets of individuals are separated
into a nesting box to highlight
differences in treatment.

l

Detained arrestees are emphasized in
red. Different colors also indicate
different treatment across racial groups.

□□□□□□□□□□
□□□□□□□□□□
□□□□□□□□□□

······~
••••••••••
•••••••••
••
••••••••••
55

1

Publicly Available Data
Provide Little Insight into
the Causes and Contours
of the Federal Bail Crisis.

This Report provides a new and far more detailed
picture of the federal bail system than the very
limited publicly available data. With existing data,
it is impossible to study key questions like whether
judges are following the law and whether race impacts
detention decisions. The limited extant public information does not hint at the reality our data reveals:
federal judges regularly deviate from the law when
jailing people prior to trial.
While the Administrative Office of the U.S. Courts (AO)
possesses vast quantities of data, it strictly limits public
access to that information.81 As noted in the Executive
Summary, the AO does not provide public access to its
federal data about race, length of pretrial detention,
and low rates of recidivism and flight. In addition, the
AO inexplicably denies taxpayers information about
the annual cost of federal pretrial detention.82 Data
tables relating to these and many other matters are
housed on the “J-Net,” the federal judiciary’s intranet
website, but are kept secret from anyone without
access to that system.
Meanwhile, the limited data the AO does publicize lumps important
and independent data points together in overly broad categories.
Although the AO provides some information about the rate at which
prosecutors seek detention and the subsequent rate of detention
56

ordered by judges, that information is not disaggregated to allow
an understanding of how those issues play out at each stage of
the pretrial process: the Initial Appearance hearing, the Detention
Hearing, and post-Detention-Hearing proceedings.83 Nor does the AO
data provide any information about prosecutors’ bases for seeking
detention or judges’ bases for detaining people.
Although the Bureau of Justice Statistics (BJS) publishes some
reports about federal pretrial detention, its publications suffer from
their own host of problems. For example, while BJS tracks the type of
crime charged, it often uses general categories, like “violent” offenses
or “property” offenses, and never links offense types to the BRA’s
provisions. This makes it impossible to accurately discern basic
facts about federal pretrial detention, including: the percentage of
people charged with offenses that render them statutorily eligible
for a Detention Hearing84 versus the percentage of people charged
with offenses that do not qualify for a Detention Hearing,85 as well
as the detention rate for each of these groups; the rate at which
people are charged with offenses that carry a statutory presumption
of detention;86 detention rates in presumption-of-detention cases;
and detention rates in non-presumption cases. Moreover, some BJS
reports combine many years’ worth of data into one statistic, making
it impossible to identify annual trends.87 And BJS publishes reports
infrequently, sometimes waiting nearly a decade between reports.88
AO and BJS data fail to shed light on numerous aspects of the
federal pretrial detention process. For many pretrial detention
measures, BJS does not disclose race or citizenship data, which
obscures racial disparities. For example, BJS provides no data about
the race and citizenship status of people subjected to financial
conditions of release, much less a racial breakdown of those detained
because they are too poor to meet a monetary condition. Moreover,
many BJS reports make it impossible to understand the link between
race or citizenship data and pretrial detention outcomes for specific
federal offenses.89 While BJS’s most recent federal pretrial detention
report links race with pretrial detention outcomes by offense,90 it
does not disaggregate Latino arrestees from white arrestees, further
obscuring potential racial disparities.91
As Professors Siegler and Zunkel wrote in 2019, “[w]hile there is
disturbingly little published data about the race effects of federal
pretrial detention, the few studies that exist show consistent racial
disparities over time, with people of color being detained at higher
rates than [w]hite people.”92 For example, a 2018 study found that
57

in the federal system, Latino and Black individuals are detained
at rates of 88% and 60%, respectively, while white individuals are
detained at a rate of 45%.93 Although these data are helpful to understanding the racial disparities exacerbated by the federal pretrial
system, the particularity and frequency of such publicly available
research remains insufficient. In fact, “the first nationwide analysis
of race- and gender-based disparity in federal pretrial detention” was
not published until 2021.94
Our Report provides a comprehensive picture of the federal pretrial
detention system that fills in many of these gaps. It further shows
that it is possible for entities like the AO and BJS to give the public
far more information than they currently make available. To address
the lack of publicly available data about race and federal pretrial
detention, we gathered information about the observable race of every
person whose hearing we watched.95 This revealed racial disparities
in many aspects of the process. Additionally, by recording whether
or not an arrestee was identified as a U.S. citizen, we discerned the
variations in pretrial detention processes and outcomes experienced
by citizens and noncitizens.96
We hope that the AO and BJS will revise their data dissemination
practices to ensure that the public has access to the most comprehensive data possible. Without this information, stakeholders
and members of the public are unable to evaluate the necessity
and efficacy of proposed reforms, thereby allowing the culture of
detention to continue unabated.

58

2

The COVID-19 Pandemic
Did Not Change the
Federal System’s
Culture of Detention.
Although we conducted our courtwatching study
during the COVID-19 pandemic, our findings are
broadly generalizable. The pandemic took an unprecedented and deadly toll on people in jail and prison.
It also led many federal courts to provide remote
access to Initial Appearances and Detention Hearings,
a development that enabled us to courtwatch in
4 geographically distinct districts. Despite the
remote format and the ever-present threat of
infection, empirical and qualitative evidence show
that the federal culture of detention is not unique
to the pandemic.
The pandemic neither exacerbated nor mitigated the
gulf between law and practice that exists in the federal
system. The 4 central problems highlighted in our
Findings & Recommendations appeared long before the
pandemic and will persist long after, unless and until
federal judges take action. When we shared our courtwatching findings with judges and federal defenders
in 36 federal courts, they uniformly confirmed that the
divergence between the statutory text and courtroom
custom is a longstanding dilemma in the pretrial arena
that predated the pandemic. That observation corresponds with the FCJC faculty’s own experience in two
decades of federal court practice.
59

National data and our quantitative and qualitative research show
that the pandemic also did not meaningfully mitigate the federal
system’s overuse of pretrial detention.97 During the early years
of the pandemic, the average length of time people spent in federal
jail increased markedly, from 253 days in 2019 to 346 days in 2021.98
Nevertheless, the potentially deadly effects of the pandemic were
rarely discussed at pretrial hearings or during judges’ detention and
release determinations. Our courtwatching data reveal that defense
counsel mentioned the pandemic in only one Initial Appearance out
of the 343 in our study,99 defense counsel mentioned the pandemic in
just 11% of Detention Hearings, and arrestees were detained in 76% of
those cases. As for the health risks posed by pretrial detention during
the pandemic, defense counsel argued that an arrestee was particularly vulnerable to COVID-19 in only 6% of the handful of Detention
Hearings where the defense mentioned the pandemic, and 73% of
those arrestees were detained.
During our interviews, judges and Federal Defenders across the
country reported that, in their experience, COVID-19 did not make
judges any less likely to jail someone.100 One judge candidly declared
that very few judges “considered the pandemic as a reason not to
detain somebody if they deserve[d] to be detained.” One Defender
described how remote hearings posed an additional barrier to release:
“[B]ecause [being in person] really gives much more nuance to who
the individual is . . . [it] make[s] it a little harder to say no [to release]
in someone’s face.” That Defender reported no increase in leniency
during the pandemic. While there was an initial push for release in
the early days of the pandemic, one Chief Federal Defender observed
that “[a]s the year [2020] went on . . . [prosecutors] started ratcheting
up in terms of their detention requests. . . . [B]y July [2020], they were
asking for detention for everybody.”

60

3

A Judge’s Decision to
Jail Someone Pretrial
Has Damaging and
Enduring Effects.
I had a woman yesterday who was ordered detained, and you know,
there was 37 pounds of meth. But this is a woman who had been
a prostitute at least twice, she had a property crime about four
years ago, these are all clustered together, at the same time she was
divorcing her husband who sexually abused one of her kids—now
he’s in prison for a long time—and she was not a well-educated lady.
I think she had some other stuff maybe about 15, 16 years earlier, and
she had some traffic warrants, but my point to the judge was, I can
clear these warrants, this is a woman [whose] . . . crimes are crimes
of poverty and desperation. . . . These people’s lives are abysmal, I
mean you don’t do that [prostitute yourself] because you’ve got power.
You do it because you’re desperate, and you know, you’re willing
to basically let someone use you. So anyway, the judge, . . . he just
detained, and I don’t . . . I realize 37 pounds of meth, that’s a lot of
meth, but they tried to sort of characterize her as this sophisticated
kind of person. I think she was sort of put up to what she did by
somebody else. This woman lives in a trailer with her mother. . . . [S]he
was supervis[able], and you can see that.

This account by one Defender typifies the stories we
heard about the human beings who are jailed in the
federal system—their lives, their heartbreaks, and their
ability to come to court and comply with the law, if only a
judge would let them return home to their families.

61

Pretrial detention has myriad pernicious consequences, harming
individuals, as well as their loved ones and communities.101 First,
it prevents people from accessing necessary healthcare, subjects
them to dangerously overcrowded living conditions, exposes them
to physical violence, and increases their vulnerability to infectious
diseases. Second, people who are jailed pending trial in the federal
system are more likely to be convicted, sentenced to longer terms of
incarceration, and sentenced pursuant to a mandatory minimum
than their released peers. Finally, pretrial detainees experience
employment and housing instability at a higher rate than their
released peers, and they can lose custody of their children after even a
few days in jail.
The harmful effects of pretrial detention cannot be justified as
permissible consequences of protecting the community, since
research shows that pretrial detention—for any amount of time—is
correlated with an increase in recidivism.102 In fact, “researchers found
that pretrial detention is associated with a ‘consistent and statistically significant increase’ in the likelihood of new arrest pending
trial,” although “people who are never detained pretrial are no less
likely to appear for court.”103 Another study found that “defendants
who are detained before trial are over ten percentage points more
likely to be rearrested for a new crime up to two years after the initial
arrest.”104 Far from mitigating crime, “detaining a person pretrial can
do more harm than good and have cascading negative consequences
on a community’s safety.”105 Although these studies focus on state
systems, their results are generalizable to the federal context—especially considering the state system’s significantly higher proportion
of overall violent offenders.106
By contrast, pretrial release has been shown to improve case
outcomes and mitigate the deleterious effects of facing federal
criminal charges. Research demonstrates that pretrial release reduces
future crime in two ways. First, it avoids the criminogenic effects of
pretrial detention (including harsh prison conditions and negative
peer effects). Second, release increases the likelihood of current and
future employment, which in turn supports economic stability and
discourages further criminal activity.107

62

A

Federal Pretrial Detention Exposes
Presumptively Innocent People to
Dangerous Conditions.
The presumption of innocence that applies to anyone facing trial
prohibits punishment prior to conviction. Nevertheless, courts
have ruled that jailing someone prior to trial does not constitute
“punishment.”108 However, when conditions are so harsh that “jail
officials place detainees in situations that pose specific threats to
their life or health,” some circuits have deemed pretrial detention to
be punishment—and thus constitutionally impermissible.109 Recent
evidence shows that the federal system has come dangerously close
that line, and probably even crossed it.
The brutal reality is that people detained pretrial in federal jails
are deprived of necessary medical care, live in dangerously overcrowded conditions, are subjected to violence, and suffer high
rates of COVID-19 infection.
First, individuals detained pending trial have trouble accessing
basic healthcare. One study found that approximately one-fifth of
all people suffering from medical conditions were unable to see a
doctor while in federal or state jail.110 Relatedly, pretrial detainees
are often unable to acquire essential prescription medication.111 This
same study found that one-third of people with chronic conditions
were not taking their prescription medication in jail, most commonly
because the jail’s doctor did not find the medication necessary or the
facility would not provide it.112
Second, rampant overcrowding in federal jails exacerbates medical
neglect while introducing incarcerated people to a litany of other
risks. Nearly half of all federal jails have experienced overcrowding
in the last decade.113 Some, like the Metropolitan Correctional Center
in downtown Manhattan (MCC Manhattan), have held twice their
stated maximum number of detainees.114 There, jail staff take up to
two months to respond to a sick call request.115 Overcrowding and
understaffing at MCC Manhattan and the Metropolitan Detention
Center in Brooklyn (MDC Brooklyn) has led to the two jails sharing
“a single psychiatrist and . . . a handful of psychologists” to “treat the
63

nearly 500 inmates suffering from significant psychiatric illnesses.”116
Overcrowding poses special risks for individuals who are detained
in older federal jails. For example, in the brutally hot months of July
and August, people detained in the United States Penitentiary in
Atlanta (USP Atlanta)—which opened in 1902117—were “held in an
overcrowded pod (unit) with approximately 100 other people,” often
without “air conditioning or airflow.”118
Mass detention has also led to hazardous understaffing issues in
federal jails, which exacerbate the effects of overcrowding and further
impede adequate access to medical care. For example, in the Miami
Federal Detention Center, understaffing led to detainees who were
“diabetics, hypertensives, [and suffered from] cardiomyopathy and
HIV . . . not being provided their medication. It was reported that over
750 prescriptions [went] unfilled.”119
A recent investigation of abuse, corruption, and misconduct at the
USP Atlanta federal jail reveals that people jailed before trial in the
federal system are often subjected to horrific living conditions.120
In testimony before the Senate, Assistant Federal Public Defender
Rebecca Shepard described an environment where pretrial detainees
are deprived of clean bathing and drinking water, go months without
hot meals, suffer emaciation from poor nutrition, go weeks without
clean clothes, experience months-long lock downs, and are denied
access to medication and mental health professionals.121 Conditions
like these are replicated at BOP pretrial detention facilities
across the country.122
These appalling conditions also deprive people of their constitutional right to the effective assistance of counsel. Attorneys at USP
Atlanta and other poorly managed federal jails struggle to speak to
or exchange mail with their clients.123 Lack of access to counsel in
turn causes court dates to be continued, drives up the time that each
person spends behind bars, and increases the cost to taxpayers.124
In the brief periods when people awaiting trial are released from
their cells during lock downs at USP Atlanta, they must make the
“cruel but routine” choice of “whether to shower, speak to a loved one
or lawyer, or study their legal case, before returning to a cramped,
infested, and often too-hot cell.”125
Third, people incarcerated in federal and state jails pretrial are
often exposed to violence and physical threats. In 2018, approximately 28,000 individuals locked in federal and state jails and
prisons reported sexual victimization while in custody.126 For every
64

1,000 individuals held in jails and prisons, nearly 13 alleged an act of
sexual victimization.127 The actual number of assaults is likely much
higher, as staff members perpetrated 56% of these reported assaults.128
These numbers represent an increase of more than 300% from 2011129
and an increase of more than 200% in sexual violence at the hands
of correctional staff.130 Aside from the threats an individual may
encounter directly, witnessed violence can also result in significant
harm, including post-traumatic stress, anxiety, depression,
avoidance, hypersensitivity, hypervigilance, suicidal thoughts,
flashbacks, and challenges with emotional regulation.131
The effects of experienced and witnessed violence exacerbate preexisting mental health conditions, which are more common among
those in prisons and jails than among the general public.132 One
former prisoner—who spent 13 years in a federal prison—described
the jail where he was held before trial as “by far the worst place”
to face incarceration.133
Fourth, the ongoing COVID-19 pandemic compounds the health and
safety crisis in federal jails. Mass incarceration—and mass pretrial
detention—have created the “perfect breeding ground” for the virus.134
The virus transmits most easily in confined spaces; “perhaps no
space contains a fixed population less capable of dispersing than
a detention facility.”135
In 2020, the COVID-19 infection rate was over 5 times higher for the
incarcerated population than for the general U.S. population.136 Early
in the pandemic, federal jails quickly became hotspots for the virus.
In May 2020, cases exploded at the Metropolitan Correctional Center
in Chicago’s federal jail (MCC Chicago), with over one-sixth of the
people detained there contracting the coronavirus.137 Just downstate
from MCC Chicago, hospitals were overwhelmed by infected individuals needing emergency treatment.138 In other federal jails,
COVID-19 spread even more rapidly. At the San Diego Metropolitan
Correctional Center (MCC San Diego), for example, over 70% of the
550 detainees were infected in a single month.139
As the pandemic progressed, each new variant renewed concerns
about the health and safety of people jailed before trial. By
August 2022, the BOP acknowledged that more than one-third of
those in BOP-managed institutions or community care facilities had
been infected by COVID-19 since the start of the pandemic (49,316 of
141,252 people in federal jails and prisons combined).140 Research and
congressional testimony by Professor Alison Guernsey has raised
65

concerns that such BOP data seriously underestimate the number
of infections and deaths.141 For example, the BOP reported that, as
of January 31, 2022, COVID-19 was responsible for 284 deaths in
BOP facilities, but Freedom of Information Act requests revealed
numerous additional unreported deaths.142
BOP mismanagement during the COVID-19 pandemic extended
beyond outbreaks and underreporting. For example, at the
Seattle-Tacoma Federal Detention Center (FDC SeaTac), detainees
reported a complete month-long lockdown because of a policy of
FDC SeaTac’s which imposed an automatic two-week lockdown
for the entire jail anytime a staff member or incarcerated person
caught COVID-19.143 For an entire month, detained individuals were
“restricted to their cells for 22 to 24 hours a day, without access to
education, recreation, or communication with family.”144
Amid a new outbreak of the monkeypox virus across the country,
medical professionals have continued to raise concerns that federal
jails are hotspots for the transmission of dangerous diseases. The
monkeypox virus spreads “through close, personal, often skin-to-skin
contact,” including “[c]ontact with respiratory secretions.”145 As the
former medical chief of the New York City jail system writes, “jails,
prisons and detention centers are sites of common close physical
contact that also remain purposefully removed from our community
health systems and oversight.”146 Noting the BOP’s mismanagement
of the COVID-19 pandemic, medical professionals have cautioned
“with people in tight spaces like a jail, a lot has to change in order for
people to not give one another monkeypox.”147
The COVID-19 pandemic revealed additional inadequacies
plaguing our federal pretrial detention centers, but it did not
create them. As pretrial detention rates have soared in recent
decades, federal jails have become more crowded, more understaffed,
and more dangerous. The people living in these egregious conditions
are presumed innocent under the law, yet these conditions threaten
their health and impede their ability to succeed at trial, prolonging
the ill effects of their pretrial detention.

66

B

Federal Pretrial Detention Leads to
Longer Sentences.
Jailing a person pending trial has major consequences for the
outcome of their case. As United States District Court Judge Jed
Rakoff has written:
In the majority of criminal cases, a defense lawyer meets her client
only when or shortly after the client is arrested so that, at the outset,
she is at a considerable informational disadvantage to the prosecutor. If, as is very often the case (despite the constitutional prohibition of ‘excessive bail’ and so-called ‘bail reform’ in a few localities),
bail is set so high that the client is detained, the defense lawyer has
only modest opportunities, within the limited visiting hours and
other arduous restrictions imposed by most jails, to interview her
client and find out his version of the facts.148
This restricted access hampers the attorney’s ability to provide
a robust defense and inhibits the development of a strong
client-attorney relationship. In interviews, defenders and judges
alike consistently echoed the idea that the pretrial detention
determination “[is] a really, really significant, pivotal moment in
any given case.”
The consequences of pretrial jailing extend far beyond the initial
period of detention and permeate each subsequent aspect of an
individual’s case, including the ultimate determination of guilt
and any attendant sentence. In regard to case outcome, a judge we
interviewed observed that pretrial detention “creates a tremendous
amount of pressure to plead [guilty] . . . [and] accept a disposition that
will result in a more prompt release because now, as a defendant,
I’ve experienced custody. I don’t like it. It also gives the government
a lever.” Another judge we spoke to recalled instances in which an
arrestee’s only hope of avoiding decades-long mandatory minimum
sentences was to cooperate with prosecutors: “[F]rom time to time,
an AUSA would brazenly say something like, ‘we have no interest in
cooperation [with the arrestee] if you release them [pending trial].’”
Data corroborate these observations. Strikingly, one study of federal
and state arrestees found that “[p]retrial detention increases a
defendant’s likelihood of conviction by 55%,”149 even when controlling
for confounding factors.150 Relatedly, a study in state court found that
67

individuals who are released pretrial are substantially less likely to be
found guilty, to plead guilty, and to be incarcerated at all.151
Should an arrestee be convicted, the pretrial detention decision
directly and adversely affects the length of their sentence, according
to a 2020 study.152 Additionally, the study found a causal relationship
between pretrial release and receiving a below-guidelines sentence
or avoiding a mandatory minimum.153 Pretrial release also increases
the likelihood that an arrestee will receive a sentence reduction for
providing assistance to prosecutors.154
As one Federal Defender we interviewed said:
[When someone] is on bond and doing well . . . it’s likely for them to
get a lower sentence when it comes time to sentence. Because you at
least have set them up where . . . they’re with family members, they’re
not getting into trouble . . . and so, when it comes time to make those
mitigation arguments, you have some concrete things—in addition
to . . . . putting their lives in context—also showing what they’ve done
while they’ve been on bond.
Another Defender observed that, at the sentencing stage, “it’s
very hard for a judge to put somebody in [prison] who has been
doing so well on the outside,” as the arrestee has shown that they
can be successful at home. Pretrial release also provides a judge
with humanizing context about the arrestee. In these instances,
a judge can “see the [arrestee’s] family, [and] they get to see [the
defense attorney] interacting with the [arrestee] without any sort of
handcuffs or anything.”
In contrast, when someone who has been detained pending trial
appears for sentencing, the judge has only ever seen them in a jail
uniform and has never witnessed the person living successfully
at home. Pretrial detention deprives a person of the opportunity
to show the sentencing judge “that they can (and more likely will)
live by society’s rules.”155 At the sentencing hearing, one Defender
noted that an arrestee who has been jailed pretrial can only make
“arguments about what they will do when they get out, this is the job
you hope they’ll have, and things like that, versus actually being able
to show and demonstrate what they’re actively doing” if they instead
had been released.
In short, the decision whether to put an arrestee in jail pretrial not
only influences the entire trajectory of that person’s case, but also
determines whether they will spend many more years behind bars
after their case is over.
68

The decision whether to put
an arrestee in jail pretrial
not only influences the
entire trajectory of that
person’s case, but also
determines whether they
will spend many more years
behind bars after their case
is over.

69

C

Even Short Periods of Pretrial
Detention Have Long-Term Negative
Personal Consequences.
Beyond leading to lengthier sentences, pretrial detention negatively
impacts the lives and well-being of detained individuals, their
families, and their communities in many long-lasting ways.
People locked in jail pretrial are more likely to lose their jobs and
incomes, even if they are detained for just a few days before ultimately being released. Because federal pretrial detention disproportionally affects poor people of color,156 the collateral consequences of
detention further aggravate this disparity along socioeconomic and
racial lines. One state study found that 3 to 4 years after pretrial bond
hearings, individuals who had been released were 25% more likely to
be employed than their detained counterparts.157
Even when people experience only brief periods of pretrial detention,
their long-term earnings are significantly lower than those of people
who are never detained.158 As one state study explains, “[T]ypically,
those who are working when given even a short jail sentence are in
low-wage positions and are easily replaceable.”159 Individuals detained
for 3 or fewer days reported employment disruption 17% of the time,
yet nearly 60% of those detained for more than 3 days reported
such disruption.160 In addition to the risk of short-term job loss, a
person jailed pending trial could be at risk of experiencing long-term
unemployment, particularly if they are “in their peak wage-earning
years (20–40)” or face “disruption of education and job training.”161
This finding about the impact of short-term jailing is especially
salient for the federal system, where our study finds that the vast
majority of arrestees (77%) are detained at the Initial Appearance and
jailed for several days until a Detention Hearing.162

70

Moreover, the effects of time spent in jail and subsequent periods of
lost earnings compound over the course of an individual’s lifetime,
fueling intergenerational poverty. As one study reports, “formerly
imprisoned people earn nearly half a million dollars less over their
careers than they might have otherwise.”163 This study further found
that such economic losses are more acute among Black and Latino
individuals, worsening economic disparities across racial lines.164
Employment instability, compounded by other factors, can also
cause housing instability when detained individuals are later
released. People who are detained pretrial often have housing
instability to begin with, and any time spent in jail can worsen the
situation.165 “[A]n arrest record and time in jail can result in denial
from a landlord or the inability to stay with family members who
live in public housing where living with a person with a criminal
record is banned.”166 Moreover, individuals who have experienced
pretrial detention are less likely to “hold a lease or mortgage after
release compared to their pre-incarceration status and are more
likely to experience homelessness after release from jail, even when
charges were dismissed.”167
Pretrial jailing fuels a pernicious cycle of intergenerational incarceration.168 Pretrial detention has devastating and lasting effects on
the lives of detainees’ children. When a parent is detained for even
a short time, they can lose custody and their child can be placed in
the foster care system.169 Once a child is in foster care, a detained
parent is less likely to regain custody than their released counterparts.170 These harmful consequences become more acute if a person
is convicted and faces incarceration post-conviction—an outcome
that is statistically more probable if the arrestee is detained pending
trial rather than released. A study of the effects of federal prisons
found “children of incarcerated parent[s] . . . exhibit more negative
behavioral, academic, and emotional outcomes, and are more likely
than their peers to end up in prison.”171

71

Today’s system of mass pretrial detention places significant
burdens on individuals, families, and communities, yet the data
show that these costs provide little societal benefit. Bail reform
efforts undertaken by a number of cities—including New York,
Houston, and Los Angeles—provide ample evidence that reducing
pretrial detention does not increase rearrest rates, including for
serious or violent crimes, and saves taxpayers millions of dollars.
In New York, for example, crime rates for various categories of
offenses decreased after bail reform—including rates of grand larceny,
robbery, and rape172—and the state is projected to save an estimated
$638 million on incarceration costs.173 One study analyzing New
York’s bail reform concluded that “80,000 people may have avoided
jail incarceration due to cash bail because of the 2019–20 reforms
and went on to pose no documented threat to public safety.”174 Studies
in other places, such as Harris County, Texas, found comparable
results: bail reform reduced pretrial detention “without adversely
impacting public safety,” and “did not fuel a spike in crime.”175
To begin to remedy the many harms of mass pretrial detention
and support the positive outcomes of pretrial release, federal
judges must shift the culture from one prioritizing pretrial
detention to one prioritizing pretrial release.

72

4

Legal Standards at
Federal Bond Hearings
The high federal pretrial detention rate is not what
Congress intended. The BRA reflects Congress’s desire
to detain a “small but identifiable group of particularly dangerous defendants.”176 It was only for this
“limited group of offenders that the courts [needed
the] . . . power to deny release pending trial.”177 Congress
considered at length arrestees’ pretrial liberty interests,
and concluded that the constitutional concerns
with pretrial detention required a narrowly tailored
statute to secure community safety and appearance
in court.178 The BRA, then, codified Congress’s desire
to detain just a fraction of the people charged with
federal crimes. In upholding the Act as constitutional, the Supreme Court confirmed that the law was
intended to “operate[] only on individuals arrested for
a specific category of extremely serious offenses.”179
The Act therefore “carefully limits the circumstances
under which detention may be sought to the most
serious of crimes.”180
To balance a person’s pretrial liberty interest with the
prosecution’s interests in preventing danger and flight,
the BRA sets substantive limits on pretrial jailing in
addition to its “exacting . . . procedural protections.”181
These substantive limitations apply at the two sequential
bond hearings that take place in federal court: the Initial
Appearance hearing and the Detention Hearing.
73

At a person’s Initial Appearance—the first court appearance in a
federal criminal case—the law provides a presumption of release.182
If, and only if, specific enumerated criteria are met at the Initial
Appearance, a judge may proceed to a Detention Hearing and detain
the arrestee in the interim.183
At the Detention Hearing, a presumption of release continues to
apply for most arrestees; a judge may jail a person only if they find
that no conditions of release will “reasonably assure” the person’s
appearance in court and the safety of the community.184 At both
hearings, the BRA imposes yet another substantive limitation,
prohibiting judges from imposing any financial requirement
that the arrestee is too poor to satisfy.185 Other statutes provide
an additional procedural protection for indigent individuals: the
right to be represented by appointed counsel for the entirety of the
Initial Appearance hearing.

74

A

The Initial Appearance Hearing
The first substantive limit imposed by the BRA applies at the Initial
Appearance hearing and is contained in § 3142(f). That provision
defines the narrow category of offenses that are eligible for pretrial
detention, instructing: “The [judge] shall hold a [detention] hearing”
only “in a case that involves” one of 7 factors. Under this law, a judge
may jail an arrestee at the Initial Appearance only if one of seven
preconditions to jailing in § 3142(f) is satisfied—we call these the
“(f) factors.”186 These 7 (f) factors serve as gatekeepers to detention; if a
judge detains someone without an applicable (f) factor, the resulting
detention is flatly illegal.187 If no (f) factor applies, a judge is not even
allowed to hold a Detention Hearing. Instead, they must release the
arrestee at the Initial Appearance.188
The factors in § 3142(f) create a federal “detention eligibility net”—
the set of offenses and factors for which Congress has authorized
pretrial detention.189 Section 3142(f)(1) defines the scope of the BRA’s
offense-specific factors, circumscribing the offense-specific bases for
holding a Detention Hearing. Section (f)(1) authorizes a judge to hold
a Detention Hearing, upon the prosecutor’s request, for crimes of
violence; most drug offenses; offenses involving guns, a minor victim,
or terrorism; offenses with a maximum penalty of life in prison or
death; and certain instances of recidivism.190 Section 3142(f)(2) adds
two additional bases for detention that we categorize as non-offense-specific risk factors: a judge is authorized to hold a Detention
Hearing if there is a “serious risk that [the] person will flee,”191 or a
“serious risk” that the person will obstruct justice.192
Many judges and practitioners mistakenly refer to the types of
offenses within the § 3142(f)(1) eligibility net as “presumption cases.”
This is a dangerous misnomer that fuels the culture of detention.
The BRA does not prescribe any presumption of detention at the Initial
Appearance. The lone offense-specific presumption of detention in
the entire statute applies only during the Detention Hearing, not
the Initial Appearance.193 In fact, a number of the offenses listed in
§ 3142(f)(1) do not qualify for a presumption of detention during the
Detention Hearing, including crimes of violence and gun possession
offenses.194 It is therefore patently wrong to bundle together all of the
§ 3142(f)(1) offenses under the umbrella term “presumption cases.”
75

B

Right to Counsel at the
Initial Appearance Hearing
Federal law requires judges to ensure that every arrestee is represented by counsel during the Initial Appearance, and to appoint
counsel to represent poor arrestees during that hearing. Specifically,
18 U.S.C. § 3006A requires that “[a] person for whom counsel is
appointed shall be represented at every stage of the proceedings from
his initial appearance before the United States magistrate judge or
the court through appeal.”195 Federal Rule of Criminal Procedure 44
requires the same: “A defendant who is unable to obtain counsel
is entitled to have counsel appointed to represent the defendant
at every stage of the proceeding from the initial appearance
through appeal.”
To comply with these laws, courts must ensure that, at every single
Initial Appearance, a lawyer is present and actively representing the
arrestee throughout that hearing. It is not enough to appoint a lawyer
who is not present at the hearing or to have a lawyer in the courtroom
who is not representing the arrestee. The legislative history of
§ 3006A plainly states that the law “requires . . . appointment of
counsel for any person under arrest,” and that this representation
“necessarily precedes the stage of formal appointment of counsel” at
the Initial Appearance.196 This early and meaningful appointment
of counsel is the only way to “assure that everyone, rich or poor, will
have the opportunity to utilize the services of an attorney as early in
the proceedings as is possible.”197 In addition, every arrestee must be
represented by counsel during the Initial Appearance to ensure that
they are not detained in violation of the substantive requirements
of § 3142(f).

76

C

The Detention Hearing
At the Detention Hearing, additional substantive and procedural
protections apply. When an arrestee is detained at the Initial
Appearance, the statute states that “[t]he [detention] hearing shall
be held immediately upon the person’s first appearance . . . unless
that person, or the attorney for the Government, seeks a continuance.”198 A judge may not order a continuance of their own accord.
When the prosecution seeks a continuance, that continuance is
limited to 3 days, barring good cause.199 When the individual seeks a
continuance, however, the limit is 5 days, again barring good cause.200
As at the Initial Appearance, there is a presumption of release at
the Detention Hearing, except in cases covered by § 3142(e)(2)–(3),
where a rebuttable presumption of detention applies.201 However, the
presumption of detention does not mandate detention. Instead, at
all Detention Hearings—even when the presumption of detention
applies—the court is prohibited from detaining the accused pretrial
unless it finds that no conditions of release will reasonably assure
the person’s appearance in court and community safety.202
The most common rebuttable presumption of detention is contained
in § 3142(e)(3). Under this provision, an arrestee faces a presumption
of detention at the Detention Hearing in most cases involving drug
offenses, § 924(c) gun violations, offenses against minor victims,
and terrorism.203
An additional rebuttable presumption of detention applies under the
following exceedingly rare circumstances: the accused is (1) charged
with a crime of violence, drugs, guns, an offense against a minor
victim, or terrorism, (2) the accused was previously convicted of any
one of those offenses while on pretrial release, and (3) fewer than
5 years have passed since the date of conviction or release from prison
for that prior offense.204

77

Because the presumption of detention is rebuttable, the court is
required at every Detention Hearing where it applies to determine
whether the evidence the defense has presented is sufficient to
“rebut” that presumption—even if the defense never mentions the
word “rebuttal.” The bar for rebuttal is low: the law requires courts
to find the presumption rebutted as long as the defense presents
“some evidence” relating to the factors listed in § 3142(g).205 Some
courts have allowed judges to consider even a rebutted presumption
as a factor favoring detention.206 However, that approach has no basis
in the statute and may be constitutionally problematic, as it impermissibly relieves prosecutors of their burden of persuasion.207 And
even if the presumption of detention has not been rebutted, courts
may not automatically order detention, but must instead consider
the presumption “together with the factors listed in § 3142(g).”208

78

D

Conditions of Release
When a judge releases an arrestee, the BRA establishes a
presumption that the person will be released on personal recognizance or subject to an unsecured bond.209 However, if a judge
determines that such conditions will not reasonably assure
appearance and community safety, the judge may impose additional
conditions of release to mitigate those concerns.210 In that case,
the judge must impose the “least restrictive” condition(s) that will
reasonably assure appearance and safety.211
Section 3142(c)(1) contains a list of conditions that a judge is required
to consider before detaining an arrestee pretrial. The first condition
is commonly referred to as a “third-party custodian”—a judge can
order the arrestee to “remain in the custody of a designated person,
who agrees to assume supervision and to report any violation
of a release condition to the court.”212 Other conditions include
restrictions on travel, employment requirements, and drug treatment.
The statute also allows judges to impose additional conditions
not listed in § 3142(c)(1).213
Although the statute lists certain financial conditions of release, it
has an overarching command: “The judicial officer may not impose a
financial condition that results in the pretrial detention of the person.”214
This prohibition is intended to restrict the use of financial conditions
and to prevent indigent arrestees from being locked in jail because
they do not have the money or the means to pay for their own release.

79

80

FINDINGS &
RECOMMENDATIONS
81

1

Judges Must Follow the
Correct Legal Standard
at the Initial Appearance
Hearing and Stop Jailing
People Unlawfully.

Our study found that there is a severe misalignment
between the legal standard that applies during Initial
Appearance hearings and the practice that unfolds
in courthouses around the country. Judges routinely
ignore the legal standard in § 3142(f) and sometimes
jail people unlawfully.

While courtwatching, we observed a problematic feedback
loop that results in illegal detention. Prosecutors
frequently make improper detention requests that do not
comply with the statute, judges neither question those
requests nor adhere to the statutory requirements, and
some arrestees are jailed unlawfully.215 See Figure 5. The
results of our 2019 Chicago pilot study were strikingly
similar, further illustrating the nationwide scope of
this problem.216 Our qualitative interviews confirmed
that judges across the country mistakenly assume that
prosecutors are entitled to a Detention Hearing in every
case, ignoring explicit statutory language to the contrary.
The unlawful detentions we uncovered contribute to
rising detention rates and disproportionately fall on
people of color.
To address these misapplications of the law, federal
judges can—and must—bring federal bail practices back
in line with the original intent of the BRA.217
82

Our study found that there
is a severe misalignment
between the legal standard
that applies during Initial
Appearance hearings and
the practice that unfolds
in courthouses around
the country.

83

A

Takeaways

• In 81% of Initial Appearances in our study where the prosecutor requested detention, the prosecutor asked the judge to
hold a Detention Hearing without citing any legal basis under
§ 3142(f). In some of these cases, prosecutors cited invalid bases for
requesting a Detention Hearing, such as danger to the community
or non-serious risk of flight. See Figure 9.

81%

• In over 99% of Initial Appearances where the prosecutor requested
detention without citing a valid basis under § 3142(f), judges
detained people without questioning prosecutors’ grounds for
detention. See Figure 9. This created a problematic feedback
loop in which a prosecutor’s request for detention at the Initial
Appearance almost always resulted in a judicial order of
detention, even when based on improper grounds. See Figure 5.

• In 12% of Initial Appearances where the prosecutor
was seeking detention, judges entered a detention
order even though no statutory basis for detention
existed under § 3142(f). These detention orders,
therefore, were flatly illegal under the BRA.
See Figure 9.

99%
84

99%

12%

p

Racial Disparities
at the Initial Appearance
White people
People of color

White people
Black and Latino

21%
44%

• Prosecutors sought detention in cases that did
not qualify for a Detention Hearing under § 3142(f)
(1) more than twice as frequently for non-white
arrestees. See Figure 10.

23%
35%

• Prosecutors similarly cited improper grounds for
detention more frequently against Black and Latino
arrestees. See Figure 11.

Citizenship Status Disparities
at the Initial Appearance
Citizens

68%

Noncitizens

85%
79%
78%

• Prosecutors requested detention nearly 20% more frequently
if the arrestee was identified as a noncitizen. At 79% of these
hearings, prosecutors failed to cite a valid basis for detention
under § 3142(f). See Figure 15.
• In noncitizen cases, no judge questioned the prosecutor’s
grounds for detention when they failed to cite a valid basis for
detention at the Initial Appearance, leading judges to detain
arrestees in 78% of these cases.

85

THE STATUTE: 18 U.S.C. § 3142(f)
(f) Detention Hearing.—The judicial officer shall hold a hearing to determine whether
any condition or combination of conditions set forth in subsection (c) of this section
will reasonably assure the appearance of such person as required and the safety of any
other person and the community—
(1) upon motion of the attorney for the Government, in a case that involves—
(A) a crime of violence, a violation of section 1591, or an offense listed in
section 2332b(g)(5)(B) for which a maximum term of imprisonment of 10 years
or more is prescribed;
(B) an offense for which the maximum sentence is life imprisonment or death;
(C) an offense for which a maximum term of imprisonment of ten years or
more is prescribed in the Controlled Substances Act (21 U.S.C. 801 et seq.), the
Controlled Substances Import and Export Act (21 U.S.C. 951 et seq.), or chapter
705 of title 46;
(D) any felony if such person has been convicted of two or more offenses
described in subparagraphs (A) through (C) of this paragraph, or two or more
State or local offenses that would have been offenses described in subparagraphs (A) through (C) of this paragraph if a circumstance giving rise to
Federal jurisdiction had existed, or a combination of such offenses; or
(E) any felony that is not otherwise a crime of violence that involves a minor
victim or that involves the possession or use of a firearm or destructive device
(as those terms are defined in section 921), or any other dangerous weapon, or
involves a failure to register under section 2250 of title 18, United States Code;
or
(2) upon motion of the attorney for the Government or upon the judicial officer’s own
motion in a case, that involves—
(A) a serious risk that such person will flee; or
(B) a serious risk that such person will obstruct or attempt to obstruct justice,
or threaten, injure, or intimidate, or attempt to threaten, injure, or intimidate,
a prospective witness or juror.

86

B
The Law: The Bail Reform Act Carefully
Limits the Cases Eligible for Detention.
The BRA divides the bail process into two stages, an Initial
Appearance hearing and a Detention Hearing, each of which is
governed by a separate legal standard. In all cases, the statute
contains “a presumption in favor of pretrial release” at the Initial
Appearance.218 It states that a judge “shall order the pretrial release
of [an arrestee] on personal recognizance” at the Initial Appearance
subject to the least restrictive conditions of release that will
“reasonably assure” that person’s appearance in court and the safety
of the community.219 To detain an arrestee at the Initial Appearance
and proceed to a Detention Hearing, a judge must find certain predicates satisfied at the Initial Appearance itself.
Although the Initial Appearance is the first court hearing in a federal
criminal case, it does not come first in the statute. Instead, the
legal standard for Initial Appearances is buried in the middle of the
statute, in § 3142(f).
There is a widespread misperception that prosecutors are entitled to
a Detention Hearing every time they request one. However, under the
BRA, the prosecutor may move for detention at the Initial Appearance
only if authorized by one of the factors in § 3142(f) (the “(f) factors”).
The BRA says that “the judicial officer shall hold a [detention]
hearing” only “in a case that involves” one of the 7 (f) factors.220
“If none of the § 3142(f) factors are satisfied, however, the [judge]
is prohibited from holding a detention hearing or detaining the
defendant pending trial.”221 If no (f) factor applies, the arrestee must,
as a matter of law, be released at the Initial Appearance. In this sense,
§ 3142(f) “serve[s] as a gatekeeper to [pretrial] detention.”222
The Supreme Court reaffirmed this limitation in Salerno, saying,
“detention hearings [are only] available if [the] case involves crimes
of violence, offenses for which the sentence is life imprisonment
or death, serious drug offenders, or certain repeat offenders.”223 A
central reason the Supreme Court upheld the BRA as constitutional
in Salerno was the gatekeeping function of the (f) factors.224 The
Court opined in pertinent part, “the Bail Reform Act carefully limits
the circumstances under which detention may be sought to the most
serious of crimes.”225
87

➔

Every federal court of appeals that has examined this issue agrees
that one of the § 3142(f) factors must be proven at the Initial
Appearance in order to hold a Detention Hearing.226 As the Second
Circuit recently held, before a federal judge can detain an arrestee at
the Initial Appearance, “the Government must establish by a preponderance of the evidence that it is entitled to a detention hearing” by
presenting a valid (f) factor.227 The First Circuit agrees: “Congress did
not intend to authorize preventative detention unless the judicial
officer first finds that one of the § 3142(f) conditions for holding a
detention hearing exists.”228 The U.S. Attorney’s Manual that guided
federal prosecutors for decades concurred: “Section 3142(f) does not
authorize a detention hearing in the absence of one of the six situations set forth [in the statute].”229
There are two broad categories of (f) factors: the 5 offense-specific
factors in § 3142(f)(1) and the two non-offense-specific risk
factors in § 3142(f)(2).230

Offense-Specific Factors
In § 3142(f)(1), the BRA identifies 5 categories of offenses that
authorize a judge to hold a Detention Hearing, if and only if the
prosecutor requests one: crimes of violence; most drug offenses;
offenses involving guns, a minor victim, or terrorism; offenses with
a maximum penalty of life in prison or death; and certain instances
of recidivism.231 The statute does not authorize the judge to detain
an arrestee or hold a Detention Hearing without the prosecutor
moving for one.232 If an (f)(1) factor applies and a prosecutor requests
pretrial detention, the judge is required to hold a Detention Hearing.
The statute further requires the judge to detain the arrestee between
the Initial Appearance and Detention Hearing if a party’s request
for continuance is granted.233
Consequently, the prosecutor’s decision to charge a § 3142(f)(1) offense
and seek detention ties the judge’s hands at this stage, effectively
mandating detention during any continuance between the Initial
Appearance and the Detention Hearing.234 In an analogous context,
a growing body of research in the federal system shows that prosecutorial charging decisions create disparities—including racial
disparities—at sentencing.235

88

Automatic detention under § 3142(f)(1) arose during the 1980s War
on Drugs, motivated by concerns about rising crime rates.236 Despite
significant evidence to the contrary, key decision-makers attributed
this uptick to individuals on pretrial release.237 Since Congress passed
the BRA, people of color have most felt the effects of automatic jailing
at the Initial Appearance under § 3142(f). For example, in 2019, over
75% of people who qualified for automatic detention at the Initial
Appearance in cases where the (f)(1) factors applied were people of
color.238 See Figure 7.

Figure 7: People of Color Disproportionately Face Charges that
Expose them to Jailing at the Initial Appearance Under § 3142(f).

0%

20%

40%

60%

80%

100%

National
Level

+@ii+-

I
Other I Unknown

Federal Bailwatching
Project

White

Additionally, evidence shows that offense type is not a good proxy for
a person’s risk of recidivism or risk of flight.239 One government study
concluded that § 3142(f) “was a poorly defined attempt to identify
high risk defendants based primarily on their charge, relying on the
belief that a defendant’s charge was a good proxy for that defendant’s
risk.”240 Another study found that although “[a]ssuming that offense
seriousness correlates to flight risk has intuitive appeal, . . . decades
of bail studies challenge that claim.”241 In fact, “defendants charged
with more serious offenses . . . do not, in fact, fail to appear
at higher rates.”242

Non-Offense-Specific Risk Factors
The second category of (f) factors requires a judge to make a subjective
determination about whether an arrestee poses a serious risk. If no
offense-specific § 3142(f)(1) factor is present, the prosecutor may move
89

Since Congress passed
the BRA, people of color
have most felt the effects
of automatic jailing at
the Initial Appearance
under § 3142(f).

90

for a Detention Hearing only on one of the grounds in § 3142(f)(2).
The prosecutor must establish by at least a preponderance of the
evidence243 that the arrestee is either a serious risk of flight or poses a
serious risk of obstructing justice.244
It is critical to understand that neither these non-offense-specific
risk factors nor the offense-specific factors allow a judge to proceed
to a Detention Hearing based on (1) danger to the community, or (2) a
non-serious risk of flight.
First, although at the Initial Appearance the prosecutor frequently
asks the judge to detain the arrestee as a “danger to the community,”
the courts of appeals have stated unambiguously that when none
of the (f) factors are met, “pretrial detention solely on the ground of
dangerousness to another person or to the community is not authorized.”245 As the Ninth Circuit has held, “We are not persuaded that
the [BRA] authorizes pretrial detention without bail based solely on a
finding of dangerousness. This interpretation of the Act would render
meaningless 18 U.S.C. § 3142(f)(1) and (2).”246 “[I]f none of the offensespecific § 3142(f)(1) factors apply, a judge is prohibited from relying
on “danger to the community—including financial danger—as a
basis for detaining a defendant at the Initial Appearance or holding a
detention hearing.”247 Release on conditions is mandatory.248
Second, although prosecutors often improperly seek detention at
the Initial Appearance on the bare assertion that the arrestee poses
a “risk of flight,” ordinary risk of flight is similarly an impermissible
basis on which to order detention at the Initial Appearance or
schedule a Detention Hearing.249 Instead, the statute requires prosecutors to establish by at least a preponderance of the evidence that
the arrestee poses a serious risk of flight.250 This means that “the
prosecutor must proffer some evidence to demonstrate that the
case actually ‘involves’ a risk of flight that is serious rather than the
baseline risk posed in any federal criminal case.”251 As one court has
stated, an arrestee may be jailed “only if the record supports a finding
that he presents a serious risk of flight.”252 The BRA’s legislative
history makes clear that detention for serious risk of flight should
occur only in “extreme and unusual circumstances.”253
A Detention Hearing is authorized only in a narrow subset of
“non-(f)(1) cases”—those cases that do not involve an offense-specific
factor in § 3142(f)(1). In non-(f)(1) cases, the court may detain the
arrestee at the Initial Appearance and hold a Detention Hearing only
if one of the § 3142(f)(2) risk factors is present. Non-(f)(1) cases include
91

➔

all fraud and financial crimes, money laundering, postal theft, bank
theft, alien smuggling, illegal reentry, and straw purchaser offenses,254
as well as offenses that a court of appeals has declassified from the
category of crimes of violence.255 Non-(f)(1) cases comprise nearly
60% of all federal pretrial cases.256 See Figure 8. Accordingly, in the
majority of cases nationwide, there is a real question of whether
the law allows a judge to hold a Detention Hearing at all.

Figure 8: Non-(f)(1) Cases Where Judges Are Not Automatically Authorized to
Hold a Detention Hearing Comprise Nearly 60% of All Federal Pretrial Cases.

0%

20%

40%

60%

80%

100%

Immigration I Fraud I Public order I Property (non-fraud) I Violent I Weapons IDrug

In non-(f)(1) cases alleging a financial crime, the law prohibits the
court from detaining someone at the Initial Appearance or holding
a Detention Hearing on the basis of “financial danger.”257 Yet despite
this prohibition, judges detain 37% of those charged with fraud
offenses at the Initial Appearance.258 One judge we interviewed
candidly admitted to holding a Detention Hearing in a fraud case on
this impermissible basis: “In fact, I did make that mistake, and [the
defense] filed a motion after I detained someone in a fraud case based
in part on dangerousness. They moved to reconsider, and ultimately
the individual was released.”

92

C

The Practice: At the Initial Appearance
Hearing, Federal Judges Jail
People Unlawfully.

D-

Our study revealed a nationwide problem in the federal courts:
judges do not understand the legal standard that applies at the
Initial Appearance and at times detain people unlawfully. In
practice, the application of the § 3142(f) factors departs from the
statutory strictures and from appellate courts’ guidance, frequently
resulting in unlawful detention.
Our interviews demonstrated that judges throughout the country
mistakenly believe that prosecutors are entitled to a Detention
Hearing whenever they request one—a position that disregards
both the statute and well-established appellate case law. Most of
the judges we interviewed either admitted to not knowing the legal
rules that apply at the Initial Appearance or evinced a misunderstanding of the legal standards. More than half of the Chief Federal
Defenders we interviewed likewise expressed confusion about the
legal standard that applies at the Initial Appearance. Over and over,
judges and attorneys alike were mystified when we began asking
questions about the § 3142(f) requirements, and some expressed
surprise at the basic idea that there is, in fact, a legal standard that
applies during the Initial Appearance.
We watched many Initial Appearances in which judges promptly
granted prosecutors’ requests for detention, jailed the arrestee, and
scheduled a Detention Hearing for a later date, all without first
determining whether the statute authorized a Detention Hearing. We
saw judges, prosecutors, and even defense attorneys proceed through
Initial Appearances in an extremely cursory, rote fashion:

93

JUDGE: Government, what is your position on bond?
PROSECUTOR: We request pretrial detention based on risk of flight and
danger to the community.
JUDGE: Okay sounds good to me. The Detention Hearing is scheduled for
next Wednesday, February 2. Anything further?
PROSECUTOR: No.
DEFENSE COUNSEL: No, your honor.259

Unfortunately, the above exchange typifies many of the Initial
Appearance colloquies we witnessed. Often, there was no meaningful consideration of the legal standards that apply at the Initial
Appearance, no recognition of the statutory presumption of release,
and no effort to safeguard arrestees’ pretrial liberty interests. Instead,
as in this colloquy, prosecutors requested detention for reasons not
authorized by the statute and judges automatically detained people
at the Initial Appearance whenever the prosecutor requested it. As
in this example, judges detained in a knee-jerk fashion without first
ascertaining whether there was a statutory basis for (1) holding a
Detention Hearing at all, and (2) detaining the arrestee for several
days until that Detention Hearing.
Nationwide, there is a 78% detention rate at the Initial
Appearance stage.260 In our study, 77% of arrestees were detained at
that initial stage.

Prosecutors Regularly Request Detention at the Initial
Appearance on Improper Grounds.
Our study revealed a pattern of prosecutors and judges ignoring and
misapplying the legal standard at the Initial Appearance—a pattern
that sometimes resulted in unlawful detentions.
In 81% of the Initial Appearances where the prosecutor sought
detention, they failed to cite a valid statutory basis for detention.
See Figure 9. In 38% of cases, the prosecutor cited statutorily invalid

94

bases, including “danger to the community,” non-serious “risk of
flight,” and an arrestee’s status as a non-citizen.261 See id. Our data
showed that there was, in fact, a legitimate, offense-specific basis for
detention under § 3142(f)(1) in many of these cases.

D--

Figure 9: At the Initial Appearance, Prosecutors
Regularly Request Detention on Improper
Grounds and Judges Detain People Unlawfully.

Cases where the prosecutor
sought detention (233/343)

In 81% of these cases, the prosectutor
failed to cite a valid statutory basis for
detention (189/233).
Judges detained arrestees in over
99% of Initial Appearances where the
prosecutor failed to cite a valid basis
for detention.

12% of these detentions were illegal
under the BRA as no statutory basis for
detention existed under§ 3142(f) (27/233) .
81% of these individuals were detained for
the duration of the case (22/27).

•••
••••••••••

••••••••••
••••••••••
•••••••••••
••••••••••
•••••••••••
•••••••••
••••••••••
••••••••••
•••••••••••
•••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
•••••••
••••••••••
••••••••••

99%

Detained for the whole duration 7

95

However, in our courtwatching, we observed that problematic
feedback loop: when the prosecutor cites improper grounds for
detention, judges rarely question whether there is a proper statutory
basis for holding a Detention Hearing or detaining the arrestee, and
at times they jail people unlawfully. See Figure 5. Our interviews
confirmed the nationwide scope of this problem, with many of the
judges we interviewed admitting that they did not ask prosecutors
to cite a § 3142(f) factor to justify holding a Detention Hearing. Some
judges even acknowledged that it is their practice to schedule a
Detention Hearing whenever the prosecutor requests one.
This feedback loop is both a consequence and a driver of the
culture of detention, with in-court practices diverging from the
legal requirements. One judge, speaking of their experience as a
prosecutor, told us: “I don’t recall the subject of Detention Hearing
eligibility ever really coming up. Instead, it seemed as though the
prosecutors, the defense bar, and the [judges] assumed that if the
prosecutor was moving for detention, it was permitted a continuance
of up to three days for the Detention Hearing to take place, and the
Detention Hearings were scheduled.” One Defender we interviewed
characterized this trend as “sloppiness,” noting that “folks [have]
gotten into bad habits.” Another Defender, concurring in this evaluation, labelled the issue as “complacency culture.”
One judge we interviewed expressed frustration with the idea that
prosecutors must clearly state the basis for detention when seeking
to detain someone at the Initial Appearance. The judge noted that
“it’s kind of superfluous to have to repeat [the basis for detention]”
because the judges know the statute (a perspective belied by our
findings). Another judge told us, more circumspectly, that after
decades on the bench, they felt that “everybody kind of understands
when you say ‘danger,’ ‘risk of flight,’ . . . everybody can understand
what you’re talking about, and it’s just a little—it’s shorthand.” To
their credit, this judge acknowledged that “it’s not right” to use a
shorthand that is inconsistent with the statute.

96

Prosecutors are more likely to request detention at the Initial
Appearance for arrestees of color. In the non-(f)(1) cases we observed,
the prosecution sought detention 43% of the time for people of color,
but only 21% of the time for white arrestees. See Figure 10. Alarmingly,
looking across all Initial Appearances, prosecutors cited an improper
basis for detention 35% of the time for arrestees of color as compared
to 23% of the time for white arrestees. See Figure 11. Arrestees of color
are more likely to be illegally jailed at the Initial Appearance than
their white counterparts.

D-

Figure 10: Prosecutors Sought Detention in
Non-(f)(1) Cases More Than Twice as Frequently Against
People of Color Than Against White Arrestees.

• • • • • • • Black and Latino

·····•• □□ GWhite

·····•• □□□
·····•• □□□

······•• □□
·······•• □
·······•• □
·······•• □

Prosecutors
sought
detention

Figure 11: Prosecutors Are More Likely to Request Detention at
the Initial Appearance on Improper Grounds for People of Color.

000
000
000
00
00
00
00
00
00
00

ooo••·····
••·····
••·····
o••·····

••••••••••••
••••••••••••
••••••••••••
•••••••••••••
•••••••••••••
•••••••••••••
•••••••••••••
•••••••••••••
•••••••
•••••••
•••••••••••••
••••••• •••••••••••••

Prosecutors cited an
improper basis for
detention 23% of the
time for white arrestees,
compared to 35% of the
time for arrestees of
color.

Prosecutors cited improper
basis for detention.

97

Our most troubling finding
was that, in 12% of Initial
Appearances where the
prosecutor was seeking
detention, judges detained
people illegally.

12%
98

Judges Misapply the Law and Illegally Detain Arrestees.

D-

Our most troubling finding was that, in 12% of Initial Appearances
where the prosecutor was seeking detention, judges detained
people illegally. See Figure 9. These were non-(f)(1) cases where, as
a matter of law, there was no offense-specific ground for detention
under § 3142(f)(1). In these non-(f)(1) cases, prosecutors did not cite or
present evidence of any § 3142(f)(2) factor to support their detention
request (the only possible statutory basis for detention). Instead, 97%
of the time the prosecution cited improper grounds, like danger to the
community or ordinary risk of flight.
Despite the prosecution providing no legitimate basis for detention
in the non-(f)(1) cases in our study, every single one of the arrestees in
those cases was detained at the Initial Appearance and subjected to
an improper Detention Hearing. See Figure 12. These detentions at
Initial Appearances—and the Detention Hearings that followed—
were flatly unlawful, in that they were not supported by any showing,
or even an assertion, that the arrestee posed a serious risk of flight or
obstruction of justice.

Figure 12: At the Initial Appearance, Prosecutors’
Improper Detention Requests in Non-(f)(1) Cases
Lead Judges to Detain People Unlawfully.

•••••
••••••
••••••
••••••
••••••

D
Despite the government providing no
legitimate basis for detention in 97%
of non-(f)(l) cases, every single
arrestee was detained and subjected
to an improper Detention Hearing.

Such unlawful detentions are the inevitable consequence of the
Initial Appearance feedback loop we identify, in which prosecutors
seek a Detention Hearing on improper grounds and judges automatically grant that request without checking the statute.

99

Moreover, these unlawful detentions were carried out in a racially
disparate way. Among non-(f)(1) cases, 56% of Black arrestees and
45% of Latino arrestees were jailed on improper grounds at the
Initial Appearance in non-(f)(1) cases, compared to just 21% of white
arrestees. See Figure 13.

Figure 13: Unlawful Detentions Are Carried
Out in a Racially Disparate Way.

Unlawfully detained:
56%
45%

21%

Black arrestees
Latino arrestees
White arrestees

The unlawful detentions we observed are just the tip of the iceberg.
Nationwide, 76% of arrestees in non-(f)(1) cases are detained at the
Initial Appearance262—a strikingly high rate considering that the
only valid ground for detention in such cases requires the prosecutor to prove that the arrestee poses a “serious risk” of flight or
obstruction. See Figure 14. Setting this high national detention rate
alongside our observation that judges detained every single non-(f)(1)
arrestee for whom prosecutors sought detention, it is very likely that
judges elsewhere in the country are likewise jailing people unlawfully
in such cases. Especially considering our qualitative finding that
many judges automatically acquiesce to prosecutors’ requests to jail
people at the Initial Appearance, the non-(f)(1) cases in our study are
truly the canaries in the coal mine.
Given that more than 1 in 10 arrestees in our study were unlawfully
jailed when the prosecutor requested detention at the Initial
Appearance, it is clear that judges are ignoring or misapplying the
law. See Figure 9. This is not just an isolated situation in which a few
judges or attorneys slightly misunderstand the law; rather, there is
a pervasive, systemic deprivation of liberty that is not authorized by
statute or case law.

100

Figure 14: Nationwide, 76% of Arrestees in Non-(f)(1)
Cases Are Detained at the Initial Appearance.

59%of
federal
cases are
non-(f)(l)
cases

D-

76% of arrestees

were detained
at the Initial
Appearance in
non-(f)(l) cases

During our courtwatching, we consistently observed judges
violating the law at Initial Appearances in non-(f)(1) cases. For
example, during an Initial Appearance in a mail fraud case involving
an indigent Black arrestee, the prosecutor cited legally erroneous
bases for detention: “We seek pretrial detention based on risk of flight
and danger to the community.”263 The judge detained the individual
and scheduled a Detention Hearing for 5 days later without objection
by the defense.264 Since this offense is not covered by § 3142(f)(1)
and no § 3142(f)(2) factors were cited, it was unlawful to detain this
arrestee at the Initial Appearance and unlawful to hold a Detention
Hearing in his case. Yet at the Detention Hearing, the judge jailed the
person for the duration of their case—a dire consequence when no
Detention Hearing was legally authorized.265
In a far less typical example, a judge improperly jailed an arrestee
at the Initial Appearance for possessing stolen mail, based that
detention on the invalid grounds of “risk of flight” and “danger to the
community,” but subsequently rectified that error.266 At the Detention
Hearing itself, the judge acknowledged that holding a Detention
Hearing would be statutorily authorized only if the prosecutor had
shown that the arrestee posed a serious risk of flight or specific
threat to a victim or witness, and concluded that the prosecution had
“failed to do so.”267 The judge consequently ordered release.268 But this
man—who was employed in the construction business and had two
children at home269—should never have been jailed in the first place.
Unfortunately, this was the only case involving unlawful detention
at the Initial Appearance where the judge realized and corrected the
error at the Detention Hearing.

101

We observed numerous other instances of illegal detention in
non-(f)(1) cases, including situations in which judges detained people
without prosecutors requesting detention or making any effort
to meet the legal standard for holding a Detention Hearing. We
repeatedly heard judges tell arrestees, “You are entitled to a Detention
Hearing”—a blatant misstatement of the law that erroneously
relieves the prosecution of its burden of establishing a § 3142(f)(2) risk
factor that warrants a Detention Hearing.
Here are additional examples of the unlawful detentions we watched:
• During the Initial Appearance in one non-(f)(1) case in which the
arrestee was solely charged with immigration offenses, the prosecutor and defense counsel stipulated to, and the judge imposed,
a $250,000 corporate surety bond that the arrestee was unable to
meet, resulting in detention for the duration of the case.270
• In a non-(f)(1) money laundering case, everyone in the courtroom
appeared to be operating on the unspoken assumption that a
Detention Hearing would be held, and the defense stipulated to
detention without the prosecutor making any request at all.271
• In a non-(f)(1) case alleging identity theft and conspiracy to commit
bank fraud, the prosecutor requested, and the judge imposed, a
$100,000 personal surety bond.272 Although no one was seeking
detention, the arrestee was jailed until the Detention Hearing
because the arrestee’s lawyer was not present, and the probation
officer was not able to reach his mother.273
• In a fourth non-(f)(1) case—which involved charges of credit
card fraud and identity theft—the judge conducted the Initial
Appearance without defense counsel present and misstated the
law, saying: “You are entitled to a bond or Detention Hearing. . . . [I]
f the government is requesting that you be held without a
bond.”274 The judge then scheduled a Detention Hearing, despite
the prosecutor never requesting one or establishing any legal
basis for detention.
In all 4 of these cases, there was no offense-specific basis for holding
a Detention Hearing under § 3142(f)(1), and the prosecutor failed
to cite to or present the requisite evidence for detention under
§ 3142(f)(2). Thus, each of these 4 cases represents an alarming
instance of a judge jailing someone illegally.

102

Such unlawful detentions occur because judges are not questioning the prosecutors’ grounds for detention. Despite the flagrant
violations of the BRA discussed above, in over 99% of the cases
in which prosecutors failed to mention § 3142(f), judges failed
to challenge prosecutors’ bases for detention. That is, among all
cases in which the prosecution requested detention at the Initial
Appearance, over 99% of arrestees were detained, even when the
prosecution had presented no valid basis for locking the person in
jail or holding a Detention Hearing. See Figure 9. That is a far cry from
the Supreme Court’s admonition that the BRA’s (f) factors “carefully
limit[] the circumstances under which detention may be sought to
the most serious of crimes.”275

D-

Some of these unlawful detentions may be motivated by judges’ fear
that someone they release will reoffend, coupled with an unquestioning reliance on Pretrial Services. When asked what explains
these unlawful detentions, one judge we interviewed said, “I think
judges who don’t want to make either the right or the hard decision
to find release conditions consistent with the Bail Reform Act
rely on Pretrial [Services’] recommendations as a basis to detain.”
The judge continued:
I think it’s laziness. I think it’s that everyone thinks you’re going to
get a Detention Hearing and the basis for it doesn’t much matter
[be]cause the government asks for a Detention Hearing, and there
is one—even if it’s not properly cited—and we’re going to get one
within three days. But, you know, could we do better? Absolutely.
Should we follow the law? Yes, I think so. I just need to be instructed.
Our courtwatching data and stakeholder interviews show that
lack of adherence to § 3142(f) is not an isolated problem in a few
districts but is instead a nearly universal problem that spans
federal courts throughout the country. A shockingly high number
of the judges we interviewed indicated that they considered “danger
to the community” to be a legitimate basis for holding a Detention
Hearing, even though every court of appeals to have considered
the issue has said otherwise. To their credit, one judge called their
district’s failure to question prosecutors’ bases for detention in
non-(f)(1) cases “a weak spot in our analysis,” which had likely led
to many unnecessary and potentially illegal detentions. The judge
went on to explain that, rather than proffering evidence as § 3142(f)
requires, “[prosecutors] would say [they are seeking detention for]
‘serious risk of flight’ and then boom, the person gets detained until
their [detention] hearing.”
103

The defense bar is also implicated in, and influenced by, the culture
of detention. Strikingly, the defense did not object to detention in
any of the cases where arrestees were jailed illegally at the Initial
Appearance. One Chief Defender candidly admitted, “I don’t think
we’ve ever raised that argument [at the Initial Appearance] that you
don’t get to the Detention Hearing at all. I don’t know that we’ve ever
raised that point here [in my district].”
However, our interviews indicate that some defense counsel
who don’t raise § 3142(f) arguments at the Initial Appearance are
responding to pressure from judges. One Defender explained that
when the defense bar objects to detention at the Initial Appearance,
judges say “save that for the Detention Hearing, counsel.” Another
Defender told us that judges typically view any argument at the
Initial Appearance regarding whether an arrestee is a serious risk of
flight “as something for them [to] determine [at] the Detention Hearing”—a position that disregards the requirements of § 3142(f)(2). That
Defender explained that, while lawyers in their office have tried to
argue that judges “shouldn’t even be having the Detention Hearing in
this circumstance,” judges are “very uninterested in that argument.”
Defense counsel waives release in 40% of cases nationally, according
to AO data.276 In our courtwatching study, defense attorneys agreed
to detention277 in 42% of the cases we observed (across Initial
Appearance and Detention Hearings). The rate at which defense
attorneys waived release at any point in the pretrial process varied
among the 4 districts, ranging from 25% to 48%.
While one reason for these high waiver rates is that a certain
percentage of cases involve some kind of detainer (often from a prior
state case or a case in which the arrestee is on probation/parole),
these high waiver rates are also driven by the problematic feedback
loop we have described. When prosecutors request detention at
extremely high rates at the Initial Appearance and judges nearly
universally respond by detaining arrestees during that hearing—even
without a valid basis—it is little wonder that Defenders describe
feeling a sense of hopelessness in the face of the culture of detention.
If judges break that feedback loop by following the statutory rules,
defense counsel may well begin requesting release at higher rates.

104

Judges Regularly Detain Noncitizens at the Initial Appearance
on Improper Grounds.

D-

In the Initial Appearances we watched, judges often detained
noncitizens and individuals with Immigration and Customs
Enforcement (ICE) detainers on improper grounds.278
While the BRA authorizes temporary detention of people with no
legal status,279 pretrial detention of noncitizens during the Initial
Appearance and Detention Hearing must meet the same legal
standards as any other case. Notably, illegal reentry and immigration
offenses are not covered by § 3142(f)(1), meaning that prosecutors
must provide evidence of a serious risk of flight or obstruction to jail
noncitizens at the Initial Appearance. The law is clear that an ICE
detainer, standing alone, is not evidence of a serious risk of flight—
flight must be voluntary.280
In practice however, one Defender described how an ICE detainer
can lead to “automatic detention. . . . [T]he judges’ view is that [an
ICE detainer] is sufficient to detain.” This Defender then told us:
“I’ve brought those [motions] and appealed those to the district court
before, and I’ve never seen anyone get out with an ICE detainer.”
DOJ data show that noncitizens do not pose a higher risk of flight
or violation of release conditions than citizens. Undocumented
individuals have the exact same low rate of non-appearance as
U.S. citizens; they fail to appear just 1% of the time.281 Compared
to U.S. citizens, undocumented arrestees are more likely to comply
with other conditions of release and significantly less likely to have
their bond revoked.282
Despite these realities, in the cases where we courtwatched, prosecutors requested detention for 85% of identified noncitizens at the
Initial Appearance, as compared with just 68% of all arrestees. In 79%
of noncitizen cases in which the prosecutor requested detention, they
either cited an improper basis in seeking a Detention Hearing or did
not provide any grounds for holding a Detention Hearing. Regardless,
when prosecutors failed to cite a valid statutory basis for detention in
these cases, the noncitizen was jailed 78% of the time. See Figure 15.
Moreover, prosecutors’ invalid detention requests resulted in judges
unlawfully detaining 20% of noncitizens at their Initial Appearances.

105

Figure 15: Prosecutors More Frequently Request
Detention for Noncitizen Arrestees, Leading Judges to
Regularly Detain Noncitizens on Improper Grounds.

000
000
JC

•••
•••
•••
•••
•••
•••
•••

•••
•••

Prosecutors requested detention for 85% of
identified noncitizens.
In 79% of these cases, prosecutors either cited an improper
basis in seeking a Detention Hearing or did not provide
any specific grounds for holding a Detention Hearing.
The noncitizen was jailed 78% of the time, resulting in
20% of noncitizens being illegally detained at their
Initial Appearance.

One Defender we interviewed corroborated these findings, noting
that prosecutors request detention more often for people who are
noncitizens, even in non-immigration cases. Similarly, a judge told
us that prosecutors often fail to demonstrate any “level of analysis”
when requesting detention in cases involving noncitizens. Instead,
the judge characterized prosecutors’ thinking as: “there’s a detainer,
we’re requesting detention.”

106

D

The Solution: At the Initial Appearance,
Judges Must Prevent Unlawful Detentions
by Following the § 3142(f) Legal Standard.

*

Judges must take action to ensure that the law is followed
and end the destructive feedback loop that results in unlawful
jailing at the Initial Appearance. One judge told us that, until we
began raising consciousness about the legal standard at the Initial
Appearance, “the (f) factors were not really foremost in most [judges’]
minds.” This needs to change.
First, judges must understand the legal standard that applies
during Initial Appearances and must recognize the strict
limitations the BRA places on the types of cases in which a judge
may hold a Detention Hearing or detain an arrestee at all. In a
case that involves an offense-specific factor in § 3142(f)(1), it violates
the law for a judge to detain an arrestee at the Initial Appearance
or proceed to a Detention Hearing without an explicit request
from the prosecution.
Adherence to the legal standard is equally important in districts
where the Initial Appearance and the Detention Hearing are held
during a single court proceeding—a practice common in highvolume urban districts like New York and Los Angeles. In such
districts, judges must be especially wary about the fact that two very
different legal standards apply at the Initial Appearance and the
Detention Hearing. When a prosecutor requests detention during
a unitary proceeding, judges should scrupulously consider whether
there is any valid legal basis under § 3142(f) for holding a Detention
Hearing before conducting a Detention Hearing.
Second, judges must be highly vigilant in ensuring that federal
prosecutors comport with each element of the BRA’s legal
standard at the Initial Appearance. When a prosecutor requests
detention and a Detention Hearing for reasons that Congress has
deemed inappropriate or otherwise impermissible—like danger or
ordinary flight risk—they are asking the judge to violate the law.
Rather than rubber-stamping this improper invitation to detain,
judges must be on guard, stand as the bulwark against illegal
detention, and deny such requests.
107

Rather than rubber-stamping
this improper invitation to
detain, judges must be on
guard, stand as the bulwark
against illegal detention,
and deny such requests.

108

Judges should ask prosecutors to state on the record what § 3142(f)
factor authorizes a Detention Hearing. If a prosecutor moves for
pretrial detention at the Initial Appearance on the allegation that
an arrestee is a danger to the community and/or an ordinary risk of
flight, the judge must explain that there is no statutory basis for that
request under § 3142(f). Dangerousness and ordinary risk of flight are
not § 3142(f) factors, and it is incumbent on judges to remind prosecutors of that basic statutory fact.

*

Third, judges must be especially careful in the types of cases
where we documented illegal detentions. These are non-(f)(1)
cases where there is no offense-specific factor that authorizes a
judge to hold a Detention Hearing, and where the only possible
basis for detention is “serious risk” of flight or obstruction of justice
under § 3142(f)(2). To determine whether an arrestee is a serious risk
of flight, judges may need to hold a more extensive hearing than
usual during the Initial Appearance. Because the law demands a
showing that any risk of flight is “serious,” the prosecution should
present some evidence to meet its burden, such as evidence about
the particular arrestee’s history and characteristics (e.g., past failures
to appear in court) or to the circumstances of the offense (e.g., the
person led the police in a high-speed chase).283 The defense should be
given an opportunity to present its own evidence to establish that the
arrestee does not pose a serious risk of flight or obstruction. And the
judge must not move forward to a Detention Hearing unless, at the
Initial Appearance, “the record supports a finding that [the arrestee]
presents a serious risk of flight” or obstruction.284
Fourth, judges can use their influence to mitigate the culture of
detention. “Judges should closely scrutinize prosecutors’ requests
for detention” and “ask them to explain why they are moving for
temporary detention, particularly if available information suggests
that the person poses a low risk of flight or danger.”285 This watchdog
function is especially important in cases where the arrestee played a
small role in the overall crime or has little to no criminal history.286 In
such cases, judges can ensure that release is the norm and jailing is
the “carefully-limited exception.”287
By following these recommendations, judges can uphold the
law, interrupt the culture of detention, and protect the liberty
interests of the people who appear before them.

109

2

Judges Must Stop
Unlawfully Jailing
Poor People Without
Lawyers at the Initial
Appearance Hearing.

In many federal courts, judges lock poor people in
jail without a lawyer during their Initial Appearance,
in violation of federal law. Our study uncovered a
national access-to-counsel crisis: judges in more than
one-quarter of the 94 federal district courts do not
provide every arrestee with a lawyer to represent them
during the Initial Appearance. See Figure 6. In fact, 72%
of the districts where we interviewed or surveyed stakeholders deprive at least some individuals of counsel at
this first bail hearing.288 While the scope of the problem
varies across districts and divisions, in every court that
exemplifies this particular crisis, arrestees are jailed
without counsel. These widespread deprivations of
counsel contribute to the culture of detention and drive
high jailing rates at Initial Appearances nationwide.289
State courts’ failures to provide counsel at first
appearance are well documented,290 but our study is the
first to report and catalogue parallel failures in the federal
system. In one district where we courtwatched, we found
that 11% of arrestees did not have a lawyer during their
Initial Appearance. See Figure 16.
110

In many federal courts,
judges lock poor people in
jail without a lawyer during
their Initial Appearance,
in violation of federal law.

111

Figure 16: In Miami, Indigent Arrestees Were Not
Represented by Counsel in 11% of Initial Appearances.

•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
•••••••••••••••••••••••
Additionally, in interviews with and surveys of stakeholders, we
learned that in at least 26 federal districts, judges fail to ensure that
arrestees are represented by counsel at the Initial Appearance in
some—and often many—cases. In certain courts, 100% of arrestees
are deprived of counsel during their Initial Appearance. Our
findings surely understate the scope of this particular crisis, as we
were unable to interview stakeholders in 58 federal districts (62% of
the total districts).
Locking people in jail without counsel violates the U.S. Code and
the Federal Rules of Criminal Procedure, may be unconstitutional,
and is not sound public policy. It is therefore imperative that
federal judges throughout the country stop holding uncounseled
Initial Appearance hearings.
Our data show that these legal failures come with serious consequences. Every uncounseled Initial Appearance we observed ended
in pretrial detention—a far higher detention rate than at the Initial
Appearances where arrestees were represented by counsel. See
Figure 17. And nearly every single arrestee we observed who faced an
Initial Appearance without a lawyer was either Black or Latino, exacerbating the already pronounced racial disparities in the criminal
system. See id. These findings are particularly troubling given that
most people charged with a federal crime do not have the money
to hire their own lawyer and therefore rely wholly upon judges to
appoint counsel at their Initial Appearances.291

112

Figure 17: At the Initial Appearance, Arrestees Deprived of Lawyers
Are Jailed at Far Higher Rates than Those With Lawyers.

Every unrepresented arrestee was
detained at their Initial Appearance.
92% were people of color.
White
30% of arrestees
were partially
represented at
their Initial
Appearance.

-a - - - -

~--------- 1
People of Color

100%
(25/25)

~---------

00000000000
89%of these
arrestees were
detained.

------------------------------------------------------------------------][ ll ,:

60% of arrestees
were represented
throughout their
Initial Appearance.

, : Jr
r

)

IC

l[

r

,

IC

ll ll :

,' ll H
]l ll

IC

,l ll
'1 l[
,l ll
'1 l[
,[ ll
'1 l[

,c

r

'

r

IC

,

,c

r

r

IC,

r

,

,

---------------r

67%ofthese
arrestees were
detained.

)

r

ll H

,, Jr
r

l[

r

:

c

'

,[

][

89%
(92/103)

:

r

'

-------------------------------------------------------------------------------------------------------------

67%
(137/206)

113

Federal judges have a duty to follow the law and uphold constitutional rights. Therefore, it is imperative that judges appoint a
lawyer to actively represent every indigent arrestee at every Initial
Appearance, and throughout that entire hearing.292 Whenever any
discussion related to pretrial detention takes place—including
potentially incriminating financial questioning—the arrestee must
be represented by counsel. This is the only way to fully uphold the
law and protect the liberty interests of the indigent accused.

THE STATUTE: 18 U.S.C. § 3006A(c)(1)
A person for whom counsel is appointed shall be represented at every stage of the
proceedings from his initial appearance before the United States magistrate judge or
the court through appeal, including ancillary matters appropriate to the proceedings.

114

115

A

Takeaways

• Our interviews and survey data revealed that more
than one-quarter of federal district courts fail to
appoint a lawyer for every arrestee at the Initial
Appearance, with at least 26 of the 94 federal districts
exhibiting this problem. See Figure 6.

• In one district where we courtwatched, 11% of
arrestees went unrepresented for the entirety of
their Initial Appearances, with no lawyer by their side
to advocate for their liberty interests. See Figure 16.

• In that district, every single individual who faced their Initial
Appearance without a lawyer was jailed after the hearing, a
100% detention rate. See Figure 17.

• In all 4 districts in our courtwatching study, when
arrestees were forced to proceed without counsel for
some part of their Initial Appearance, there was a
notable increase in pretrial detention: across courtwatched districts, partially represented individuals
were detained 89% of the time, while fully represented individuals were detained 67% of the time.
See Figure 18.

100%
116

28%

11%
100%

89%
67%

p

Racial Disparities
in the failure to provide counsel before
deprivations of liberty

92%

• 92% of the arrestees who were unrepresented at
their Initial Appearances were people of color.
See Figure 17.

• Judges unlawfully detained unrepresented individuals in violation
of § 3142(f) in some of the Initial Appearances we observed,
compounding the harm of not providing a lawyer.
• We also observed Initial Appearances where arrestees made
incriminating statements while judges questioned them
without a lawyer, jeopardizing the person’s Fifth Amendment
right against self-incrimination and their ability to fight the case
in the future.

117

B

The Law: Arrestees Have a Legal Right to
Be Represented by Counsel at
the Initial Appearance.
Federal Law Entitles Indigent Arrestees to Representation by
Counsel at the Initial Appearance.
Every time a judge fails to provide an indigent individual with
a lawyer to represent them throughout the entirety of their
Initial Appearance hearing, she violates federal law. There are
at least two laws that explicitly direct courts to appoint counsel
to stand up on behalf of indigent arrestees during the Initial
Appearance: 18 U.S.C. § 3006A and Rule 44 of the Federal Rules
of Criminal Procedure. Each provision requires judges to provide
appointed counsel “at every stage of the proceeding from initial
appearance through appeal.”293 These laws are explicitly directed
at the judges responsible for appointing counsel, not at the
federal public defender.294
Section 3006A states in mandatory terms that every indigent
arrestee “shall be represented . . . from initial appearance.” The plain
text of this law requires several things. First, it mandates “represent[ation]”—meaning a lawyer must actively appear on behalf of
every arrestee and represent them as counsel, not just passively
standby in an advisory capacity. Second, the plain language of
the statute requires that each person be represented during their
Initial Appearance.295 Dictionary definitions and case law define
the word “from” inclusively, as a “starting-point” for a series.296 It
follows that requiring representation “from initial appearance”
does not mean a judge can provide counsel toward the end of that
hearing, let alone after that hearing concludes. Rather, the law
requires judges to provide each arrestee with a lawyer to stand up
on their behalf and represent them for the entire duration of the
Initial Appearance hearing.297
The purpose and legislative history of § 3006A further demonstrate
that it requires judges to ensure that all individuals are represented
by counsel during the Initial Appearance. When Congress enacted
§ 3006A through the Criminal Justice Act of 1964 (the CJA), it
made clear that the statute was meant to guarantee “counsel at

118

Every time a judge fails
to provide an indigent
individual with a lawyer to
represent them throughout
the entirety of their Initial
Appearance hearing, she
violates federal law.

119

every stage of the proceedings,” the first of which is “the initial
appearance.”298 Congress emphasized this point when it amended
the CJA in 1970, noting that the law “requires . . . appointment of
counsel for any person under arrest,” and that this representation
“necessarily precedes the stage of formal appointment of counsel”
at the Initial Appearance.299 These reports further the purpose
of § 3006A: “assur[ing] that everyone, rich or poor, will have the
opportunity to utilize the services of an attorney as early in the
proceedings as is possible.”300
In short, the text, interpretation, and stated purpose of § 3006A
show that all arrestees are entitled to have a lawyer actively
representing them—not passively advising—during their Initial
Appearance hearing.
While we use the term “indigent arrestee” as a shorthand for someone
who cannot afford to pay for a lawyer, it is clear from the CJA’s legislative history that judges are required to appoint counsel for anyone
who cannot afford quality representation. As then-Attorney General
Robert F. Kennedy said in his testimony in support of the CJA: “[This
bill] seeks to guarantee competent legal representation and services
to every accused person whose lack of funds prevents him from
providing for his own defense.”301 He continued: “[This bill] studiously
avoids the term ‘indigent.’ Instead, it adopts the test of financial
inability to secure a necessary part of adequate representation.”302
This broad mandate for appointment of counsel squares with the
CJA’s goal of ensuring representation for arrestees at every stage of
criminal proceedings. Attorney’s fees in federal criminal cases can
easily exceed tens or hundreds of thousands of dollars—a cost many
above the poverty line are unable to afford. As a result, judges cannot
deny representation to those who cannot afford it, irrespective of
their socioeconomic status.
The Model Plan that implements the CJA also clearly requires
representation by counsel during the Initial Appearance hearing.
In a section pointedly called “Timely Appointment of Counsel,”
the Plan specifies:
Counsel must be provided to eligible persons as soon as feasible in
the following circumstances, whichever occurs earliest: (1) after they
are taken into custody; (2) when they appear before a magistrate or
district court judge; (3) when they are formally charged or notified of
charges . . . ; or (4) when a . . . judge otherwise considers appointment
of counsel appropriate under the CJA and related statutes.303
120

The “whichever occurs earliest” command means that the moment
an indigent arrestee first “appear[s]” before a magistrate judge
during the Initial Appearance hearing (circumstance (2)), the
judge must “provide[]” that person with counsel—even before the
point in that hearing when the arrestee is formally notified of the
charges (circumstance (3)).
The Federal Rules of Criminal Procedure likewise require judges
to provide counsel to represent every arrestee during the Initial
Appearance. The text of Rule 44, which governs appointment of
counsel, states that any arrestee unable to retain counsel shall have
an attorney appointed to “represent the defendant at every stage
of the proceeding from initial appearance through appeal.”304 In its
notes on the 1966 amendment to Rule 44, the Advisory Committee
explained, “The phrase ‘from his initial appearance before the
commissioner or court’ is intended to require the assignment of
counsel as promptly as possible after it appears that the defendant
is unable to obtain counsel.”305 Rule 5, which discusses Initial
Appearances specifically, also states that “[t]he judge must allow the
defendant reasonable opportunity to consult with counsel” during
the Initial Appearance hearing itself.306 Much like the CJA, these
rules establish that every arrestee has the right to be actively represented by a lawyer for the entire duration of the Initial Appearance.

Arrestees Must Be Represented by Counsel During the Initial
Appearance Because the Bond Determination Implicates
Important Legal Standards that Laypeople Cannot Apply
on Their Own.
Beyond the laws requiring representation by counsel, every arrestee
needs a lawyer to enforce the complex legal standard that applies
at the Initial Appearance, ensure that the judge does not hold an
unwarranted Detention Hearing, and prevent the person from being
locked in jail in violation of the BRA. Every federal Initial Appearance
involves a determination of detention or release governed by the
BRA. When the prosecution moves for detention at the Initial
Appearance, § 3142(f) sets forth the only lawful bases for detention
and the holding of a Detention Hearing; courts are required to release
the arrestee when no § 3142(f) factor is present.307 Unrepresented
arrestees cannot effectively enforce their own rights under that
statute, make their own bond arguments, or hold prosecutors and
judges to the law’s strictures.
121

➔

Yet our interviews suggest that many judges and practitioners are
operating under the misconception that the Initial Appearance is a
perfunctory administrative proceeding, rather than a court hearing
governed by a panoply of statutory rules.308 A core tenet of our legal
system is its adversarial nature—there must be an attorney representing each side’s interests.309 Without a lawyer, someone charged
with a crime has little chance of opposing a prosecutor’s request
that they be locked in jail. And without counsel, there is an even
higher risk that an arrestee will be detained unlawfully, in violation
of the BRA’s rules.310 In short, an unrepresented individual cannot be
expected to understand the complex legal standards that have long
eluded highly educated judges and lawyers.

The Sixth Amendment Right to Counsel Attaches at
the Initial Appearance.
In addition to the federal rules and statutes that require representation by counsel during the Initial Appearance, indigent arrestees
likely have a Sixth Amendment right to counsel during that hearing.
The Sixth Amendment “embodies a realistic recognition of the
obvious truth that the average defendant does not have the professional legal skill to protect himself when brought before a tribunal
with power to take his life or liberty, wherein the prosecution is
presented by experienced and learned counsel.”311 In Gideon v. Wainwright,312 the Supreme Court held that the Sixth and Fourteenth
Amendments entitle indigent people charged with serious crimes
to court-appointed counsel.313
In Rothgery v. Gillespie County,314 the Supreme Court unequivocally
held that the Sixth Amendment right to counsel attaches at the
Initial Appearance: “[A] criminal defendant’s initial appearance
before a judicial officer, where he learns the charge against him
and his liberty is subject to restriction, marks the start of adversary
judicial proceedings that trigger attachment of the Sixth Amendment
right to counsel.”315 A federal arrestee has a constitutional right to
appointed counsel at any “critical stage” of the proceedings after the
right to counsel has “attached.”316 The Court has never settled the
question of whether the federal Initial Appearance qualifies as a
“critical stage” and is therefore a procedure during which a person has
a Sixth Amendment right to be represented by counsel.317

122

The federal Initial Appearance is a critical stage. All available
evidence illustrates that the outcome of the Initial Appearance
may adversely affect an arrestee’s rights, and that counsel is
necessary to protect those rights under the BRA and navigate the
pretrial labyrinth. The attachment rule is supposed to ensure that a
person has a right to counsel at the point when “the accused ‘finds
himself faced with the prosecutorial forces of organized society, and
immersed in the intricacies of substantive and procedural criminal
law’”—such as the intricacies of the legal standard that applies
during the federal Initial Appearance.318
Case law is clear that a proceeding qualifies as a critical stage if it
meets two criteria: the outcome could substantially prejudice an
individual’s rights, and an attorney could remedy that potential
prejudice. “Critical stages” are those that hold “significant consequences for the accused”319 where there is “a need for counsel’s
presence.”320 At a critical stage, counsel is needed to assist in “coping
with legal problems or . . . meeting [the defendant’s] adversary.”321
It does not matter whether the hearing is “formal or informal.”322
Rather, what matters to the critical stage analysis is “whether
potential substantial prejudice to a defendant’s rights inheres
in the particular confrontation and the ability of counsel to help
avoid that prejudice.”323
First, the outcome of an Initial Appearance has the potential to
significantly prejudice the rights of the accused—most prominently,
the right to liberty. Courts recognize that the potential loss of one’s
liberty is sufficiently significant to trigger the Sixth Amendment
right to counsel.324 Federal and state courts alike have acknowledged
that, given the potential loss of liberty involved, a state bail hearing is
a critical stage requiring appointed counsel.325
That the Initial Appearance may prejudice a person’s liberty is
aggravated by the fact that pretrial detention adversely affects the
disposition of one’s entire case, making conviction at trial more
likely. When a person is detained pretrial, they are more likely to
be convicted because they have a diminished ability to confer with
their lawyer and gather evidence.326 Additionally, individuals who are
detained pretrial are more likely to receive a higher sentence, a within-Guideline sentence, and a mandatory minimum sentence than
those who are released.327 Pretrial detention may also impede access
to counsel at later stages of the case.328
The risk of prejudice to an arrestee at the Initial Appearance extends
far beyond the deprivation of liberty. Even a few days in pretrial
123

➔

detention can cause people to lose their jobs, their homes, and
custody of their children.329 These consequences accrue regardless
of whether counsel is ultimately appointed at a later critical stage.
The Initial Appearance therefore can substantially prejudice
an arrestee’s rights.
Second, the Initial Appearance is a critical stage because a lawyer
can remedy the inherent potential for prejudice by ensuring that
the judge adheres to the proper legal standard and does not detain
the arrestee unlawfully in violation of § 3142(f). Arrestees cannot
be expected to be familiar with this legal standard, to understand
that the law authorizes a judge to hold a Detention Hearing only in
certain types of cases, or to hold prosecutors to their burden under
§ 3142(f).330 This Report shows that even judges and prosecutors often
misunderstand the legal nuance of the BRA’s provisions. Arrestees
are simply not intimately familiar with the legal intricacies of the
pretrial detention process. Because “the function of counsel as a
guide through the complex legal technicalities”331 of the BRA is
essential to remedy the potential for prejudice that arises during the
Initial Appearance, the second prong of the critical stage test is met.
Two circuit cases—both of which predate Rothgery—erroneously
suggest that the Sixth Amendment does not require representation
at the Initial Appearance. Their reasoning, however, supports
the position that the Sixth Amendment does require counsel at
the Initial Appearance.
In United States v. Perez,332 the Ninth Circuit considered whether
an arrestee has a right to counsel at the Initial Appearance before
the court asks the arrestee their name. “Significantly, counsel
was appointed when the proceedings may have affected his rights
following the initial determination of the defendant’s name.”333
The court concluded that Perez did not have a right to counsel
before being asked his name because “[n]othing at this stage of the
proceedings . . . impairs the defense of the accused.”334 That understanding may have been plausible in 1985 when Perez was issued,
but as described above, more recent evidence shows that the deprivations of liberty that commonly occur during Initial Appearances
are causally related to an increased risk of conviction.335 In United
States v. Mendoza-Cecilia,336 the Eleventh Circuit similarly disregarded the importance of the Initial Appearance, calling it “largely
administrative” and noting that the bail hearing is “not a trial on
the merits.”337 Perez and Mendoza-Cecilia unfairly and incorrectly
minimize the role that an attorney can play in safeguarding an
arrestee’s liberty, and other courts have not followed these cases
124

when evaluating the equivalent of the federal Initial Appearance in
state systems.338
As this Report has shown, the Initial Appearance is anything but
administrative. Instead, it is a critical stage at which arrestees
can and do suffer unlawful deprivations of liberty if not represented by counsel. Without counsel present to insist that judges
and prosecutors follow the oft-ignored legal standard in § 3142(f),
arrestees are jailed unlawfully and are subjected to unauthorized
Detention Hearings.
Some judges nevertheless perpetuate the view that the Initial
Appearance is a benign administrative procedure rather than a
critical stage. One judge we interviewed echoed these misconceptions, musing, “Maybe I shouldn’t even be giving people lawyers
[at the Initial Appearance]. But I think [the practitioners] would be
amenable to [universal representation by counsel], because honestly
having a lawyer helps; it just helps everything go faster.” It should
go without saying that a lawyer’s job at the Initial Appearance is
not to make things “go faster,” but instead to protect the accused’s
presumption of innocence and profound interest in liberty.
The Due Process Clause of the Fifth Amendment also protects
an arrestee’s right to be represented by a lawyer during the Initial
Appearance. Due process requires “constitutionally sufficient”
procedures to prevent against unnecessary or erroneous deprivations
of liberty.339 At the Initial Appearance, an arrestee has a clear liberty
interest against being detained pretrial. And, as explained above,
the arrestee cannot hope to vindicate that liberty interest without
a lawyer present. The legal technicalities of § 3142(f) are beyond the
grasp of people without legal training. Accordingly, several courts
have held that the Due Process Clause requires representation by a
lawyer at the state equivalent of an Initial Appearance.340

Conflict-of-Interest Concerns Do Not Prevent Courts
from Providing All Arrestees with Representation at
the Initial Appearance.
The plain legal requirements of § 3006A, Rule 44, and the
Constitution override any practical or ethical concerns with
ensuring that every arrestee is represented by counsel during
the Initial Appearance.
In the course of our interviews, some judges claimed that they did
not have enough lawyers on hand to provide representation to every

125

➔

arrestee during the Initial Appearance. One judge described the
problem this way: “What happens logistically here in the courtroom
is that the CJA lawyers are not there. And so, the PD [federal public
defender] is there, so anytime you can appoint the PD, you appoint
the PD. But, if the PD is not there, I have no way [to provide counsel],
unless I tell the CJA lawyers.” When asked why the judges in that
particular district allow arrestees to be jailed without counsel during
the Initial Appearance, the judge blamed the culture, saying that it is
“certainly an issue of the practice in the district.”
Beyond this logistical objection, practitioners in many districts told
us about a second major barrier to ensuring universal representation
at Initial Appearances: judges refuse to allow an appointed lawyer
to represent multiple co-defendants during the Initial Appearance,
for fear of creating an ethical conflict of interest. As one Chief
Defender exclaimed when describing this issue: “[The judges are]
more concerned about the conflict [of interest] than they are about
the Sixth Amendment!”
Our interviews with stakeholders from other districts show that
any logistical or ethical obstacles can be overcome with a bit of
planning and coordination. Every federal district in the country can
easily follow the example of the many districts that ensure that all
arrestees are represented by counsel during the Initial Appearance,
such as the Northern District of Illinois, the Southern District of
California, and the Eastern and Western Districts of Wisconsin.
These districts and more allow the federal defender duty attorney
to represent arrestees in conflict cases, and even to represent
multiple co-defendants in the same case—with the explicit, on-therecord caveat that such representation is for the purposes of the
Initial Appearance only.341
In districts where the duty defender represents multiple arrestees
during the Initial Appearance, the attorney does not discuss the
details of the offense with anyone, focusing instead on matters
relevant to the legal standard at the Initial Appearance. In some
districts, the attorney also obtains a verbal waiver of any conflict
of interest from all arrestees. As soon as practicable—and well in
advance of any Detention Hearing—CJA counsel are appointed
to represent all but the one co-defendant who continues to be
represented by the federal defender’s office. This approach is
supported by the 1979 amendment notes to Rule 44, which clarify
that “[t]he defendant should be fully advised by the trial court of the
facts underlying the potential conflict [of interest] and be given the
opportunity to express his views.”342
126

Allowing a duty defender to provide temporary representation of
multiple clients with the informed consent of each is markedly preferable to forcing arrestees to face their Initial Appearances alone. One
Chief Defender described the process as follows: The federal defender
duty attorney “is prepared to represent multiple people at an Initial
Appearance . . . with the court’s permission, and the court almost
always permits it.” The duty defender avoids conflict of interest
concerns by telling “all the clients that [they are] meeting with at
the time, ‘I’m not going to talk to you about the facts of the case. . . . I
don’t know which of you, yet, I will be representing. I want you to get
your court appearance today, so . . . with your permission, I’m going
to represent you and the others, but because of that, I can’t get into
the facts. . . .’ [A]nd they’ll agree because they want it—they’ll want
their court appearance.”
While interviewing a judge in a district where arrestees often go
unrepresented at the Initial Appearance, we asked if the judges there
would be amenable to the federal defender duty attorney simply
representing every arrestee during that hearing. The judge considered
this solution and responded, “I’m not sure why they couldn’t do
that.” It became clear during our interviews that stakeholders in
that district did not consider the absence of counsel at the Initial
Appearance to be a problem, and therefore had made no effort to
craft a solution.
Another Chief Defender described an even better method for
ensuring representation and avoiding conflicts of interest: at Initial
Appearances in multi-defendant cases, the duty defender represents
one person, and the judge or defender’s office secures CJA panel
attorneys to represent the remaining co-defendants.
This Defender noted that her district would “have [a federal
defender] duty attorney and a CJA duty attorney assigned [to Initial
Appearances] for each day and ready to roll.” The Defender said the
court had also set up a “fairly complicated computer system” to
ensure that CJA attorneys could always be reached for Initial Appearances.343 Finally, the Defender explained that it was possible to secure
extra attorneys for multi-defendant cases with a bit of prior warning:
“[E]ven in some of our smaller offices . . . they [would reach] out to me
and [say], ‘Hey, we have a big case coming in, we don’t have enough
people on the [CJA] panel. Can you tell us some people [from another
district/division] who would be available?’ . . . [T]here’s no reason to
be scrambling.” Though the Defender recalled some initial pushback
from judges when this practice began, “it turned out, they loved it.”
127

➔

C

The Practice: At the Initial Appearance
Hearing, Federal Judges Unlawfully Jail
Poor People Without Lawyers.
Our study uncovered a nationwide access-to-counsel crisis in
the federal system. During our courtwatching, we observed judges
violating federal law by holding uncounseled Initial Appearances
where indigent individuals appeared pro se across from federal prosecutors, and by jailing indigent individuals without first appointing
them a lawyer. We also observed Initial Appearances where indigent
arrestees were represented during only part of the hearing and made
incriminating statements before being given a lawyer.
During interviews and conversations with defenders and judges, we
learned that this problem extends far beyond the districts where we
watched deprivations of counsel unfold before our eyes. In fact, more
than one-quarter of the 94 federal districts nationwide, and possibly
more, do not ensure that every single arrestee is represented by a
lawyer at the Initial Appearance. See Figure 6. These widespread
deprivations of counsel violate the law, increase an arrestee’s
chances of being detained or incriminating themselves, and exacerbate existing racial disparities in the federal system.

Judges in Many Federal Courts Lock Poor People in Jail Without
Lawyers at the Initial Appearance.
a.

Our Courtwatching Data Illustrate the
Access-to-Counsel Crisis.

Among the 4 federal districts where we courtwatched, we found a
concerning access-to-counsel problem. Arrestees were fully represented by counsel in only 60% of Initial Appearances across the
4 districts. In 30% of cases, arrestees were not provided counsel until
partway through the hearing.344 Most troubling, we found that in 7%
of cases, arrestees were entirely deprived of counsel and were forced
to appear pro se during the Initial Appearance.345

128

During our courtwatching,
we observed judges
violating federal law by
holding uncounseled Initial
Appearances where indigent
individuals appeared
pro se across from federal
prosecutors, and by jailing
indigent individuals without
first appointing them
a lawyer.

129

In fact, more than onequarter of the 94 federal
districts nationwide, and
possibly more, do not ensure
that every single arrestee is
represented by a lawyer at
the Initial Appearance.

130

In some districts, however, the problem is far worse. Of the 4 districts
in our study, the Southern District of Florida was the only court
where we observed judges holding Initial Appearances and locking
people in jail with no defense lawyer present. See Figure 16. In that
district, indigent arrestees were not represented by counsel in 11%
of Initial Appearances, and they received only partial representation about 40% of the time. See id. That is, in fewer than half of
the Initial Appearances we observed in the Southern District of
Florida, judges ensured that arrestees were represented by counsel
during the entire hearing.

D-

Remarkably, our conversations with stakeholders from the Southern
District of Florida suggest that these statistics actually underestimate the problem. One Defender told us that the pandemic
actually improved this situation, and that the portion of arrestees
who went unrepresented in that district was “much lower than
it would be in normal times.” Likewise, one judge estimated that,
before the COVID-19 pandemic, around 30% of Initial Appearances
occurred without a defense lawyer representing the accused at
any point of the hearing.
The following example typifies the many Initial Appearances we
witnessed where judges locked poor people in jail without a lawyer.
In this case, there was no defense lawyer present during the hearing.
The prosecutor requested a Detention Hearing on invalid grounds,
and the judge unquestioningly detained the arrestee for 3 days:

131

JUDGE: Ok, I find at this time that you qualify for the appointment
of an attorney. So what we will do since your co-defendant is already
being represented by the [Federal Public Defender], we will have to
appoint you a . . . CJA attorney . . . . [T]he court will contact him, let him
know of his appointment so he can contact you. The next matter we
discuss . . . is the right to a bond hearing. In this matter, I’m going to hear
from the government.
PROSECUTOR: The government is requesting pretrial detention based on
risk of flight and danger to the community, but if we have a couple of days
I can speak with [appointed counsel] and I’m sure we can figure it out.
JUDGE: That’s a matter that your attorney will be handling on your
behalf. I’m going to set that [Detention] Hearing for three days from today,
that would be February 25th in the afternoon, so your attorney will be in
contact with you regarding that hearing. And that will be the next hearing
in your case. It may be that he will be discussing that with the government
ahead of time.
ARRESTEE: Ok.346

The failure to provide counsel during the Initial Appearance in this
case and others not only violated multiple laws, but also ran afoul
of the timing requirements for appointment in Southern District of
Florida’s CJA Plan, which are identical to those in the Model Plan.347
Rather than automatically setting Detention Hearings and jailing
unrepresented people until those hearings, judges must adhere to the
law and ensure that each arrestee is represented for the entirety of
their Initial Appearance.

b.

Stakeholder Interviews with Defenders Across the
Country Demonstrate the National Scope of this
Access-to-Counsel Problem.

The deprivation of lawyers during the Initial Appearance is a
national problem. Our qualitative research confirms that in more
than one-quarter of the 94 federal districts in this country, judges
fail to provide a lawyer to all arrestees at the Initial Appearance.348
This access-to-counsel problem spans at least 26 federal district
courts and 9 federal circuits.349 See Figure 5. It is likely this underestimates the problem, as our interviews covered less than half of
the federal districts.
132

We confirmed through our research that in at least 16 districts,
judges do not provide every indigent individual with an appointed
attorney to represent them during the Initial Appearance: the
Southern District of Florida, the Western District of Texas, the
Southern District of Texas, the Northern District of Mississippi, the
Western District of Missouri, the Middle District of North Carolina,
the Southern District of Ohio, the District of Delaware, the Central
District of Illinois, the Eastern District of North Carolina, the District
of South Carolina, the Eastern District of Virginia, the District of
Arizona, the Northern District of California, the District of New
Mexico, and the District of the Virgin Islands.350 In addition to these
16 named districts, in 10 more districts, we heard from at least one
Defender or CJA attorney that people are not always represented by
counsel for the entirety of the Initial Appearance, bringing the total
number of districts to 26—which is 28% of all of the federal district
courts in this country.351
Courts in these 26 districts not only violate the federal laws
discussed above and risk unconstitutional deprivations of counsel,
but also contravene the timing requirements for appointment of
counsel in their own CJA Plans.352
In at least two of the districts named above, 100% of arrestees are
not represented during the Initial Appearance. One Defender told us
that, in nearly a decade with the office, he had never heard of anyone
representing an arrestee at the Initial Appearance; all indigent
arrestees in his district are forced to appear pro se during that first
hearing. In a third listed district, over 50% of people are not represented. In a fourth named district, a Defender estimated that 40%
of arrestees currently go without counsel at the Initial Appearance,
but explained that this was an improvement over the past, when
zero arrestees were represented during that proceeding. A judge in
that same district told us that even when federal defenders were
present in the courtroom in one of the divisions in the district, “they
would sit there, and unless the question [of whether counsel would
be appointed] came up, the person [would] always go through the
Initial Appearance unrepresented.”
That vignette describes a common scenario in a number of the
16 districts listed by name in this Report—a federal defender
duty attorney is physically present in the courtroom, but serves
“primarily in an advisory capacity” or “standby” capacity, according
to interviewees. Notably, that attorney does not actively represent
arrestees during the Initial Appearance or present legal arguments
on their behalf.
133

Instead, the duty attorney speaks with the arrestee before the
hearing, informs them of their rights, perhaps fills out the financial
affidavit with them, and then sits back and does not otherwise
participate in the hearing. One Defender described the defense
lawyer as being simply “a warm body.” This practice must also
change, as it is simply another variation on the denial of counsel.
An attorney standing by as a warm body is not playing their
requisite adversarial role.
In other named districts—including the District of South Carolina,
the Eastern District of Virginia, the District of Arizona, the Northern
District of California, the District of New Mexico, and the District
of the Virgin Islands—arrestees are often represented by appointed
counsel during the Initial Appearance. However, in each of these
6 districts, there are certain types of cases in which arrestees are not
represented by appointed counsel. For instance, in some districts, the
practice in multi-defendant cases is for the federal defender’s office to
represent one arrestee during the Initial Appearance while the others
go unrepresented. In other districts, people without legal status in
the United States are not consistently represented during the Initial
Appearance. In still other districts, the practice recently changed
in an effort to provide counsel, but people continue to go without
lawyers during the Initial Appearance in a subset of cases.353
We also learned that several additional districts beyond these
26 historically did not provide all indigent arrestees with counsel
during the Initial Appearance. However, those districts changed their
practices fairly recently to provide lawyers for all arrestees during
the Initial Appearance. The districts where the practice has changed
include the District of Puerto Rico, the Eastern District of Louisiana,
the District of Colorado, the District of Wyoming, and the Northern
District of Indiana.354 In all 5 of these districts, there was previously
a blanket practice of never appointing counsel to represent arrestees
during the Initial Appearance. Accordingly, until fairly recently,
one-third of federal districts locked indigent individuals in jail
without counsel.355
A major driver of this access-to-counsel problem is that most of
the judges and many of the Chief Federal Defenders we interviewed
admitted to not knowing the legal rules that apply at the Initial
Appearance under § 3142(f).356 In fact, some were surprised to learn
that there is, indeed, a legal standard that governs during that
hearing. These misapprehensions help explain why judges, and
even some defenders, do not see a problem with people being jailed
134

without lawyers. During our interviews, we heard many variations
on the question asked by one Chief Defender: “What are the kinds of
legal arguments that are typically made at the Initial Appearance?” In
districts where the judges and defense bar are not aware of the legal
standard, they also may not understand why a lawyer is necessary to
guarantee that standard is followed.

D-

In the many districts that allow poor people to be jailed without
lawyers, Defenders nearly universally reported that federal prosecutors “move to detain everyone” at the Initial Appearance. And why
wouldn’t they, when there is no lawyer on the other side to raise a
legal objection to jailing?

Arrestees Without Lawyers at the Initial Appearance Are
Detained at Far Higher Rates than Those with Counsel.
Unsurprisingly, our data show that detention rates at the Initial
Appearance increase as representation decreases. Arrestees who
received full representation at the Initial Appearance were detained
in 67% of cases. Arrestees who received partial representation, in
contrast, were detained in 89% of cases357—a 23-percentage-point
increase. Finally, when the accused had no lawyer at all, every single
Initial Appearance ended with the arrestee going to jail. See Figure 18.

Figure 18: When Arrestees Were Forced to Proceed
Without Counsel at Their Initial Appearance, There
Was a Dramatic Increase in Pretrial Detention.

Fully represented individuals
were detained 67% of the time.

67%

89%

100%

Partially represented individuals
were detained 89% of the time.
All unrepresented individuals

were detained.

135

A failure to provide counsel
at the Initial Appearance
acts as a de facto
detention order.

100%
136

This finding suggests that a failure to provide counsel at the Initial
Appearance acts as a de facto detention order. When courts deny the
accused the assistance of a lawyer, they not only violate the accused’s
statutory and constitutional rights, but they effectively condemn
that person to jail time—and all the adverse consequences that
come with it.
To make matters worse, unrepresented individuals are frequently
prevented from advocating on their own behalf during the Initial
Appearance, with judges and prosecutors alike expressing reservations about speaking directly with the accused. A judge we interviewed candidly said: “[I]f I see a defendant is going to do anything
at that Initial Appearance [when] they didn’t have an attorney . . . I’m
just not going to let them talk, you know, I’m going to stop them. . . . I
think that most magistrate judges I know are like that, no matter
where they fall on that law-and-order spectrum.” This approach shuts
down self-advocacy, an accused’s only possible avenue to release
when the judge fails to appoint a lawyer.
At the outset of one Initial Appearance hearing we watched, the prosecutor appeared ready to agree to the arrestee’s release on a personal
surety bond. But after the arrestee mentioned that she was homeless
during questioning by the judge, the prosecutor changed his mind
and requested detention. Because the arrestee did not have counsel to
assist her, she tried to speak up on her own behalf:

JUDGE: The detention hearing on Wednesday at 10 a.m., and I will set
the arraignment—
ARRESTEE: Can I speak?
JUDGE: —also. Yes, ma’am.
ARRESTEE: I am at this time at a friend’s house. And I swear that I am
not going to flee. I am not going to do anything bad. My dog is staying
with them as well.
JUDGE: All right, ma’am. Mr. [Prosecutor], does that change the
government’s position?
PROSECUTOR: Your Honor, quite frankly I would like to have a
conversation with appointed defense counsel about this as opposed to
the defendant.358
137

Of course, since there was no lawyer representing the woman during
her Initial Appearance, there was no one for the prosecutor to have
that conversation with, and the woman was ultimately detained.
In other cases, judges detained unrepresented arrestees unlawfully,
without any valid basis for detention under the BRA. A judge should
take extra pains to ensure that an unrepresented arrestee is, in fact,
eligible for detention under § 3142(f) at the Initial Appearance, since
the accused is not protected by adversarial safeguards. Nonetheless,
we observed two cases where the court both failed to provide counsel
and failed to observe the law, which did not authorize pretrial
detention. Cases like these are especially egregious, as judges
violate the law twice over: they violate the accused’s legal right to
be represented by counsel and detain the accused unlawfully in
counsel’s absence.

Unlawful Detention Squared: Judges Detain
Arrestees Illegally and Without Lawyers
In a fraud case in which there was no offense-specific basis for
detention at the Initial Appearance, a judge ordered the arrestee
detained, even though the prosecutor did not request detention, let
alone meet the legal standard. The prosecutor did not say a single
word outside of his introduction, but the judge entered an illegal
detention order without a second thought. In fact, when detaining
this individual, the judge even acknowledged that the arrestee did
not have a lawyer:

JUDGE: Okay. So I’ll appoint [name of lawyer] on this case.
He’s not present, but he will be in contact with you regarding
your case.
COURTROOM DEPUTY: You want to set it for—
JUDGE: So I’m going to set your case for arraignment for a
detention hearing and for arraignment. So your next court
hearing will be November 19th, and before that your attorney
will be in contact with you to discuss your case with you.359

138

Lack of Counsel at the Initial Appearance Exacerbates
Racial Disparities.

D-

In our study, the vast majority of people whom judges unlawfully
deprived of counsel at the Initial Appearance were Black and
brown individuals. In 92% of the Initial Appearances we witnessed
where the court failed to appoint a lawyer, the arrestee was either
Black or Latino. In fact, the proportion of people of color who faced
pro se Initial Appearances exceeded the representation of people
of color in our sample writ large.360 Only two white arrestees in our
sample had to face the Initial Appearance without counsel, compared
to 23 people of color. See Figure 17.
Black and Latino individuals were also more likely than white
individuals to face the Initial Appearance without the full assistance
of counsel. While just 13% of white arrestees were only partially
represented at their Initial Appearance, over 30% of Black and Latino
individuals faced the same problem—more than twice as many.361
Moreover, white arrestees received full representation in
80% of their Initial Appearances, while Black and Latino
arrestees received full representation in 60% and 53% of their
Initial Appearances, respectively.
Even being deprived of a lawyer for part of the Initial Appearance
can seriously prejudice the accused. A person can inadvertently
incriminate themselves while responding to a judge’s questions
about their financial circumstances, even after the judge advises
them of their rights under the Fifth Amendment. Language barriers
can further complicate the fraught nature of facing federal criminal
charges without a lawyer. In one case, for example, the court questioned a Venezuelan man without counsel. He was charged with
a drug crime while aboard a vessel subject to U.S. jurisdiction and
answered the judge’s questions through an interpreter:

139

JUDGE: Other than that property, do you have any other personal property
or bank account, financial account greater than 5,000 US dollars?
ARRESTEE: The properties I don’t have, vehicles I don’t have those either.
Bank account I do have.
JUDGE: What is the balance in your bank account?
ARRESTEE: Currently, it’s probably zero. That was the reason why
I accepted that job.
JUDGE: Don’t say anything about anything that happened to you. So let
me find that you qualify for appointment of counsel, and I will therefore
appoint the—well, for now the Federal Public Defender is to represent you
in connection with this proceeding.362

By the time this man was given a lawyer, he had already incriminated
himself. As long as courts fail to provide representation to everyone
from the start of their Initial Appearance onward, people accused
of serious crimes will continue to make incriminating statements.
And as long as Black and Latino are given lawyers less often than
white arrestees—as was the case in our study—these incriminating
statements will disproportionally harm people of color.

140

D

The Solution: At the Initial Appearance,
Judges Must Follow the Law and Appoint
Lawyers to Actively Represent Every
Indigent Arrestee.

D-

The responsibility for this access-to-counsel crisis falls squarely
on the judges’ shoulders. Federal law and district CJA Plans task
judges with appointing counsel to represent indigent individuals
during Initial Appearance hearings. The text and purpose of Rule
44 and § 3006A(c) clearly establish that every arrestee must have a
lawyer to represent their interests during their Initial Appearance.
Every time a federal judge fails to provide a lawyer for an arrestee
during the Initial Appearance and forces that person to appear pro se
across from a prosecutor, they violate the law. Locking people in jail
without lawyers is a perversion of the adversarial system.
Courts must ensure that every person accused of a federal crime
is actively represented by a lawyer from the start of their Initial
Appearance, and at least before any detention or release determination is made. Providing universal representation at federal Initial
Appearances will stem high detention rates and avoid jailings that
are simultaneously unlawful and uncounseled.
The best practice is to have a duty federal defender or CJA lawyer
actively represent arrestees at every Initial Appearance, and for judges
to secure additional CJA lawyers whenever the federal defender’s
office is conflicted out. This has been the practice in the District of
Massachusetts for decades; the Western District of Virginia also uses
this model, and the District of Colorado recently began endeavoring
to follow it as well.
Many courts already provide counsel for every indigent arrestee
from the very start of the Initial Appearance, even before a judge
reviews the financial affidavit. Other courts, however, stand on
ceremony by refusing to allow counsel to represent an arrestee at
the Initial Appearance until pursuing a lengthy colloquy regarding
the indigency determination. This latter practice must cease, as it
jeopardizes an arrestee’s liberty interest and Fifth Amendment right
against self-incrimination.
141

It also disregards the fundamental purpose of the CJA, which was to
provide an appointed attorney for every arrestee “whose lack of funds
prevents him from providing for his own defense,” not just for those
who fall below the poverty line.363
Stakeholders we interviewed in districts that adhere to best practices
described various ways courts determine whether a person is entitled
to appointed counsel without questioning an unrepresented arrestee.
One judge described how he appoints counsel before the Initial
Appearance to facilitate that determination: “[W]hat I started doing
was, I didn’t care if they qualified or not, I was just appointing CJA
to help facilitate them getting an attorney. . . . [T]he staff determines,
after Pretrial Services interviews them, whether they’re going to
qualify, and whether the public defender has a conflict. And if the
public defender has a conflict, my case manager will ask me, and
we’ll appoint a CJA attorney before the initial conference.” A Chief
Defender discussed similar practices in his district: “[I]n terms of the
indigency determination, the clients fill out a financial affidavit, and
it’s submitted to Pretrial Services, who then submit it to the Court.”
He continued: “And I don’t—honestly in my seven years, six years
here, I’ve never seen—I don’t care how much the client makes—I’ve
never seen a judge deny a client court appointed counsel, because,
you know, it costs $90,000 just to fight a federal case, you know.
So, that’s how they do [it], and then in terms of the questioning,
usually the only question that is posed is: do you want counseling?
That’s really it.”
Among the districts that fail to provide counsel to each arrestee at
the Initial Appearance, many Defenders told us that their offices had
campaigned for years to convince judges to allow them to represent
arrestees during these hearings, but to no avail. When we asked
Defenders why judges were so resistant to following the law and
allowing them to provide representation at the Initial Appearance,
this statement of one interviewee was echoed by many others: “They
think we’re a fly in the ointment. They think having lawyers there will
slow down their process.” Another Chief Defender concurred: “It’s
very efficient this way. The judges say it’s too much trouble to have
lawyers there [representing people during the Initial Appearance].”
However, others told us that efficiency concerns cut in the opposite
direction, in favor of providing lawyers. One Chief Defender said that
the culture in their district began to change during the pandemic:
“They let us participate [in Initial Appearances] because they
suddenly realized how helpful we are.” Another Chief Defender told
us that judges in their district appoint counsel in every case because
142

Locking people in jail
without lawyers is a
perversion of the
adversarial system.

143

Courts must ensure that
every person accused of
a federal crime is actively
represented by a lawyer
from the start of their
Initial Appearance, and at
least before any detention
or release determination
is made.

144

it makes things run more smoothly: “Judges want experienced
counsel there [at the Initial Appearance] for efficiency.”
Our courtwatching data and qualitative interview findings
show that this access-to-counsel crisis also stems from judges
prioritizing district culture above the law. “[The judges] don’t want
anything to change,” said one Chief Defender. “They’ve always done it
this way,” said another. This sentiment was echoed in our interviews
with judges. When we asked a judge in a district where judges
regularly fail to provide poor people with lawyers why that practice
had persisted for so long, they candidly responded, “Sometimes we do
things around here just because—[and] I imagine this is the truth
everywhere . . . —‘Oh that’s just the way we do it.’” This statement
provides an explanation, but it is not an excuse for violating
people’s legal rights.

*

The Department of Justice is also to blame for allowing this unconscionable situation to continue. Our research establishes that, in
over one-quarter of federal courts, Assistant United States Attorneys
regularly ask judges to deprive unrepresented individuals of their
liberty. Under the ethical rules, the prosecutor serving as the lawyer
for the United States government “has the responsibility of a minister
of justice and not simply that of an advocate,” and “this responsibility carries with it specific obligations to see that the defendant is
accorded procedural justice.”364 In a criminal case, there can be no
procedural justice without defense counsel.
Federal prosecutors should never participate in any hearing in a
criminal case when there is no advocate on the other side. Instead,
there should be a blanket DOJ policy requiring prosecutors to insist
on the appointment of counsel at the Initial Appearance before any
hearing is held—especially any hearing where they seek to jail the
accused. If DOJ adopts such a policy, it is imperative that judges
and other stakeholders hammer out the logistics so that counsel
is provided promptly and the Initial Appearance is not delayed
to procure counsel.
The obstacles to providing representation at the Initial Appearance
are bureaucratic, but the right to a lawyer is ensconced in well-settled
law, and the liberty interest protected by defense lawyers at such
hearings is fundamental. Federal courts must prioritize the law
over the culture of detention and guarantee that every person
charged with a federal crime in this country is represented by a
defense lawyer—playing their proper adversarial role—from the
beginning to the end of their Initial Appearance.
145

3

Judges Must Follow the
Correct Legal Standard in
Presumption-of-Detention
Cases to Reduce Racial
Disparities and High
Federal Jailing Rates.

The statutory presumption of detention that applies
to many federal offenses during the Detention Hearing
is a primary driver of sky-high pretrial detention rates.
The practice surrounding the presumption—and the
presumption itself—contradict the sacred promise of
“innocent until proven guilty”365 and reverse the BRA’s
“clear preference for pretrial release.”366 Courts misunderstand and misapply the presumption of detention,
further contributing to the culture of detention.
The most common presumption of detention in the
BRA reads: “Subject to rebuttal by the person, it shall be
presumed that no condition or combination of conditions
will reasonably assure the appearance of the person as
required and the safety of the community if the judicial
officer finds that there is probable cause to believe that
the person committed” most federal drug offenses, gun
offenses charged under § 924(c), minor victim offenses,
and several other crimes.367
146

Congress intended this presumption of detention
to capture only the “worst of the worst” offenders.368
“[L]egislators wanted the drug presumption to prevent
rich people suspected of high-level drug trafficking from
fleeing.”369 But in practice, the presumption now applies
in a high percentage of federal cases—including 93% of
federal drug cases370—very few of which pose any special
risks of flight or recidivism. In our Chicago pilot study,
nearly 90% of the presumption cases we observed were
drug cases.371 Given the prevalence of presumptiontriggering charges, it is especially important that judges
hew carefully to the limitations the statute and case law
place on the presumption.
Our study finds that judges routinely misapply the
presumption of detention, giving it more weight than
the law allows. As a legal matter, the presumption is
rebuttable, but judges fail to treat it as such. This appears
to be a nationwide problem, as our courtwatching data
are supported by interviews with stakeholders from
around the country. In fact, many of the judges we interviewed made statements indicating that they either
misunderstand the presumption of detention or do
not apply it correctly.
Our findings corroborate a recent government study’s
observation that the presumption of detention has
become “a built-in bias for incarceration that feeds the
federal system’s colossal detention rates and stark racial
disparities,” and “deprives nearly every person awaiting
trial in a federal drug case of their liberty.”372
To ensure that the presumption of detention does
not swallow the presumption of release, courts must
follow the letter of the law by treating the presumption
as rebuttable, giving the presumption the low weight
to which it is assigned by case law, and holding the
prosecution to its burden of proving that detention
is genuinely necessary.

147

A

Takeaways

• In our study, arrestees facing a presumption of
detention were detained at a rate of 72%, which
exceeded the rate of detention among arrestees to
whom a presumption did not apply by nearly 20%.
See Figure 19.

• Judges detained arrestees facing a presumption of detention in
87% of the cases in which the prosecutor explicitly invoked the
presumption during the Detention Hearing, compared to 76% of
the time when the prosecutor did not invoke the presumption.

• In 95% of the contested Detention Hearings we observed where the
presumption of detention applied, judges either failed to mention
whether the presumption of detention was rebutted or concluded
that the presumption was not rebutted. See Figure 20.

• In 100% of Detention Hearings where the judge found that the
presumption had not been rebutted, the judge detained
the arrestee.

148

72%
55%
87%
76%
95%

100%

Detained when
presumption
applied
Detained when
presumption
did not apply

p

Racial Disparities
in presumption-of-detention cases

89%

• In the cases where a presumption of detention applied,
89% of the arrestees were people of color.

97%

• Among the Detention Hearings where prosecutors
invoked the presumption of detention, 97% of the
arrestees were people of color. See Figure 21.

100%

• Prosecutors erroneously invoked the
presumption of detention exclusively against
Black or Latino arrestees.

People of color
detained

73%

White people
detained

68%

• Judges detained people of color at higher rates
than white individuals: the detention rate in
presumption-of-detention cases involving people
of color was 73%, while the detention rate in
presumption-of-detention cases involving white
arrestees was just 68%. See Figure 22.

100%
149

THE STATUTE: 18 U.S.C. § 3142(e)(3)
Subject to rebuttal by the person, it shall be presumed that no condition or combination of conditions will reasonably assure the appearance of the person as required
and the safety of the community if the judicial officer finds that there is probable
cause to believe that the person committed—
(A) an offense for which a maximum term of imprisonment of ten years or
more is prescribed in the Controlled Substances Act (21 U.S.C. 801 et seq.), the
Controlled Substances Import and Export Act (21 U.S.C. 951 et seq.), or chapter
705 of title 46;
(B) an offense under section 924(c), 956(a), or 2332b of this title;
(C) an offense listed in section 2332b(g)(5)(B) of title 18, United States Code, for
which a maximum term of imprisonment of 10 years or more is prescribed;
(D) an offense under chapter 77 of this title for which a maximum term of
imprisonment of 20 years or more is prescribed; or
(E) an offense involving a minor victim under section 1201, 1591, 2241, 2242,
2244(a)(1), 2245, 2251, 2251A, 2252(a)(1), 2252(a)(2), 2252(a)(3), 2252A(a)(1),
2252A(a)(2), 2252A(a)(3), 2252A(a)(4), 2260, 2421, 2422, 2423, or 2425 of this title.

150

B
The Law: The Presumption of Detention
Was Intended to Apply Narrowly and to Be
Easily Rebutted by the Defense.
While the Supreme Court has expressly recognized that “freedom
should be the [norm] for people awaiting trial,” the statutory
presumption of detention leads “judges to assume people charged
with certain crimes . . . will flee and endanger the community if
released.”373 At the Detention Hearing, the presumption of detention
applies to most cases involving drugs, gun charges under § 924(c),
minor victims, or terrorism.374 Although the presumption was
supposed to apply to only the most serious of cases, it now applies
to nearly half of all federal cases.375 To temper this reality, case law
emphasizes two checks that the BRA and the Constitution impose on
the presumption: (1) there is an easy-to-meet standard for rebutting
the presumption and the prosecution always bears the burden
of persuasion, and (2) the presumption alone does not warrant
detention and must always be weighed along with other factors.

Congress Intended the Presumption of Detention to Apply
Narrowly, But It Has Become Expansive in Modern Practice.
The legislative history of the BRA makes clear that Congress
intended the presumption of detention to apply to only a narrow
set of cases—a group they thought of as the “worst of the worst”
offenders.376 The Senate Judiciary Committee, for example, argued
that there should be a presumption of detention for the “most
serious” drug offenses because it is “well known that drug trafficking
is carried on to an unusual degree by persons engaged in continuing
patterns of criminal activity” who “have both the resources and
foreign contacts to escape to other countries with relative ease.”377
However, the BRA was also drafted amidst the so-called “War on
Drugs.”378 This policy outlook, which has disproportionately impacted
people of color,379 exaggerates the risks posed by those arrested
on drug charges. The statements of the senators drafting the BRA
reflect this bias. Senator Orrin Hatch, for example, “without citing
151

➔

any evidence, contended that bail reform in drug cases was needed
because drug offenders ‘historically abused bail proceedings more
than most criminal offenders.’”380 Senator Lawton Chiles echoed this
point, explaining that “a presumption of detention would establish
‘special considerations’ for bail in drug cases that would ‘go a long
way toward closing that revolving door and would help assure that
drug dealers who are arrested are brought to justice.’”381
Deputy Attorney General James Knapp raised alarm bells at the time,
foreshadowing that the drug presumption would apply even to those
“without any prior record or evidence of criminal history.”382 But
Congress failed to heed his warning.383
Forty years later, DOJ’s prescient concerns have been borne out
in practice, especially as drug-related prosecutions increased as a
percentage of the overall federal docket. Drug crimes are now the
most common federal offenses. They make up the largest share of
the federal criminal caseload,384 and their share of that docket has
increased by 50% since the enactment of the BRA in 1984.385 This
escalation in drug cases means an increase in presumption cases,
as the presumption applies in most drug cases. A recent study by
Professor Stephanie Holmes Didwania found that “the presumption
was responsible for a 19 percentage-point increase in the probability
of detention” after controlling for other variables.386
Today, not only does the presumption of detention apply more
broadly and inflexibly than intended, but it creates racial disparities
as well. Didwania finds that “Black and Hispanic men are more
likely to be subject to presumptive detention than similarly situated
white men.”387 The presumption falls disproportionately on people
of color, as they are more likely to face charges that fall under
the presumption’s domain.388
Understandably, the Judicial Conference of the United States has
expressed concern that the rise in presumption cases is “unnecessarily increasing the detention rates of low-risk defendants,
particularly in drug trafficking cases.”389 The Judicial Conference
relied on a study finding that the presumption increased detention
orders by 26% for people in the lowest Pretrial Risk Assessment
(PTRA) category.390 The same study found that detention rates in
gun presumption cases climbed from 66% in 1995 to 86% in 2010.391
These findings strongly suggest that judges place undue weight on
the presumption of detention.

152

Although Congress intended the presumption of detention to
apply to only the most serious of cases, there is no evidence that
people in presumption cases pose any greater risk than those in
non-presumption cases of either being rearrested for any offense,
rearrested for violent offenses, failing to appear, or revocation of
pretrial release.392 In many cases, in fact, individuals charged in
presumption cases are actually at a lower risk of absconding than
those in non-presumption cases.393
Thus, the presumption of detention has largely outgrown its intended
purpose, and judges should do their best to lessen its impact.

The Constitution, BRA, and Case Law Provide Several Checks
on the Presumption.
The law furnishes safeguards that can prevent the presumption
from having an outsized effect on pretrial detention. In particular,
the presumption is subject to two checks that should guard against
the over-incarceration of low-risk individuals: (1) the presumption
is rebuttable and places a minimal burden of production on the
arrestee, and (2) an unrebutted presumption alone is insufficient
to order detention.394
Serious constitutional issues arise when judges disregard these
checks and treat the presumption like a de facto order of detention.
To comport with due process, judges must recognize that the
presumption is rebuttable and must give arrestees a meaningful
opportunity to rebut the presumption in practice.395
Because “[f]reedom from bodily restraint has always been at the core
of the liberty protected by the Due Process Clause,”396 the Supreme
Court has held that infringements on liberty must be “narrowly
tailored to serve a compelling state interest.”397 Applying this
heightened scrutiny to the bail context, Salerno “upheld the constitutionality of a bail system where pretrial defendants could be detained
only if the need to detain them was demonstrated on an individualized basis.”398 But when judges treat the BRA’s presumption as
“irrebuttable,” they deprive arrestees of their due process right to an
“individualized evaluation,” which inevitably “results in the deprivation of liberty even when not necessary.”399 If a judge fails to treat
the presumption as rebuttable, individuals “are categorically denied
bail based solely on their status”—not based on an individualized
determination of their risk of flight or dangerousness.400 This runs
afoul of the Constitution.
153

➔

The presumption of
detention has largely
outgrown its intended
purpose, and judges should
do their best to lessen
its impact.

154

Fortunately, the two checks on the presumption help judges
avoid these constitutional concerns. The first check is that the
presumption is rebuttable and the defense bears only a “light
burden of production” to rebut the presumption.401 As the Seventh
Circuit explained in United States v. Dominguez, the presumption is
“‘rebutted’ when the defendant meets a ‘burden of production’ by
coming forward with some evidence that he will not flee or endanger
the community.”402 “Any evidence” at all that comes within the
broad scope of § 3142(g) meets the low bar for rebuttal, “including
evidence of [the arrestee’s] marital, family and employment status,
ties to and role in the community, clean criminal record and other
types of evidence encompassed in § 3142(g)(3).”403 Courts recognize
that this standard for rebutting the presumption “is not a heavy
[burden] to meet.”404 The Fifth Circuit has held that the arrestee’s
ties to the community, standing alone, definitively rebut the
presumption: “Where the defendant has presented considerable
evidence of his longstanding ties to the locality in which he faces
trial, as did [this defendant], the presumption contained in § 3142(e)
has been rebutted.”405
Therefore, in every presumption-of-detention case, it is incumbent
upon the judge to assess whether evidence has been introduced that
rebuts the presumption. The defense need only produce evidence or
facts; the law does not require the defense attorney to argue specifically that those facts “rebut” the presumption as a matter of law.
Once “some evidence” has been produced, the judge must make
a legal determination about whether the evidence in the record
is sufficient to “rebut” the presumption. The Seventh Circuit has
reversed a lower court’s pretrial detention determination because
“the judge did not hew to [this] statutory framework.”406 Since any
“evidence of economic or social stability” meets the defense’s burden
and rebuts the presumption,407 in most presumption cases the judge
should conclude that the presumption has been rebutted.
The second important legal check on the presumption is that “an
unrebutted presumption is not, by itself, an adequate reason to order
detention.”408 Both the Fifth and Seventh Circuits have held that
the presumption of detention, standing alone, cannot support a
detention order.409 If the presumption alone could justify detention,
“there would be no need for Congress to have specified ‘the weight
of the evidence against the person’ as a separate factor for the court
to consider.”410 Instead, courts must consider the presumption
“together with the factors listed in § 3142(g),” even if the presumption
has not been rebutted.411
155

➔

In reaching the ultimate release or detention determination, the
judge must remember that, even in a presumption case, “the burden
of persuasion always rests with the government.”412 That is, the
defense never has to persuade the judge that there exist conditions
of release that will reasonably assure the accused’s appearance
and the safety of the community, let alone guarantee their client’s
compliance.413 Instead, it is always the prosecutor’s burden to
show, by “clear and convincing evidence,” that no such conditions
exist.414 When the prosecution requests detention on the basis of
dangerousness, it is the prosecutor’s burden to “prove[] by clear and
convincing evidence [both] that [the] arrestee presents an identified
and articulable threat to an individual or the community”415 and that
no conditions of release would mitigate that particularized threat.
Notably, the judge must never expect the defense to “‘rebut’ the
government’s showing of probable cause to believe that [the accused]
is guilty of the crimes charged,” nor “demonstrate that [the type of
crime charged] is not dangerous to the community.”416
As a practical matter, these two checks come into play at a Detention
Hearing in a presumption case because § 3142(f)(1) authorizes
temporary detention at the Initial Appearance and the holding of a
Detention Hearing in every presumption case. Adherence to these
procedural checks would prevent the presumption from promoting
unnecessary pretrial detention.

The Law Requires Judges to Follow Two Steps in
Presumption-of-Detention Cases.
In connection with these two checks on the presumption, case
law sets forth two legal steps that a judge must follow in every
presumption case.
At Step 1, the judge should determine whether the presumption has
been rebutted, ideally stating their finding on the record. Under the
law, as long as the defense presents “some evidence” that weighs in
favor of release—including any evidence of ties to the community—
the judge should conclude that the presumption has been rebutted.417
While the law requires the defense to produce facts, it does not
require the defense to make the legal argument that those facts
“rebut” the presumption. Rather, in every presumption case, the

156

judge must evaluate the record as a whole to determine whether the
presumption is rebutted as a matter of law. In reaching that determination, the judge should ideally examine the Pretrial Services Report
for evidence that rebuts the presumption.
At Step 2, after a judge has decided whether the presumption has
been “rebutted,” they must then determine whether detention is
warranted after considering the other pretrial release factors listed
in § 3142(g).418 The ultimate question is whether the prosecution has
carried its burden of proving that there are no release conditions that
would reasonably assure both (1) this particular arrestee’s appearance
in court, and (2) by clear and convincing evidence, that this particular
arrestee does not pose “an identified and articulable threat” to a
specific individual or the community.419 Importantly, the judge must
consider all of the § 3142(g) factors whether or not they found the
presumption rebutted at the first step.420

157

➔

C

The Practice: Federal Judges Misapply the
Presumption of Detention.
Even though judges have the power and legal responsibility to limit
the impact of the presumption of detention, they seldom do. Instead,
our research shows that judges routinely ignore the legal checks that
the BRA provides and give the presumption of detention more weight
than the law allows. This appears to be a nationwide problem, as our
courtwatching data were supported by interviews with stakeholders
in many additional districts. This misuse of the presumption causes
many more people to be incarcerated pending trial than necessary
and results in stricter conditions of release in the rare cases where
people obtain release. These impacts also contribute to racial
disparities in the federal criminal system, since people of color face
presumption charges more often than white arrestees.

Judges Give the Presumption of Detention More Weight Than
the Law Allows.
Our study revealed that, too often, judges base their detention
decisions on the presumption instead of the person. In deciding
whether pretrial detention is appropriate, the law requires judges
to engage in an individualized assessment of whether there are
any conditions of release that will reasonably assure the person’s
appearance in court and the safety of the community, even when the
presumption of detention applies. In practice, however, judges rarely
look past the relevant offense to decide whether the individual in fact
poses a flight risk or danger to the community. Instead, they treat the
presumption as a de facto detention order.
Our courtwatching data show that arrestees were detained at much
higher rates in presumption-of-detention cases than non-presumption cases. The presumption of detention applied in 58% of the
cases we observed. Within those presumption cases, arrestees faced
a 72% detention rate. This detention rate exceeded the detention
rate for non-presumption cases by nearly 20%. See Figure 19. This
is consistent with the disparity Didwania found in her study.421

158

Our study revealed that,
too often, judges base their
detention decisions on the
presumption instead of
the person.

159

Detention rates grew even higher when we confined our analysis
to Detention Hearings. Arrestees facing a presumption were
detained 87% of the time when the prosecutor explicitly invoked the
presumption during the Detention Hearing, compared to 76% of the
time when they did not.

Figure 19: Arrestees Are Detained More Frequently in
Presumption Cases Than in Non-Presumption Cases.

The presumption of
detention applied
in 58% of the cases
we observed.

•••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
•••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••
••••••
••••••
••••••
•••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
•••••

Within those
presumption cases,
arrestees faced a 72%
detention rate .
In non-presumption
cases, arrestees faced
a 55% detention rate .

Our supplemental legal research revealed that this presumption
problem is potentially even more pronounced in other districts. A
team of FCJC researchers led by Professor Judith Miller obtained
every district court judicial opinion written in a presumption case
in the First Circuit from the enactment of the BRA in 1984 through
late 2021.422 The team found that 92% of the 250 written opinions in
presumption cases in that circuit ordered the arrestee detained. Two
districts within the circuit had even higher proportions of detention
orders. In one district, judges detained arrestees in 97% of their
written opinions in presumption-of-detention cases; in the other
district, judges ordered detention in every single written opinion, not
granting release to a single arrestee in a presumption case.423
When judges allow their detention pronouncements to be driven
primarily by the presumption, jailing decisions are not limited to the
cases in which pretrial detention is truly warranted by the law and
the facts. Allowing the presumption alone to control detention determinations means that judges are not adhering to their obligation to
consider the particular individual’s flight risk or danger level, as well
160

as the availability of conditions that would mitigate those concerns
in a way that might “reasonably assure” compliance. Systematically
overvaluing the type of crime charged, at the expense of giving individualized consideration to a person’s particular characteristics,
promotes mass incarceration.

D-

There are many reasons why a judge might mistakenly treat the
presumption as a mandate for pretrial detention. Perhaps judges
misunderstand the presumption and how it should impact their
analysis. Several Defenders we spoke with expressed this view. One
Defender said judges “definitely don’t apply” the proper legal standard
in presumption cases. “They definitely nod their heads and don’t say
you’re wrong,” the Defender continued, “and then they don’t [apply
the law]. So I don’t know, maybe they get it and they don’t care, but
they sure as heck are not applying it correctly.” Another Defender
lamented that it is “sort of a waste of time to do all the prep [for the
Detention Hearing in a presumption case] because you know it’s a
fait accompli.” In judges’ defense, some scholars also misunderstand
the presumption and improperly place a burden of persuasion on the
defense.424
Other Defenders expressed specific complaints with how judges
apply the rebuttal standard (Step 1 of the legal test). One explained
that judges do not understand that the burden of persuasion always
rests with the prosecutor, and routinely shift it to the defense, in
violation of the law. Another noted that judges conflate the rebuttal
standard with the burden of persuasion: “[T]hey’re not going to say
that it got rebutted, and then not release.” Yet another Defender
summarized their experience as follows: “I think you’re just so far
behind the eight ball when you start in a presumption case that it’s
really, really hard. The odds of you getting someone out if you got a
presumption case are vastly lower.”
Our courtwatching data confirm that judges are not following the
law in presumption cases. Consider, for example, the requirement
at Step 1 that the presumption need only be rebutted by “some
evidence.”425 As explained above, this standard for rebuttal is meant
to be easy to satisfy.426 Yet in 95% of presumption cases that held a
Detention Hearing in our study, judges either concluded that the
arrestee had failed to rebut the presumption or did not mention
whether the presumption was rebutted.427 See Figure 20.
Our data also suggest that judges often disregard Step 2 of the legal
test, where they are supposed to weigh the rebutted or unrebutted
161

presumption along with all of the other pretrial release factors in
reaching the ultimate release or detention determination. Of the
cases where judges did not affirmatively find the presumption
rebutted, 81% resulted in detention. And at every single Detention
Hearing where a judge definitively found that the presumption had
not been rebutted, the judge detained the arrestee. See Figure 20.
These numbers are surprising given that almost any evidence
can rebut the presumption, and an unrebutted presumption

Figure 20: Judges Are Not Following the
Law in Presumption Cases.

000000

The presumption applied in 113 Detention Hearings.

••••••••••
••••••••••
•••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••

81%ofthese
cases resulted
in detention.

At Step 1, in 95% of
Detention Hearings in
presumption cases,
the judge did not find
the presumption
rebutted (107/113) .

••••••••••
..
I

When the Judge exphc1tly found that the
presumption was not rebutted, 100% of
these cases resulted in detention.

does not mandate detention.
When we asked a judge from a district where we had courtwatched why it is so rare for courts to find the presumption of
detention rebutted, the judge explained: “I don’t understand [the
presumption]. . . . I really don’t think judges, including this judge,
even though I did some research into trying to understand what it
means—what does it mean? What do you need to provide to rebut
it? I don’t think that’s litigated enough. I think that it needs to be
brought to the court’s attention more often to know what standard
applies.” This sentiment was echoed in a Detention Hearing we
observed, in which the presiding judge remarked: “Candidly, I don’t
know whether the presumption is rebutted or not. Case law is
murky as to what level of evidence is required to rebut . . . . I don’t
know if that is met here. So, I’ll just base this [detention order]
on preponderance of evidence that there is no condition that
would prevent flight.”428

162

These statements elucidate why only 5% of judges found that
the presumption of detention was rebutted at the Detention
Hearing: judges don’t understand the legal standard for rebutting
the presumption, and some unfairly put the onus on the defense
to explain it.
In fairness to judges, defense counsel argued that the presumption
was rebutted or rebuttable in less than 20% of presumption cases
where the prosecutor or judge mentioned the presumption. Additionally, the defense waived the Detention Hearing, stipulated
to detention, or otherwise did not contest detention in 55% of
the presumption cases we observed, as compared with 26% of
non-presumption cases.
However, at contested Detention Hearings where judges did not
affirmatively find the presumption of detention rebutted,429 judges
detained arrestees at a rate of 78%. Therefore, the exorbitant pretrial
detention rates we observed in presumption cases are primarily
driven by judges misapplying the law. Judges ultimately are required
to make a finding about whether the presumption has been rebutted
at every contested Detention Hearing, regardless of the defense
presentation. Yet nearly every contested Detention Hearing in our
sample lacked this finding.
Another example of the misapplication of the presumption of
detention occurs at the Initial Appearance. Although there is
some overlap between the § 3142(f) factors and the types of cases
that qualify for the presumption of detention under § 3142(e), the
presumption itself applies only at the Detention Hearing stage.
But the presumption was improperly mentioned by a judge in 5%
of Initial Appearances in our study; we also witnessed several cases
in which the presumption was invoked where it did not apply as
a matter of law. Five percent may seem a relatively small number,
but given that the law unambiguously cabins the presumption to
the Detention Hearing, 5% of Initial Appearances is 5% too many.
We also observed 4 cases in which a prosecutor or a judge invoked
a presumption at the Detention Hearing where, as a matter of law,
no presumption applied.430
Judges may also give the presumption of detention too much weight
because they are risk averse and fear adverse consequences.431 Judges
“retain [an] incentive to be too cautious releasing people pretrial,”432
and that incentive can loom especially large in presumption cases.
A Chief Federal Defender in one district where we courtwatched
163

expressed this view. “I think sometimes the presumptions are used
to get to a place that [judges] think they want to get to, and I think
there’s a risk averse-ness, sort of,” the Defender noted. “It’s safer, it’s
easier for them to detain. And I don’t think they totally understand
the magnitude of the impact that they’re having—or maybe they
don’t care—but I think they very much worry if they release and if
something bad happens—I think that is triggering. . . . [J]udges are
more worried about taking a risk and the fear that they might get
burned.” A judge we spoke with discussed these types of misgivings:
“[I]t’s always in the back of my mind in a case like that. That, you
know, this is a presumption case, and am I doing the right thing
if I am deciding that all of the other information in front of me is
suggesting the person should be released?”
Finally, judges may feel compelled to overweigh the presumption of
detention because, in presumption cases, the statute ties detention
to the type of offense, rather than to individual characteristics.
For example, the drug presumption applies to any drug offense
that carries a maximum term of imprisonment of ten years or
more.433 This means that “[v]irtually all federal drug offenses subject
individuals to these heightened detention standards . . . without
regard to a person’s role, culpability in the offense, or lack
of prior convictions.”434
A Chief Federal Defender we interviewed discussed the unfair challenges the generalized nature of the presumption poses. “It’s not an
absolute that [presumption case arrestees are] going to be detained,
but it’s much harder and really in sort of an unfair way,” he remarked.
“It’s just . . . the cases aren’t worse than so many other cases that are
non-[presumption], and you’ve got this uphill battle that you have
to fight with it. So, it’s a lot harder. It’s a problem.” One judge put
the point even more explicitly: “I just don’t think [the presumption]
is worth anything. Especially [since] there’s such a wide variety of
drug and gun cases, etc., from that list, and the fact that everyone’s
presumed detained just seems absurd. Why would you do that to
people? . . . I can’t even really wrap my mind around that reasoning.”

164

When Judges Do Release Presumption Arrestees, They Impose
Harsher Conditions than in Non-Presumption Cases.

D-

Even in the rare case when an arrestee facing a presumption of
detention was released in our study, they faced more stringent
conditions of release than their counterparts. For example, the
median bond amount in presumption cases was much higher than
the median bond for non-presumption cases; in presumption cases,
it was $250,000 at the Detention Hearing, which is double the
non-presumption median of $125,000.435 Thus, the presumption
appears to result in harsher conditions of release as well as
higher detention rates.
Practitioners from districts beyond the ones where we courtwatched confirmed that judges impose harsher release conditions
in presumption cases. One Chief Federal Defender we spoke with
noted that, “[o]n drug cases, if you have someone to vouch for the
client, . . . we’ve got [a] pretty good shot at getting people out.” But she
continued by noting that the arrestees “might be [placed] on GPS
[monitoring], which can be a real hassle. [The judges] love GPS—
that’s like the new thing that makes them feel good about what
they’re doing.” Another Defender described a similar experience.
“If we win [release in a presumption case],” she remarked, “what
we will win is a high bond with [financial] conditions, you know,
requiring co-sureties.” The Defender characterized this change as
“[t]he difference between a flat-out detention, and . . . a detention
because they can’t meet the conditions.” In this way, the presumption
continues to promote the culture of detention even when the arrestee
apparently overcomes it.

The Presumption of Detention May Contribute to Racial
Disparities in the Federal System.
The burden of the presumption of detention is not borne equally by
all arrestees. Using sentencing data as a proxy, we estimate that most
arrestees facing a presumption are people of color, since they make
up 75% of those convicted of qualifying drug offenses nationwide.436 In
this study, the proportion of people of color was even higher: 89% of
presumption cases involved people of color, while only 11% involved
white arrestees.437 And in our Chicago pilot study, 94% of presumption
cases involved people of color.438

165

The presumption is especially pernicious for people of color because
the offenses to which it is tied can dovetail with racial disparities
and stereotypes. One judge spoke candidly about this problem.
“[A]nother type of case we get in our district a lot are . . . so-called
‘gang cases,’” the judge remarked, “and I think there’s a lot of racism
involved in calling someone a gang member and then arguing about
their dangerousness. Because as soon as you label someone that way,
the thought that comes into . . . a judge’s mind is, ‘Oh, this person is
violent. They’re living the kind of lifestyle of violence and committing
crimes and that type of thing.’ . . . [O]ne of the things I think is really
important to do is not to have things devolve into these types of
meaningless labels.”
Our data support the idea that the presumption may fall even
more heavily on arrestees of color. People of color received different
treatment from prosecutors when facing a presumption charge. Of
the cases in which prosecutors invoked the presumption during the
Detention Hearing, 97% of the arrestees were people of color, and 3%
were white. See Figure 21.

Figure 21: People of Color Receive Different
Treatment When Facing a Presumption Charge.

Of the cases in which
prosecutors invoked the
presumption during the
Detention Hearing, 97°/o
of the arrestees were
people of color, and 3%
werewhite.

■■

•

Latino

■■■
■■■

■■■
■■■

•

■■■
Black
■■■■

••••
••••

■■■ G-White

166

In addition, judges detained people of color at higher rates than white
individuals: the detention rate in presumption cases involving people
of color was 73%, while the detention rate in presumption cases
involving white arrestees was just 68%. See Figure 22. In our Chicago
pilot study, every arrestee detained in a presumption case was a
person of color.439

D-

Figure 22: The Presumption of Detention May Contribute
to Racial Disparities in the Federal System.

Judges detained
people of color at
higher rates than
white individuals in
presumption cases .

•••••••
••••••••••
••••••••••
••••••••••
••••••••••
•••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
0000000000
00000
0000000

The detention rate
involving people of
color was 73% .

•i•hi·- @:)@ii
••Hr\3-

And 68% for white
individuals.

Finally, when we observed judges and prosecutors making serious
legal mistakes regarding the presumption, the arrestee was always
a person of color. As mentioned above, we observed two cases in
which a prosecutor invoked a presumption at the Detention Hearing,
despite a presumption not applying as a matter of law—and in
both cases, the arrestee was Black. (In one of those cases, the judge
erroneously affirmed the prosecutor’s claim that a presumption of
detention existed.) Relatedly, we observed two other cases in which
the judge invoked a presumption where none applied by law. In both
of those cases, the arrestee was Black.

167

D

The Solution: At the Detention Hearing,
Judges Must Adhere to the Low
Standard for Rebutting the Presumption
of Detention and Never Treat the
Presumption as a Mandate for Detention.
Judges can mitigate the presumption of detention’s ill effects by adhering
to the letter of the law. Both the scholarship and the data show that the
presumption has largely outgrown its intended purpose and its usefulness.
Congress may take action to rein in the presumption at some point.440 Courts
may even be forced to confront the presumption’s constitutionality. Until then,
judges must “hew to the statutory framework” in presumption cases, an approach
that will also insulate their decisions from reversal.441
In every presumption case, judges should follow this two-step process:
At Step 1, the judge should make a finding about rebuttal. As a matter of law, if
the defense has presented “some evidence” of the arrestee’s history and characteristics (such as ties to the community, family ties, or employment) or some
evidence that mitigates the circumstances of the offense, the judge should find
the presumption rebutted.442 If the defense has not presented any evidence,
the judge should examine the Pretrial Services Report to determine if there is
any evidence that “rebuts” the presumption. Following this first step in every
presumption case will limit the damage the presumption causes, minimizing its
weight from the outset—as Congress intended.
At Step 2, the judge should consider all of the § 3142(g) factors together, viewing
the presumption as, at most, just one factor in the mix—and never treating the
presumption as the determinative factor.443 The judge should recognize that the
existence of a rebutted—or even an unrebutted—presumption does not limit
their discretion to release. The rebuttable presumption must never operate like a
mandatory minimum or a mandate for detention.
By following these simple legal rules, judges can prevent unnecessary detention
and shift the culture of detention.

168

Our courtwatching data
confirm that judges are
not following the law in
presumption cases.

169

4

Judges Must Stop
Unlawfully Jailing People
for Poverty Through
Excessive Financial
Conditions.

Our courtwatching study shows that federal judges
consistently impose financial conditions of release
that result in pretrial detention. This practice violates
the explicit statutory language of the Bail Reform Act,
perpetuates a system where wealth buys release and
people are jailed for poverty, and has a disproportionate
racial impact. These detentions, which violate the law,
contribute to rising detention rates as well as racial and
socioeconomic disparities in the federal system.
Judges must bring the practice back in line with the law
by ceasing to impose financial conditions that arrestees
cannot meet.

THE STATUTE: 18 U.S.C. § 3142(c)(2)
The judicial officer may not impose a financial condition that results in the pretrial
detention of the person.

170

Our study found that
federal courts regularly
impose inequitable and
burdensome financial
conditions, perpetuating
a system in which money
buys freedom and poverty
ensures incarceration.

171

A

Takeaways

• In 37% of cases in our study, judges imposed
financial conditions, such as personal surety bonds,
corporate surety bonds, and the upfront posting of
collateral. See Figure 23.

37%

• In 34% of all cases and 91% of cases where financial
conditions were imposed, judges required arrestees
to post a secured bond, reintroducing the evils of cash
bail systems that the BRA sought to avoid.
See Figure 23.

34%

• In one district where we courtwatched, arrestees
were detained in 40% of cases involving financial
conditions solely because they did not have the money
to pay for their release. See Figure 25. Judges in that
district regularly imposed federal bail bonds known as
corporate surety bonds (CSBs). In 92% of cases where
a CSB was imposed, the accused was locked in jail
because they were unable to obtain a bail bond.
Every single individual subjected to a CSB was a
person of color. See Figure 24.
• Across all 4 districts, arrestees did not have the
money to meet financial conditions in 36% of cases
where such conditions were imposed. See Figure 23.
In fact, 21% of all arrestees detained at the Initial
Appearance remained in jail after the Detention
Hearing because they could not meet financial
conditions of release. For these individuals, the
financial conditions acted as de facto detention orders,
in violation of the law.

172

40%
92%

36%
21%

p

Racial Disparities
in the imposition of financial conditions of release

95%

• Black and Latino arrestees were much more likely
to face financial conditions of release than white
arrestees; among arrestees on whom secured bonds
were imposed, 95% were people of color. See Figure 26.

95%
173

B

The Law: The Bail Reform Act Prohibits
Financial Conditions of Release that
Result in Detention.
The BRA unambiguously prohibits judges from imposing financial
conditions of release that people are unable to meet. The statute
states concisely: “The judicial officer may not impose a financial
condition that results in the pretrial detention of the person.”444
The BRA contemplates no exceptions or loopholes to this
requirement. And yet, courts regularly evade, pervert, and ignore
this unambiguous provision.
Judges are allowed to require conditions of release only when an
unsecured bond will not reasonably assure an arrestee’s appearance
and community safety.445 In addition, a judge may not impose
financial conditions of release to address safety concerns; such
conditions may only be used to secure someone’s appearance in
court.446 Judges are therefore prohibited from issuing “a detention
order without a proper finding of risk of flight or danger to the
community” and from “granting bail but setting an exorbitant
financial condition that the defendant [cannot] meet.”447
The BRA thus strictly cabins the use of financial conditions, authorizing them only when necessary to assure appearance—not safety—
and limiting them to arrestees who are solvent enough to meet
the financial obligation imposed. The BRA allows judges to require
arrestees to “execute a bail bond with solvent sureties” only when
such a condition is reasonably necessary to ensure appearance.448
Courts have interpreted this statutory provision to authorize two
different types of financial conditions: (1) federal bail bonds known
as corporate surety bonds (CSBs), and (2) personal surety bonds
commonly referred to as “solvent surety” bonds. Congress intended
that judges would reserve both bail bonds and solvent surety bonds
for the rare case,449 to avoid detaining people simply because
they are poor.450
The legislative history of the BRA shows that the prohibition on
financial conditions that result in detention was of central concern
to the drafters.451

174

The BRA unambiguously
prohibits judges from
imposing financial
conditions of release that
people are unable to meet.

175

The BRA strictly cabins the
use of financial conditions,
authorizing them only
when necessary to assure
appearance—not safety—
and limiting them to
arrestees who are solvent
enough to meet the financial
obligation imposed.

176

As the main Senate report notes, the “primary purpose of the [1966
BRA] was to deemphasize the use of money bonds in the federal
courts, a practice which was perceived as resulting in disproportionate and unnecessary pretrial incarceration of poor defendants.”452
The Senate report emphasizes that one “concern [animating the
enactment of the BRA] . . . was the argument that stringent financial
conditions of release . . . would be used to avoid the limitations and
procedural requirements” for pretrial detention.453 The committee
noted that “bail procedures often result in pretrial detention through
the arbitrary use of high money bail,” requiring “restrictions on the
use of financial conditions.”454
In fact, one of the reasons the 1984 Act authorized preventive
detention on dangerousness grounds was to stop judges from
“strain[ing] the law, and impos[ing] a high money bond ostensibly
for the purpose of assuring appearance, but actually to protect the
public.”455 The Senate committee explained, “[T]he pretrial detention
provisions of Section 3142 are to replace any existing practice of
detaining dangerous defendants through the imposition of excessively high money bond.”456
Accordingly, the BRA was supposed to “preclud[e] pretrial detention
through the use of high money bond.”457 Federal courts recognized
this prohibition almost immediately after the BRA’s passage, with
the Eighth Circuit stating that one of the “major differences” from
previous federal bail laws was “the [BRA’s] prohibition against using
inordinately steep financial conditions to detain defendants.”458
Other circuits followed suit.459
As a practical matter, the BRA’s prohibition means that “[i]f the
judicial officer determines that some amount of money will assure
the appearance of the defendant, then he must select an amount that
is attainable.”460 Critics of state-level cash bail systems have long
touted the BRA’s limitations on financial conditions as a salutary
and egalitarian aspect of the federal bond regime.461
Supreme Court jurisprudence also states that excessive financial
conditions are antithetical to the BRA and the Constitution.
In Stack v. Boyle,462 the Court held that judges violate the Eighth
Amendment when they impose bail “at a figure higher than an
amount reasonably calculated [to ensure the arrestee’s appearance
at trial].”463 In Salerno, the Court corroborated Stack’s holding, stating
that “when the Government has admitted that its only interest is
in preventing flight, bail must be set by a court at a sum designed to
ensure that goal, and no more.”464
177

➔

Accordingly, financial conditions of release that result in an arrestee’s
detention are always excessive: either there are no conditions of
release that will reasonably assure the person’s appearance, in which
case a court can order confinement outright, or there exist conditions
that will assure appearance, in which case the arrestee must be
released.465 A judge who detains someone at the Initial Appearance
by setting financial conditions that the arrestee cannot meet blocks
that person from availing themselves of the procedural protections of
a Detention Hearing.466
If an arrestee does not qualify for detention, the BRA commands
that the judge “shall order the pretrial release of the person on
personal recognizance” unless doing so “will not reasonably assure
the appearance of the person as required or will endanger the safety
of . . . the community.”467 If the judge determines that conditions
are needed to ensure these goals, the statute requires the “least
restrictive . . . condition”468—including the least restrictive financial
condition(s).469 Most importantly, while a judge may choose from
a menu of least restrictive conditions that will reasonably assure
appearance, a financial condition that the person cannot meet is
never an option. Judges have a variety of tools at their disposal to
reasonably assure appearance without using financial conditions,470
which inherently “privilege the wealthy over the poor.”471
To ensure that financial conditions of release are not excessive,
and therefore do not result in unlawful detention, judges must
consider the individualized circumstances of each arrestee. The
BRA’s legislative history illuminates this point, as the Judiciary
Committee “emphasized that all conditions are not appropriate to
every defendant,” and instructed “the judicial officer [to] weigh each of
the discretionary conditions separately with reference to the characteristics and circumstances of the defendant before him and to the
offense charged, and with specific reference to the factors set forth in
[§ 3142](g).”472 Federal courts concur, with the Eighth Circuit reversing
a release order that applied the same condition of release to all individuals in a multi-defendant case.473 Individualized determinations
also mitigate against “over-conditioning,” or imposing conditions of
release that are more burdensome than what is required to satisfy
the purpose of the BRA.474

178

Financial conditions of
release that result in an
arrestee’s detention are
always excessive: either
there are no conditions of
release that will reasonably
assure the person’s
appearance, in which case a
court can order confinement
outright, or there exist
conditions that will assure
appearance, in which
case the arrestee must be
released.
179

C

The Practice: Federal Judges Impose
Excessive Financial Conditions that
Violate the Law and Jail People
for Poverty.
Our study found that federal courts regularly impose inequitable
and burdensome financial conditions, perpetuating a system in
which money buys freedom and poverty ensures incarceration.475
In practice, judges frequently disregard the BRA’s bright-line rule
against predicating release on someone’s ability to pay.
Judges rely too heavily on secured bonds. In some courts, judges
purport to release indigent arrestees by imposing high corporate
surety bonds, knowing full well that the person will never be able
to obtain the bond and will therefore remain in jail. This undue
emphasis on financial conditions of release exacerbates racial
disparities within the federal pretrial system. Additionally, our data
show that financial conditions result in detentions that violate the
BRA, as some arrestees are too poor to meet the conditions imposed.
We assessed financial conditions imposed at the Initial Appearance,
at the Detention Hearing, and overall across both hearings.

Judges Are Overly Reliant on Secured Bonds.
We found that judges rely heavily on financial conditions.476 Across
our entire dataset, judges imposed financial conditions in 37% of
cases. See Figure 23. Cases involving financial conditions comprised
68% of the subset of released cases and 21% of the subset of detained
cases. Judges imposed secured bonds with alarming frequency,
using them in 91% of cases involving financial conditions across the
Initial Appearance and Detention Hearing stages. See Figure 23. In
11% of cases involving financial conditions, judges imposed cash bail
requirements (secured bonds requiring the upfront posting of cash).

180

Secured bonds come in many different forms—they include cash
bail, the upfront posting of collateral, the upfront posting of property,
personal “solvent surety” bonds, and corporate surety bonds obtained
from a bail bond company.

D-

A personal solvent surety bond requires the arrestee to find a relative
or friend with financial resources willing to bind themselves legally
to paying the bond if the arrestee flees.477 In many cases, imposing
a solvent surety bond violates the BRA, in that it is a financial
condition that results in detention. But defense attorneys often agree
to this condition in an effort to secure their clients’ release. When
this happens, it is incumbent on judges to uphold their statutory
responsibility to impose the least restrictive conditions—even if that
means disagreeing with both defense counsel and the prosecution
and imposing an unsecured bond instead.

Figure 23: Judges Impose Burdensome Financial
Requirements and Jail People for Poverty.

In 37% of cases, judges imposed
financial conditions (127/343) .
In 91% of cases where financial
conditions were imposed, judges
required arrestees to post a
secured bond (116/127).

Arrestees did not have the
money to meet financial conditions in 36% of cases (46/127) .
For these individuals, the
financial conditions acted
as de facto detention orders,
in violation of the law.

•
••••••••••
••••••••••
••••••••••
••••••••••
•••••••••••
•••••••••
••••••••••
••••••••••
••••••
••••••••••
••••••••••
••••••••••
••••••••••

91%

181

One especially troubling type of secured bond in the federal system
is the corporate surety bond (CSB). When a CSB is purchased, the
bail bond company signs a contract with family members or friends
and charges a nonrefundable fee—which in federal court is typically
between 10% and 15% of the full bail amount.478 The co-signers
are often required to post property or other assets as collateral,
sometimes in amounts totaling 150% of the bail amount.479 If the
arrestee violates the terms of release or fails to appear in court, the
person who signed the contract is responsible for paying the full bail
amount and forfeits the collateral.480
According to a review done by Color of Change and the ACLU, many
bail bond contracts include additional fees and conditions that are
difficult for the arrestee and their co-signers to meet.481 Many people
are left with loan installments and fees well after a case is resolved.482
Bail bond companies are often predatory and charge extremely high
interest rates.483 Indigent people must then choose between staying
in jail or taking on significant debt to secure their freedom.484
Most federal CSBs involve Nebbia conditions, another restrictive
condition of release.485 When a court orders Nebbia conditions, the
arrestee or the relative/friend posting bond is required to prove
that the funds for both the premium and the collateral come
from legitimate, legal means. Although a judge has determined
that the arrestee is otherwise entitled to release, the individual
will remain in jail until the arrestee has met their burden of proof
at a Nebbia hearing.
The use of CSBs results in higher rates of pretrial detention
because indigent people are unable to obtain such bonds. In our
study, CSBs were used in 14% of all cases (and only in one district,
the Southern District of Florida (Miami)). In 92% of the cases where
a CSB was imposed, the arrestee was ultimately detained pretrial
(after both the Initial Appearance and the Detention Hearing). See
Figure 24. At the Initial Appearance, 100% of cases involving a CSB
ultimately resulted in pretrial detention; at the Detention Hearing,
70% of cases involving a CSB resulted in the detention of the arrestee.
In contrast, when an unsecured bond was imposed, only 10% of
people were detained at the Initial Appearance, and no one was
detained at the Detention Hearing.

182

Figure 24: The Use of Corporate Surety Bonds Results in Higher
Rates of Pretrial Detention and Exacerbates Racial Disparities.

■■■■
■■■■
■■■■■■■■■■
■■■■■■■■■■
■■■■■■■■■■
■■■■■■■■■■

D-

In 92% of the cases where a
CSB was imposed, the
arrestee was ultimately
detained pretrial.
All those on whom CSBs were
imposed were people of color.

We observed the pervasive use of CSBs in Miami, and detention
outcomes in that district demonstrate conclusively the problems
with CSBs. Fully 40% of cases in Miami involving financial
conditions result in pretrial detention solely because the arrestee is
unable to meet financial conditions of release. See Figure 25. CSBs are
so common in Miami that the court has an official stamp that reads:
“Both sides stipulate to a $250,000 CSB with Nebbia, reserving the
right to a PTD hearing at a later date.”

Figure 25: Judges Jail Arrestees Who Are
Too Poor To Pay For Their Release.

••••••••••••
••••••••••••
••••••••••••
••••••••••••
•••••••••••
•••••••••••
•••••••••••
•••••••••••
•••••••••••
•••••••••••

40% of cases in
Miami involving
financial conditions
resulted in pretrial
detention solely
because the arrestee
was unable to meet
financial conditions
of release .

183

In our interviews with judges and Defenders, we confirmed that
federal judges across the country rely upon CSBs and cash bonds,
in derogation of the statute. One Defender from a large western
district where we did not courtwatch told us: “[T]he biggest obstacle
[to getting our clients released] is our magistrate court not setting
[financial] bonds that are a reasonable expectation for the clients to
meet.”486 As this Defender explained, in such cases “[the condition of
release] is either detention or a cash or corporate surety bond. And
then, who’s going to post that bond—that cash?” In that district,
the posting of funds must be done by the arrestee themselves or
a relative. Similarly, in a large district in the southwest where we
did not courtwatch, Defenders told us that judges often require
the upfront posting of cash, which prevents release in many cases.
Contrary to the legal standard under § 3142(c)(2), these conditions of
release often result in jailing on the basis of an arrestee’s poverty.
Through interviews with Defenders in districts where we did not
courtwatch, we also uncovered the pervasive use of the so-called
“solvent surety” condition. An arrestee must find a family or friend
to serve as a personal surety, meaning that person agrees to pay the
monetary amount of the bond if the arrestee absconds.
Requiring an indigent individual to locate a personal solvent surety
who is willing to post or co-sign a money bond often forecloses the
possibility of release for indigent arrestees—in violation of the BRA.
In districts where this practice is common, Defenders told us that
solvent sureties are required to provide numerous forms of personal
and financial information, establish their net worth, and show proof
of employment and recent paystubs.487 Contrary to the legal standard,
judges in these districts require signatories “to show actual assets,
that need[] to meet [the bond] amount.” As one Defender in a large
western district remarked, this requirement frequently “put[s] poor
people in jail.”
Even if an arrestee can secure a solvent surety, this condition
often prolongs the amount of time an arrestee spends in jail. The
process of verifying the proposed surety is both burdensome and
time-consuming. Additionally, one Defender from a large district in
the southwest where we did not courtwatch told us that prosecutors
will at times rescind their release recommendation while verification
is pending. The verification process is so prolonged in that particular
district that the Defender recalled cases in which their clients were
ordered released and yet locked in jail for 45 days as they awaited the
verification of their solvent surety.
184

It is little wonder that Defenders from across the country told us
that if they could change one thing about their district’s pretrial
processes, they would eliminate the solvent surety requirement.

D-

While conditions like electronic monitoring and GPS monitoring
may seem like alternatives to financial conditions of release, they
may still carry a high pecuniary burden for arrestees. When a judge
orders electronic monitoring, a device is attached to the releasee’s
ankle and connected to their phone. If the person leaves their home
without authorization, an alarm will alert Pretrial Services/Probation.
Arrestees are often required to pay a daily fee to the private
companies that provide this technology. This cost may be prohibitive,
forcing some people to stay in jail.488
In our study, judges ordered electronic monitoring in 42% of the cases
where an arrestee was released. Several Defenders whom we interviewed labelled electronic monitoring as the most overused condition
of release. One Defender explained that electronic monitoring was
so commonly imposed in their district that they were constantly
“running out of monitors” and having to delay release for arrestees
until “when we think we’ll have [a monitor] freed up.” One judge we
interviewed speculated that electronic monitoring is so frequently
used because “it seems like something punitive you can do to the
person while they’re out.”489

Excessive Financial Conditions Have a Disproportionate
Racial Impact.
Saddling indigent arrestees with high bonds aggravates racial
disparities in the federal system given discrepancies in household
wealth across racial groups.490
People of color are less likely to meet financial requirements, as
institutionalized racism in the form of housing, tax, financial, and
education policies has resulted in highly uneven distributions
of wealth and property ownership.491 Black Americans own just
one-tenth of the wealth of white Americans, and this wealth gap
“persists regardless of households’ education, marital status, age, or
income.”492 Although the average Black American household owes
one-third of the debt owed by the average white American household,
the interest rates for Black borrowers are noticeably higher.493
Research conducted in the Second Circuit, for example, found that
33% of white arrestees in that federal court were homeowners,
compared to just 7% of Black arrestees and 9% of Latino arrestees.494
185

Against this backdrop, it is striking that in 95% of cases in our study
where a secured bond was imposed, the arrestee was a person of
color.495 See Figure 26. Moreover, in every single case in our study
where a CSB was imposed, the arrestee was a person of color.496
See Figure 24.

Figure 26: Excessive Financial Conditions of
Release Have a Disproportionate Racial Impact.

000000
• • • • • • • • • • 95% of secured
• • • • • • • • • • bondswere
• • • • • • • • • • imposed against

••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
• • • • • • • • • •

people of color.

Financial Conditions Contribute
to Unlawful Detention.
Our findings reveal on-the-ground practices that make a mockery
of the BRA’s prohibition against financial conditions that result
in detention.
Shockingly, during our courtwatching, arrestees were detained
because they could not meet financial conditions of release in 13%
of all cases we observed, and in more than one-third—36%—of
the cases where financial conditions were imposed. See Figure 23.
Of the people who were detained in our study, 21% were detained
for the duration of the case because they could not meet financial
conditions of release.
Our study further revealed that federal judges are fully aware that
indigent arrestees cannot pay exorbitant financial conditions such as
CSBs. When setting a $250,000 CSB, one judge said to the probation
officer, on the record:497

186

22

So you have expressed a concern that he should not be

23

released on this bond, which it's my experience nobody ever

24

meets this bond and they remain detained. But we will make it

25

a matter of record that this Defendant may not be released on

1

D-

this tentative bond without further hearing.

During an interview, another judge confirmed that CSBs almost always
result in detention, describing $250,000 CSBs as “a bond they’re never
going to make, obviously. Most people can’t make that bond.”
Imposing a financial bond that a person cannot make is, by definition,
unlawful: a judge violates the BRA every time they “impose a financial
condition that results in the pretrial detention” of an arrestee.498 One
judge we interviewed stated that when they impose a bond that an
arrestee can’t meet, “I don’t feel bad about that necessarily if I think the
bond was reasonable.” But a bond that an arrestee cannot meet is, by
its very nature, unreasonable and illegal.
Because many people charged with federal crimes are indigent,
financial conditions of release run a serious risk of serving as de
facto detention orders, an indisputable violation of the BRA.499 Our
interviews suggest that this is the case in many districts across the
country. One Chief Federal Defender told us that financial conditions
in their district are so unreasonable for arrestees that they are
“tantamount to detention.” A Chief Federal Defender from another
district likewise stated that financial conditions “set[] people up for
failure or keep[] them in jail.”
In every case, judges must keep the BRA’s prohibition on financial
conditions that result in detention paramount, carefully considering
necessity and reasonableness before imposing any type of financial
condition. But many judges act on autopilot, without attention to
whether a given condition would be overly burdensome for a particular
individual. This may be a consequence of prosecutors or Pretrial
Services frequently requesting financial conditions of release. Overall,
prosecutors requested financial conditions of release in 29% of the
cases we observed (22% of Initial Appearances and 12% of Detention
Hearings). Regardless, judges must have the courage to follow the law
and depart from requested conditions that would prove onerous or
unreachable for the person standing in front of them.
187

Our interviews revealed a perception that judges do not carefully
consider a person’s financial ability before imposing financial
conditions. One Defender we interviewed observed, “Pretrial [Services]
will want to load your client down [with conditions] . . . . If the attorney
doesn’t speak up, the judge will just go with it.” A Chief Federal
Defender from another district recalled a case in which they argued
that imposing the prosecutor’s requested financial conditions would
violate the BRA because the arrestee was too poor to make bond and
would end up jailed for poverty. Rather than consider the arrestee’s
indigency, the judge “waited for the prosecutor to withdraw the
[financial] request” before altering the conditions of release.
During courtwatching, we witnessed the direct connection
between burdensome financial conditions and pretrial detention.
We watched one case where 3 fishermen from Cuba—who did not
speak English and were charged with nonviolent drug offenses—were
ordered released on CSBs of $250,000 each, with Nebbia conditions.
But none of the arrestees was able to afford the bond, so all remained
detained pending trial.
Judges sometimes explicitly told arrestees that financial conditions
served as de facto orders of detention. For example, when ordering a
$250,000 corporate surety bond with a Nebbia condition, one Miami
judge told the defense attorney:500

22

The language just explicitly states what would actually happen

23

if your client ever met the $250,000 corporate surety with a

24

Nebbia condition bond, which the likelihood of that is in my

25

educated guess zero.

These examples highlight how judges further the culture of detention
by imposing onerous financial conditions, turning purported release
orders into mandates for detention.
Our results further show that defense lawyers rarely advocate against
financial conditions of release. In cases where financial conditions
were imposed at the Initial Appearance, the defense argued against
188

those conditions just 2% of the time (2 cases total). In those 2 cases,
both of the arrestees were white. At the Detention Hearing, the
defense contested financial conditions of release in only 3% of the
cases in which financial conditions were imposed. See Figure 27.

D-

Figure 27: The Defense Bar Rarely Contests
Financial Conditions of Release.

• • • • • • ■- In cases where

••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
••••••••••
• • • • • • • • • •

• • • • • • • • • •
• • • • • • • • • •

financial conditions
were imposed,
defense attorneys
contested financial
conditions only 3%
of the time (4/127) .

Our interviews indicate that the defense may shy away from
contesting financial conditions because they are willing to agree to
just about anything to keep their clients out of jail. As one Defender
put it, “if he’s getting out, I mean Jesus, let him out. I’ll take what
I can get. And if that’s what you need to let him out, [then] let him
out.” Another Defender concurred, noting “You want to get somebody
out, and your client’s family owns a house, and they’re . . . willing to
put [the] house up, . . . you’re not raising the issue.” This Defender
explained, “you just go to the Court and say, ‘we’ve got property, judge,
if you want to use this as collateral on a bond.’ . . . [W]e’re doing it
because we’re trying to get our clients out.” However, this disempowered trend from the defense bar does not excuse the conduct of
judges. The BRA requires judges to make an independent determination as to the appropriateness of a bond and a person’s financial
ability to meet the bond, and far too often judges impose a financial
condition without such a determination.

189

D

The Solution: At Both Pretrial
Hearings, Judges Must Stop Imposing
Financial Conditions that Result
in Detention and Tailor Release
Conditions to Each Arrestee’s Individual
Economic Circumstances.
As frontline decisionmakers tasked with applying the law as
written, judges must heed the BRA’s instruction to impose the
least restrictive conditions of release. To comply with the statute,
judges should also cease imposing financial conditions of release
that people cannot meet. The only way to fully adhere to the statutory
prohibition against financial conditions that result in detention is to
entirely stop imposing financial conditions on indigent individuals.
Given that the vast majority of those charged with federal crimes are
indigent,501 many federal courts in this country need to dramatically
reduce their reliance on financial conditions. Judges should reserve
financial conditions of release for the rare case where the arrestee has
the economic means to pay for their freedom.
As a practical matter, judges should pay close attention to each
arrestee’s personal financial circumstances and fashion individualized conditions of release that are reasonable for that specific
person. Before imposing a financial condition, a judge should
carefully examine whether the particular arrestee before them has
the means and the wherewithal to meet that condition. The judge
should also evaluate whether a financial condition might impermissibly result in the arrestee’s detention. In making these individualized determinations, judges must be mindful of how financial
conditions compound existing racial and socioeconomic disparities.

190

Unless judges are vigilant, they will continue violating the BRA by
allowing the statute’s narrow authorization of financial conditions to
swallow its rule against detaining indigent individuals via financial
conditions that they cannot meet. In federal districts where it has
become commonplace to impose financial conditions of release,
often without regard for the arrestee’s economic means, judges will
need to resist the entrenched culture and refuse to impose financial
conditions proposed by the prosecution and Pretrial Services/
Probation. Judges should take a hard look at all of the financial
conditions that make it difficult—and often impossible—for
indigent arrestees to secure release.

*

Given the evils of cash bail, the federal system should unequivocally abjure corporate surety bonds, a truly retrograde vestige of
the past. Federal judges should also think carefully before imposing
any kind of secured bond, including requiring a relative to serve
as a solvent surety.
By applying the BRA as written, judges can avoid detaining people for
poverty and can mitigate racial and socioeconomic disparities within
the federal system.

191

192

CONCLUSION
193

When our Clinic first embarked on this study, we were surprised
to learn that many federal judges, prosecutors, and defense
attorneys were completely unaware of the federal bail crisis
unfolding in their midst. Stakeholders consistently told us that
the federal pretrial process was working just fine. Some even
questioned why we would choose to spend time and energy
investigating a system that was functioning as it should. Those
stakeholders were largely oblivious to skyrocketing federal
jailing rates and were astonished to learn how rare it is for
people to flee or be rearrested after being released pretrial—
demonstrating that the federal system could release far more
people pending trial without jeopardizing community safety or
the efficient administration of justice.

However, Federal Criminal Justice Clinic faculty and students knew
from our own experiences litigating federal pretrial hearings that the
system was in crisis. We also knew that the brunt of that systemic
breakdown was being borne by people of color, poor people, and
people who have no powerful lobby to challenge such injustices.
We knew that, absent intervention, mass detention would continue
unabated in the federal system. To begin to change this culture, a
catalyst was needed.
We hope this Report serves as that catalyst by providing the first
data-driven examination of the federal bail crisis. This Report
describes a pervasive culture of detention within the federal system.
Through courtwatching and interviewing stakeholders around
the country, we uncovered a freewheeling system in which judges
have allowed misguided courtroom habit to supersede the law.
During Initial Appearance hearings, we watched federal judges lock
poor people in jail without lawyers and detain others in violation of
the statute. During Detention Hearings, we watched judges ignore
the rebuttable nature of the presumption of detention. And we saw
judges purporting to release indigent people while instead burdening
them with unattainable financial conditions, thus violating the law
and impermissibly blocking any hope of freedom.

194

This Report also shows how such deviations from the written law
exacerbate racial and socioeconomic disparities in the criminal
legal system. Prosecutors cited improper grounds for detention at
Initial Appearances more frequently when the arrestee was a person
of color than when they were white, and people of color were far
more likely to be illegally detained at the Initial Appearance than
white arrestees. People of color were also significantly more likely
to be deprived of counsel at their Initial Appearance hearing than
white arrestees. At Detention Hearings, prosecutors overwhelmingly
invoked the presumption of detention against people of color—on
several occasions doing so when no presumption existed as a matter
of law. Consequently, people of color were detained at higher rates
in presumption cases than white arrestees. Additionally, the vast
majority of secured bonds were imposed against people of color,
and only people of color were subject to bail bonds (corporate surety
bonds). While each of these findings is independently concerning,
taken as a whole, they provide a damning illustration of pretrial
practices aggravating racial disparities in the system.
Our Clinic’s research has already persuaded judges to reconsider
their practices, making us optimistic that this Report will encourage
further reforms.
After the FCJC’s pilot study in Chicago, we observed promising
changes in federal bail processes. One judge told us that, as a result
of our study, judges and practitioners in Chicago had developed “a
heightened sensitivity to the whole question of whether a Detention
Hearing may take place at all under § 3142(f).” Prior to the FCJC’s
intervention, federal prosecutors in Chicago cited a valid basis for
detention in just 5% of Initial Appearances. But after we intervened
with the United States Attorney’s Office and shared our findings with
the court, federal prosecutors cited a valid statutory basis in 82% of
Initial Appearances.502 Additionally, after training the defense bar, we
observed a nearly 40% decrease in the rate at which defense attorneys
consented to pretrial detention.503
Our interventions have likewise catalyzed a culture shift in
courtrooms across the country. One judge we interviewed observed
that our efforts to educate stakeholders about the BRA are “having
a huge impact across the country, and it’s sort of chipping away at
all this and winning people over.” That judge continued by saying
that, after attending one of our bail presentations via the Federal
Judicial Center, all of the federal magistrate judges in their district

195

convened to discuss the legal standard that applies during the Initial
Appearance, and concluded that it had been a “weak spot in [their]
analysis.” The entire court decided to rectify the problem by asking
prosecutors “to give some factual basis” for their detention requests
at the Initial Appearance, especially in non-(f)(1) cases, where the risk
of unlawful detentions is greatest. Another judge told us: “Detention
rates are coming down, and much of that is attributable to you and
your efforts. You have made a real difference in how the federal
judiciary sees release and detention.” Our interventions are also
changing the way federal public defenders approach bond issues. One
Chief Defender told us that our research “is forcing [the defense] to
be better lawyers . . . make better arguments,” and “get courts to listen
to those arguments.”
By complying with the law and requiring federal prosecutors to
follow suit, federal judges can ensure that every single person
who comes before them is protected by the BRA’s substantive and
procedural safeguards. Our Report lays out 4 key recommendations
to ensure adherence to the statute and the Constitution. First,
to prevent unlawful detentions at the Initial Appearance, judges
must adhere to the limitations in § 3142(f). Second, to prevent
arrestees from being jailed without lawyers, judges must appoint
counsel to represent every indigent arrestee for the entirety of their
Initial Appearance hearing. Third, in presumption-of-detention
cases, judges must recognize the low standard for rebutting the
presumption and never treat the presumption as a mandate for
detention. Finally, at both pretrial hearings, judges must strictly
limit the use of financial conditions of release and account for each
person’s individual financial circumstances to prevent arrestees from
being jailed for poverty.
Fear may inform many of the harms we identify in this Report. It is
incumbent upon judges to act boldly and to be guided by data, not
institutional pressures. As one judge we interviewed noted, “if people
aren’t violating your bond, then you’re not letting enough people out.
The statute says ‘reasonably assure’ safety and appearance. That
phrase means there’s an acceptable risk that’s being assumed by the
system. If I’m not feeling the impact of that risk, I’ve got to recalibrate
and release more people.”

196

We hope that our Report empowers federal judges to continue
shifting the culture that has brought the federal system to this
calamitous point. One judge we interviewed characterized federal bail
hearings as “the intersection of our education crisis . . . , our mental
health crisis, our drug crisis, and race in America.” Especially given
those interactions, judges must resist the institutional pressures
and implicit biases that have led the system to prioritize detention
and must return to the presumption of release enshrined in the BRA
and affirmed in Salerno. Ultimately, federal judges have the power
to uphold the rule of law, to make detention prior to trial the rare
exception, and to be champions of liberty.

197

198

APPENDIX A:
BACKGROUND
& METHODOLOGY
199

During the 2018–2019 academic year, the Federal Criminal
Justice Clinic (FCJC) began investigating the cause of high federal
pretrial detention rates and identified a “federal bail crisis.”504
Following the success of the FCJC’s Chicago Pilot Study in
2018–2019, Professor Alison Siegler launched a national Federal
Bailwatching Project (FBP) in fall 2020 to further investigate that
crisis. This Report is the culmination of that project.

200

1

Chicago Pilot Study
(2018–2019)
In 2018, Professor Siegler, Professor Erica Zunkel, and
their students in the FCJC embarked on the first courtwatching project ever undertaken in federal court,
observing bond hearings in the United States District
Court for the Northern District of Illinois in Chicago.505
Professors Siegler and Zunkel published the findings
from this pilot project in the National Association of
Criminal Defense Lawyers’ Champion journal, in an
article titled Rethinking Federal Bail Advocacy to Change
the Culture of Detention.506 The Chicago courtwatching
project served as a pilot for the national study described
in this Report.
The Chicago pilot consisted of two primary phases.507
During Phase 1, law students in the FCJC and volunteers
from across the University of Chicago gathered and
coded data from 173 bail-related hearings in federal court
in Chicago.508 The FCJC discovered that prosecutors
frequently requested detention for unlawful reasons, and
that in nearly 10% of cases, arrestees were detained in
violation of the law.509 The FCJC also uncovered troubling
racial disparities. After completing the data-gathering
portion of Phase 1, the FCJC shared its findings with
the U.S. Attorney’s Office and federal magistrate judges
in Chicago.510 In Phase 2, which began in mid-2019, the
FCJC conducted additional courtwatching, which showed
that prosecutors, judges, and defense attorneys were all
adhering more closely to the BRA in the aftermath of the
FCJC’s interventions.511 During interviews, stakeholders
in Chicago said that our pilot project changed how judges,
prosecutors, and the defense bar approach the federal
pretrial detention process by highlighting how in-court
practices had diverged from the legal standards.

201

2

The FCJC’s
Advocacy Efforts
From 2018 to the present, the FCJC has engaged in
a multi-pronged effort to combat the federal bail
crisis, educate stakeholders and the public, and align
on-the-ground practices with the written law. These
interventions have brought the federal bail crisis into
the national consciousness:
• Legislative Advocacy: In November 2019, Professor Siegler
testified before the House Judiciary Committee about the need
for reform in the federal bail system.512 Professor Siegler’s oral
testimony513 and accompanying written testimony514 highlighted the most troubling aspects of the federal bail crisis and
proposed reforms to the BRA that would both clarify the legal
standards and revivify the norm of liberty that had motivated
the statute.515 Professor Zunkel played a critical and central
role in the Clinic’s legislative advocacy work. Many of the
reforms advocated in the testimony and in Professor Siegler
and Zunkel’s work were embodied in the Federal Bail Reform
Act of 2020, introduced by Chairman Jerrold Nadler (D-NY) as
a replacement for the Bail Reform Act of 1984.516
• Policy Advocacy: In November 2020, Professors Siegler and
Zunkel contributed to the call for federal bail reform in a
Justice Roundtable Report.517
• Executive Branch Advocacy: In December 2020, the FCJC
submitted a memo to the Biden Administration discussing
the need for federal bail reform and next steps for the Administration,518 and Professor Siegler met with members of the
Biden Transition Team.

202

• Public Advocacy: In February 2021, Professor Siegler and Kate
Harris, a student in the FCJC, published an op-ed in the New
York Times titled How Did the “Worst of the Worst” Become Three
Out of Four?519 This piece urged the Biden Administration and
Attorney General Merrick Garland to “disrupt the culture of
detention that pervades the ranks of federal prosecutors and, to
some degree, the federal judiciary.”520
• Training federal judges and other stakeholders: Since
2018, Professor Siegler has given speeches and trainings for
hundreds of federal judges, hundreds of probation officers, and
thousands of Federal Public Defenders and CJA lawyers.

203

3

Federal Bailwatching
Project (2020–2022)

In 2020, Professor Siegler launched the national
Federal Bailwatching Project (FBP). The centerpiece
of the FBP is the courtwatching study detailed in
this Report. In that study, FCJC students and others
gathered, coded, and analyzed data from 4 federal
courts and mined the federal docketing system for additional data about the courtwatched cases. The 4 federal
courts where we gathered courtwatching data were
the Southern District of Florida (Miami), the District
of Massachusetts (Boston), the District of Maryland
(Baltimore), and the District of Utah (Salt Lake City).
We also generated qualitative data from many more
federal courts through interviews with stakeholders.
These interviews were conducted in districts throughout
the country. The interviews enabled us to confirm
that the practices we witnessed while courtwatching
were replicated in other courts and revealed additional problematic practices. Interviewing also helped
our team identify best practices that might shift the
culture from one prioritizing pretrial detention to
one prioritizing release.
Professor Siegler created the overarching vision for the
FBP, recruited all collaborators, and supervised and ran
every phase of the project that culminated in this Report.
At the time of this Report’s publication, she and the FCJC
continue to pursue legislative and policy advocacy, as well
as stakeholder engagement and training.
204

A

Phase 1: Creation and Courtwatching
After the Chicago pilot, we decided to scale up our courtwatching
model to conduct a broader investigation with a more national
scope. In the fall of 2020, we began to plan the courtwatching
phase of the FBP. Because of the ongoing COVID-19 pandemic, we
were required to conduct our courtwatching entirely virtually, with
courtwatchers “attending” video hearings via Zoom, rather than
entering physical courthouses.
Professor Siegler recruited clinical law professors and students from
5 additional law schools who agreed to collaborate on the project
during this first phase of the courtwatching study. The student courtwatchers took verbatim notes during Initial Appearance hearings and
Detention Hearings and coded their data into standardized online
courtwatching forms.
Throughout the summer and fall of 2020, we engaged in exploratory
conversations with Chief Federal Defenders to set up the 4 courtwatching sites. Through these conversations, we identified a set
of federal courts where it would be possible to access virtual bond
hearings. To gain access to some courts, we or the Chief Defender
communicated directly with the judges. During the summer of 2020,
Professor Zunkel was instrumental in setting up the first courtwatching site, in the District of Massachusetts.
The contributions and insights of Professor Zunkel and the project’s
Technical Advisor, Professor Sonja Starr, provided an essential and
irreplaceable foundation for the entire data-gathering structure of
this study. In the fall of 2020, Professors Siegler, Zunkel, and Starr
created the early drafts of the courtwatching forms and online
surveys that courtwatchers used to gather and record data. Together,
they created “Verbatim Notes Forms” for courtwatchers to record
information as they watched pretrial hearings, with separate forms
for Initial Appearances and Detention Hearings. The 3 professors also
developed an early template of the online surveys into which courtwatchers ultimately coded data about the hearings. In addition, the
professors worked together on early fundraising efforts. Professors
Zunkel and Starr also served as advisors throughout the duration
of the 2-year project, providing invaluable advice and editorial
commentary on documents, protocols, and trainings.
205

Professor Siegler and students in the 2020–2021 FCJC cohort
subsequently created and honed Verbatim Notes Forms for each
of the 4 courtwatching districts.521 They also created an online
Qualtrics survey where courtwatchers answered a series of detailed
questions in order to code various aspects of every Initial Appearance
and Detention Hearing, including the result of each hearing. The
Qualtrics surveys enabled us to transform the notes taken during
court hearings into an analyzable spreadsheet format from which we
generated the findings in this Report. FCJC student Desiree Mitchell,
Professor Starr, Boston Site Manager Aliza Bloom, and Project
Manager Noadia Steinmetz-Silber all made major contributions to
these Qualtrics surveys.
With Professor Siegler’s guidance, FCJC students also created courtwatching manuals, spreadsheets where courtwatchers recorded
hearings from the court calendar and indicated hearings that they
watched, protocols to ensure the smooth transfer of the data gathered
from each courtwatching site to a central repository, FAQs for the
Qualtrics surveys, and other materials tailored to each courtwatching
site. We then prepared and presented courtwatcher training sessions
to standardize practices across all courtwatching sites.
During the 2020–2021 academic year, Professor Siegler launched the
4 courtwatching sites and ran them in collaboration and conjunction
with faculty from other law schools. Boston Site Manager Aliza
Bloom and Project Manager Noadia Steinmetz-Silber contributed
immeasurably to the courtwatching phase of this project. (The
student members of the courtwatching teams are listed in the
Authors & Acknowledgments section.):
• We launched our first courtwatching site on September 8,
2020, in the District of Massachusetts, leading a team that
included Chief Federal Public Defender Miriam Conrad;
Professor Dehlia Umunna, Faculty Deputy Director of Harvard
Law School’s Criminal Justice Institute Clinic (CJI), Clinical
Instructor and Managing Attorney of CJI Lia Monahon, and
then-Clinical Instructor Sadiq Reza of CJI; and Professor
David Rossman, the Director of Boston University School
of Law’s Criminal Law Clinical Programs (BU), as well as
BU Professor Wendy Kaplan and BU Visiting Scholar Aliza
Bloom. Law students from both Boston-based clinics served
as courtwatchers. Professor Siegler conducted trainings for
the Boston-based students before courtwatching began.

206

Bloom became the Boston Site Manager, supervising students
as they monitored court dockets, engaged in courtwatching,
and coded their results. Courtwatching in Boston ran for
5 months, ending on January 11, 2021, and netted a total of
37 relevant hearings in 29 cases.
• We launched our second courtwatching site on November 2,
2020, in the Southern District of Florida, with the help of
Chief Federal Defender Michael Caruso. The FCJC’s own clinic
students served as courtwatchers under Professor Siegler’s
supervision. FCJC students Kate Harris and Sam Bonafede
served as the Miami Site Managers and were joined in early
2021 by Project Manager Steinmetz-Silber. Courtwatching in
Miami ran for 4 months, ending on March 11, 2021, and netted
a total of 342 relevant hearings in 220 cases.
• We launched our third courtwatching site on January 11, 2021,
in the District of Maryland, with the help of Chief Federal
Defender James Wyda and Professor Maneka Sinha, the
Director of the University of Maryland Carey School of Law’s
Criminal Defense Clinic. Mr. Wyda secured the agreement of
U.S. Chief District Court Judge James Bredar of the District
of Maryland to enable the project to proceed. Following
those conversations, Professor Sinha served as the Baltimore
Site Manager, recruiting her clinic students and additional
volunteer students enrolled in criminal procedure courses
separate from the clinic to ensure that the site would be
adequately staffed. Courtwatching in the Northern Division
of the District of Maryland (Baltimore) launched on January
11, 2021, and ran for 14 weeks until April 15, 2021. Thirteen
students observed a total of 107 relevant hearings in 63 cases.
• We launched our fourth courtwatching site on January 11,
2021, in the District of Utah, with the help of First Assistant
Federal Public Defender Robert Steele. Professor Jon Powell,
the Director of Campbell Law School’s Restorative Justice
Clinic and Monica Veno of that clinic were the Utah
Site Managers; their clinic students served as the courtwatchers. Courtwatching in Salt Lake City ran for 4 months,
ending on March 11, 2021, and netted a total of 60 relevant
hearings in 31 cases.

207

In each district, courtwatchers worked pre-scheduled shifts and
were responsible for observing all hearings during their shift. This
ensured that the data collected by the courtwatchers captured as
many hearings as possible. The hearings covered a range of charges,
with a variety of federal magistrate judges presiding. The professors
responsible for each courtwatching site played an integral role by
supervising their clinic students throughout the courtwatching and
coding phases.
While observing remote hearings, courtwatchers used the Verbatim
Notes Form templates for Initial Appearances and Detention
Hearings. In Part 1 of the Verbatim Notes Form, courtwatchers
recorded detailed information about the particular case they were
watching, including the name of the arrestee, the names of every
participant, the case number, and the charges; courtwatchers also
answered questions about the presence or absence of counsel. In
Part 2 of the Verbatim Notes Form, courtwatchers took word-for-word
notes on their laptops, writing down everything said by every
participant in the hearing. To ensure accuracy, FCJC students subsequently spot-checked Verbatim Notes Forms against transcripts in a
subset of cases.
To obtain race data, courtwatchers recorded in the Verbatim Notes
Form the demographics of everyone in the courtroom based on
their observations of each person’s physical appearance, noting the
basis for that assumption in the form itself. Because demographic
information such as race and gender were not available from the
docket sheet or any other public source, we relied on courtwatchers’
perceptions and judgments. We recognized that courtwatchers might
be uncomfortable with being asked to record observable race, and we
provided a space in the Verbatim Notes Form for courtwatchers to
convey any such concerns.
Courtwatchers also recorded in the Verbatim Notes Form
whether an arrestee was identified as a United States citizen
during the hearing.522 This information allowed us to code differential practices and outcomes between citizens and noncitizens
in subsequent phases.
Prior to launching the Maryland and Utah courtwatching sites,
Professors Siegler and Bloom, along with FCJC students Bonafede
and Harris, trained volunteers on both substantive law and
courtwatching procedures. To complete their training, student
observers spent a week observing hearings informally before
the official site launch.
208

Due to the dedication of our faculty partners and student courtwatchers, our study ultimately gathered information from over
600 federal bond hearings. Because we were interested in observing
the entire pretrial detention and release process in every case, we
eliminated cases where courtwatchers had not observed the Initial
Appearance. This narrowed our dataset to the 343 relevant cases
discussed in this Report. In each of those cases, we observed the
Initial Appearance as well as the subsequent Detention Hearing, if
one was held, for a total of 536 hearings.

209

B

Phase 2: Data Coding and Analysis
The data coding and analysis phase of this study ran from spring
2021 through summer 2022. The project’s Technical Advisor,
Professor Starr, played a pivotal role in the data coding and analysis
phase. In 2021, Data Managers Morgen Miller and Dylan Baker of the
Coase-Sandor Institute of Law and Economics523 joined this project
and made many notable contributions.
Upon finishing the courtwatching phase of the project in spring 2021,
courtwatchers completed an online Qualtrics survey for each hearing
that they had observed, drawing on the information in their Verbatim
Notes Forms. These surveys focused on many substantive issues,
including those that we discuss in this Report: the prosecutor’s
stated bases for requesting detention at both hearings, the judge’s
bases for detaining or releasing at both hearings, whether and at
which point in the hearing the arrestee was represented by counsel,
whether and what financial conditions were imposed on released
individuals, and the results of the hearings. Professor Siegler, the
FCJC student team, Bloom, and Steinmetz-Silber created a coding
protocol and fielded courtwatchers’ questions in real time to troubleshoot problems and ensure accuracy. They also spot-checked cases
and re-coded as necessary.
During summer 2021, we obtained additional information for each
observed case by downloading case documents from the federal
court’s Public Access to Court Electronic Records (PACER) docketing
system. Michael Jeung served as the Project Manager in June and
July 2021, training a group of undergraduate interns how to use
PACER and input case data into Excel. Project Manager Jeung
supervised the interns as they gathered and coded data into Excel
from the Docket Sheet, Complaint, Indictment, Case Summary, and
the relevant Detention or Release Order in each of the observed cases.
Project Manager Jeung and undergraduate intern Jacqueline Lewittes
checked all Excel sheets and ensured all relevant PACER documents
were retrieved and properly entered for every case in our dataset.
FCJC student Krysta Kilinski took over as Project Manager in
August 2021, overseeing interns as they continued to clean and
validate our dataset. In fall 2021, Project Manager Kilinski created
Qualtrics surveys to code the results of the PACER research,

210

created a coding protocol, conducted a training on that protocol,
and supervised interns as they coded the data. These Qualtrics
surveys allowed us to track certain outcomes of interest across cases,
including whether the arrestee was detained or released, whether the
prosecutor cited a valid § 3142(f) factor when requesting detention,
and the type of conditions of release imposed.
For cases where a courtwatcher observed the Initial Appearance
but not the subsequent Detention Hearing, we ordered Detention
Hearing transcripts. During summer 2021, undergraduate interns
reviewed PACER dockets and flagged all “unobserved hearings”
to determine all cases that required a Detention Hearing transcript.524 This list was subsequently validated by Baker and Project
Manager Jacqueline Lewittes. Using the validated list, Lewittes
also engaged in the laborious process of ordering transcripts from
the 4 courtwatched districts, which FCJC students then coded into
additional Qualtrics surveys.
To analyze the data gathered through courtwatching, FCJC students
worked closely with Professor Starr, Coase-Sandor’s Baker, and
Technical Research Assistants Jacob Jameson and Dhruv Sinha.
First, the Qualtrics survey response spreadsheets were cleaned.
This process consisted of removing duplicates, correcting typos,
generating new variables, and standardizing identifier variables,
such as arrestee name, case number, and city. Next, the spreadsheets
were merged by matching the identifier variables, creating a master
spreadsheet that tracked all data for each case from both the Initial
Appearance and the Detention Hearing—if one was held. Most
of the data analysis was then completed using R, a widely used
statistical programming language.
Additionally, some manual calculations were performed by reviewing
the full data spreadsheet and Qualtrics response spreadsheets in
Excel. Professor Starr and FCJC students also coded every type of
crime charged in our data set. This, in turn, illuminated which cases
were technically eligible for detention at the Initial Appearance
under § 3142(f), as well as which cases qualified for a presumption of
detention at the Detention Hearing.

211

Our study’s methodology was limited in a couple of important ways.
First, given the relatively small size of our dataset, our Report relies
on summary statistics, rather than regression analyses. This was a
limitation created by the number of hearings that occurred during
our courtwatching schedule and the limitations the academic
calendar placed on the number of weeks during which students were
available to courtwatch. Second, because Pretrial Service Reports
are not made public, we lacked verifiable information regarding
arrestees’ criminal history.
Throughout the 2021–2022 academic year, FCJC students Clare
Downing and Stephen Ferro served as the Project Managers, while
FCJC student Angela Chang served as a Data Manager.525 In the fall
of 2021, Downing, Ferro, Chang, and Kilinski conducted and led an
entirely new phase of data coding and analysis that yielded far more
nuanced results. Chang’s role included incorporating into our data
analysis the results of these supplemental Qualtrics surveys, which
recorded additional information and corrected errors. Chang also
engaged in data reconciliation and cleanup, determined the appropriate variables to be used in calculating each metric, and manually
reviewed data spreadsheets to calculate the more complex metrics.
In addition, from 2021–2022, Baker, Jameson, and Dhruv Sinha spent
countless hours writing code in R and STATA to clean, merge, and
analyze our data set.

212

C

Phase 3: Qualitative Interviews
To understand and elucidate the national scope of the bail crisis
unfolding in federal courts throughout the country, we supplemented our quantitative courtwatching data with qualitative
interviews of stakeholders in many districts beyond the 4 districts
where we had courtwatched. We interviewed 48 federal magistrate
judges and federal public defenders from 36 federal courts
across 11 federal circuits.526
The goal of our interviewing project was to determine whether
the problems revealed by our courtwatching study were occurring
nationwide. Through these interviews, we confirmed that the federal
bail crisis was truly national in scope, and we learned more about the
freewheeling and varied nature of pretrial bail proceedings in federal
courts. Excerpts from these interviews appear throughout the Report
as supplemental findings to our quantitative results. To ensure
candor, we promised interviewees that we would de-identify them,
would not list their names in this Report, and would not attribute
quotes to them directly.
The interviewing project began in fall 2021. Led by Interview Team
Leaders Krysta Kilinski and Ale Clark-Ansani, FCJC students first
selected a geographically diverse pool of federal district courts that
included rural, urban, suburban, and border districts. We included
the 4 districts where we had courtwatched, then added districts that
had particularly high rates of detention, as well as some districts
with average or particularly low rates of detention. Our aim was to
identify and understand any problems that were leading to high
detention rates, as well as best practices that mitigated the culture
of detention. FCJC students compiled a spreadsheet of stakeholders
to contact and developed interview questions. Professor Siegler then
recruited the interviewees.
In spring 2022, Kilinski, Clark-Ansani, and Professor Siegler
interviewed 26 judges and federal defenders (mostly Chief Federal
Public Defenders or their second-in-command) who practiced in
13 federal district courts. With the interviewees’ approval, we recorded
interviews via Zoom and uploaded them to Box. In the few instances
where interviewees asked not to be recorded, we took detailed notes to
shore up our recollection.
213

We asked a standard set of questions of each interviewee, with
different question sets for magistrate judges and federal defenders. As
the interviewing process progressed, we added new questions sparked
by our prior interviews. When interviewing stakeholders from our
4 courtwatching districts, we added unique questions related to our
courtwatching data and observations. Project Manager Jacqueline
Lewittes then transcribed all of the interviews and analyzed them
according to key themes, including changes observed during the
COVID-19 pandemic, adherence to the legal standard during Initial
Appearances, practices relating to counsel at Initial Appearances,
adherence to the legal standard in presumption-of-detention cases,
and the use of financial conditions of release. Kilinski worked with
Lewittes to identify salient vignettes to highlight in the Report.
We were able to interview a federal magistrate judge and the Chief
Federal Public Defender in 3 of the 4 courtwatching districts
(Maryland, Massachusetts, and the Southern District of Florida). In
the fourth district—the District of Utah—we interviewed a defender,
but the federal magistrate judges we contacted in Salt Lake City did
not consent to be interviewed.
In August 2022, Professor Siegler conducted interviews of 17 additional Chief Federal Defenders and 5 Assistant Federal Defenders
to determine specifically how access to counsel at the Initial
Appearance operated in numerous federal districts. These interviews
provided a much more comprehensive and complex picture of the
national access-to-counsel crisis described in this Report.

214

D

Phase 4: Creating this Report
During the 2020–2021 academic year, FCJC student Report Team
Leaders and Primary Authors Kate Harris and Sam Bonafede led
the research and writing of a first draft of this Report. During the
2021–2022 academic year, FCJC student Report Team Leaders and
Primary Authors Clare Downing and Stephen Ferro led the research
and writing of subsequent drafts. During summer 2022, Project
Manager Lewittes led the writing and editing of the final Report draft,
with essential contributions by FCJC students Emma Stapleton and
Sebastian Torrero, as well as intern Ethan Ostrow. In August 2022,
FCJC students Jaden Lessnick and Clare Downing spearheaded the
final phase of the writing and editing process.
In the spring and summer of 2022, we partnered with Voilà:, an information design company. Professor Siegler, Data Manager Chang, and
Project Manager Lewittes identified and revalidated key data points,
which Voilà: visualized and integrated into the Report design. Additionally, Project Manager Lewittes oversaw communications between
the FCJC and Voilà: teams, including preparing full annotated drafts,
communicating revised datapoints and new visualization requests,
and relaying design feedback to Voilà:.

215

216

APPENDIX B:
DISTRICT-SPECIFIC
DATA COMPARISONS
217

1

Detention
Rates
Figure 28 shows that detention rates varied at the
Initial Appearance, at the Detention Hearing, and
overall, across the districts where we courtwatched.
Across all 3 metrics, Salt Lake City had the highest
detention rates and Baltimore had the lowest.
“Overall” numbers depict what happened across the entire bond
phase, during the Initial Appearance and during any Detention
Hearing. Specifically, the cases categorized as detained “overall” are
cases where the arrestee was detained at the Initial Appearance and
no Detention Hearing was held, as well as cases where the arrestee
was detained at the Initial Appearance and remained detained after
the Detention Hearing.

Figure 28: Detention Rates
100%

Salt Lake City
Miami
Boston
Baltimore

218

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

At Initial Appearance

63%

76%

78%

97%

At Detention Hearing

75%

78%

76%

83%

Overall

54%

62%

66%

81%

2

Types of Cases
by District

Figure 29: Proportion of Non-(f)(1) Cases, Presumption
Cases, and Drug Cases in Each District

Non-(f)(l) cases

Presumption cases

Drug cases

-

-

.

53%
43%

-

-

43%

29%

38%
47°/o

27%
33%

-

49%

.

-

51%

-

28%

34%

Miami

Boston

Salt Lake City

Baltimore

219

3
A

Race

Race of Arrestees in Each District

527

Figure 30 shows how the racial composition of arrestees varied
across districts. Baltimore had the highest proportion of Black
arrestees, while Miami and Salt Lake City had the highest proportion
of Latino arrestees. In all districts, people of color comprised a
large majority of arrestees.

Figure 30: Race of Arrestees in Each District528
0%

20%

40%

60%

80%

100%

Miami

II
■
♦Eh++

220

Salt Lake City

I■

Baltimore

■

Boston

Other I Unknown

White

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Black

76%

34%

34%

6%**

Latino

3%**

24%*

58%

52%

Other

2%**

-

<1%**

3%**

White

19%

34%*

6%

32%

Unknown

-

7%**

1%**

6%**

B

Detention Rates by Race
Looking across both hearings and all 4 districts, the overall detention
rate data did not show notable racial disparities.529 There are two
likely underlying reasons. First, most of the people federal prosecutors charge with crimes are people of color: people of color made up
81% of those charged in federal court in 2019.530 The vast majority of
arrestees in our data set—87%—were likewise people of color.
See Figure 3. Accordingly, there were very few white arrestees in our
study. Second, the small sample size of some racial groups, particularly white arrestees, means that our results may not be representative of the total cases involving white arrestees in a given district.
Despite the data limitations identified above, Figure 31 shows that
detention rates at the Initial Appearance were higher for arrestees of
color than for white arrestees in all 4 districts we observed. Figures
32 and 33 show that trends in detention rates by race varied for the
Detention Hearing and overall. Notably, in Boston, the detention
rate for arrestees of color was consistently higher than that of white
arrestees across all 3 metrics.

Figure 31: Detention Rates Overall, by Race
0%

100%

0%

100%

0%

100%

0%

100%

Boston
Salt Lake City
Miami
Baltimore

t:119

tlll1l·t

tMMlt

White

221

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Black

67%

100%

82%

100%**

Latino

50%**

86%**

77%

100%

Other

100%**

-

100%**

White

50%*

50%*

71%

+-

+-

100%**
90%

Figure 32: Detention Rates at Detention Hearing, by Race
0%

100%

0%

100%

0%

100%

0%

100%

Salt Lake City
Miami
Baltimore
Boston

•=111•

222

t·IMlt

tlll+t

White

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Black

71%

71%*

75%

100%**

Latino

-

100%*

75%

80%

Other

100%**

-

100%**

100%**

White

100%**

50%**

100%*

89%*

Figure 33: Detention Rates Overall, by Race
0%

100%

100%

0%

IIIIIIEJ

~

tum•

tlll+t

0%

--

100%

0%

100%

Salt Lake City
Boston
Miami
Baltimore

t•1 1·11t

White

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Black

54%

80%

68%

100%**

Latino

50%**

86%*

66%

81%

Other

100%**

-

100%**

100%**

White

50%*

30%**

71%

80%

223

4

Prosecutors’
Detention Requests

Figure 34 compares the percentage of cases in which
prosecutors requested detention in each district.
These numbers include cases where the prosecutor did not explicitly
request detention during the hearing, but based on context, appeared
to be seeking detention. Salt Lake City had the highest rate of prosecutorial detention requests across all 3 metrics. In Miami, the rate
of prosecutorial detention requests was artificially low because of
the high number of cases involving corporate surety bonds (CSBs).
In those cases, the prosecutor did not formally seek detention;
however, arrestees were seldom able to meet the CSB condition and
obtain release. Seeking a CSB thus effectively amounts to a detention
request, a fact that was confirmed by a judge during one hearing:
“[N]obody ever meets this bond and they remain detained.”531

Figure 34: Rate of Prosecutors Seeking Detention
Initial Appearance
& Detention Hearing

Initial Appearance

Detention Hearing

0%

0%

0%

100%

100%

100%

Salt Lake City
Miami

-224

--

Boston
Baltimore

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)532

D. Utah
(Salt Lake City)

At Initial Appearance

63%

76%

64%

97%

At Detention Hearing

88%

78%

61%

100%

Overall

60%

62%

45%

97%

5
A

Defense
Behavior

Defense Waiver
Figure 35 shows the rate at which defense counsel in each district
waived the right to seek release, submitted to detention, stipulated
to detention, agreed to detention without prejudice, waived the
Detention Hearing, or otherwise did not contest detention at each
stage, as well as overall (at any point pretrial). Baltimore—which had
the lowest rate of overall pretrial detention—had the lowest overall
rate of defense waiver. See Figure 28.

Figure 35: Rate of Defense Waiver
At any point

At Initial
Appearance

0%

0%

100%

--

100%

0%

I

■

I

100%

0%

100%

--

NA

I
I
I

At Detention
Hearing

In between ~

~

NA
NA

I

Salt Lake City
Boston
Miami

I

Baltimore

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

21%

3%**

I

D. Md
(Baltimore)

D. Ma
(Boston)

19%

14%**

Between IA and DH

13%**

-

-

-

At Detention Hearing

4%**

56%

44%

48%

Overall

25%

48%

45%

48%

At Initial Appearance

+-

+-

225

B

Defense Argument for Release
Figure 36 shows the rate at which defense counsel in each district
made affirmative arguments for release at each hearing. Again,
Baltimore—the district with the lowest detention rate among the
districts we observed—had the highest rate of defense counsel
arguing for release at the Detention Hearing.

Figure 36: Rate of Defense Argument for Release at
Initial Appearance vs Detention Hearing
0%

•
•
•
•

226

50%

100%

•
•

Salt Lake City
Miami

•

Boston

•

Baltimore

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

At Initial Appearance

6%**

3%**

7%

6%**

At Detention Hearing

92%

33%*

39%

45%

6
A

Initial
Appearance

Prosecutors’ Bases for Requesting
Detention at the Initial Appearance
Figure 37 shows the percentage of cases where the prosecutor
cited various grounds when requesting detention at the Initial
Appearance. Some of these grounds were legal bases for detention
contained in § 3142(f), while others were not. Prosecutors in Boston
and Salt Lake City cited to the BRA when requesting detention
at the Initial Appearance at far higher rates than prosecutors in
Baltimore and Miami.

Figure 37: Grounds Cited by Prosecutors When
Seeking Detention at the Initial Appearance

Factor listed
in 3142(f)

Danger,
ordinary risk
of flight

Noncitizen
client, criminal
history, etc.

No specific
grounds stated

0%

0%

0%

0%

--

I
I

100%

100%

I
■

- ■

100%

100%

I
I
I

Salt Lake City

I

Miami

Boston

Baltimore

227

228

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Factor listed in 3142(f)

10%**

77%

2%**

67%

Danger, ordinary
risk of flight

28%

18%**

45%

0%

Noncitizen client,
criminal history, etc.

10%**

5%**

2%**

7%**

No specific
grounds stated

53%

0%

51%

27%*

B

Cases Charging Offenses Not Listed
in § 3142(f)(1) (Non-(f)(1) Cases)
Figure 38 shows, using publicly available federal sentencing
data, the percentage of cases in each district where there is
no § 3142(f)(1) factor present.

Figure 38: Proportion of Non-(f)(1) Cases in Each District533
0%

50%

75%

100%

Miami
Boston
Salt Lake City

I

D. Md
(Baltimore)

% of sentenced cases
where no (f)(1) factor 29%
is present

Baltimore

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

43%

53%

43%

229

Figure 39 shows, using data from our study, the rate at
which prosecutors sought detention in cases where, by law,
no § 3142(f)(1) factor applied.

Figure 39: Rate of Prosecutors Seeking
Detention in Non-(f)(1) Cases

0%

25%

50%

75%

100%

Baltimore
Boston
Miami
Salt Lake City

Rate of detention
requests

230

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)534

50%*

43%**

35%

-

C

Unlawful Detention at Initial Appearances
Figure 40 shows detention outcomes at the Initial Appearance in
cases where prosecutors sought detention, including the rate at which
arrestees were detained unlawfully at the Initial Appearance.

Figure 40: Detention Outcomes at the Initial Appearance,
When Prosecutor Sought Detention
0%

75%

50%

100%

Salt Lake City
-

Boston
Miami

Detained unlawfully

l

D. Md
(Baltimore)

D. Ma
(Boston)

82.5%

91%

Detained
Unlawfully

15%*

Released

2.5%**

Detained

l

Baltimore

Released

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)535

87%

100%

9%**

13%

-

0%

0%

0%

231

7

Counsel at the
Initial Appearance

Figure 41 shows that only Miami held the Initial
Appearances where some arrestees were not represented by counsel.
In Miami, 11% of the arrestees we observed were unrepresented
during the Initial Appearance; 100% of those arrestees were detained
at the Initial Appearance, and 76% were detained for the duration
of the case. See Figure 17. Figure 41 shows a correlation between
detention rates and the extent to which the arrestee was represented
by counsel at the Initial Appearance, with detention rates increasing
as representation by counsel decreases.

Figure 41: Detention Rates at the Initial
Appearance, by Representation
Fully represented

Partially represented

Unrepresented

0%

0%

0%

100%

-Full Representation

232

100%

--

NA

Salt Lake City

NA

Boston

NA

Baltimore
Miami

D. Md
(Baltimore)

D. Ma
(Boston)

62%

73%

I

100%

S.D. Fl
I(Miami)

D. Utah
(Salt Lake City)536

59%

97%

Partial Representation537 73%*

75%**

92%

100%**

Unrepresented

-

100%

-

-

8
A

Cases Involving the Presumption
of Detention in § 3142(e)(3)

Application of the Presumption of Detention
The presumption of detention applied in 67% of Baltimore cases, 52% of
Boston cases, 56% of Miami cases, and 55% of Salt Lake City cases.
Figure 42 compares the proportion of arrestees of color in presumption
cases versus non-presumption cases. In every district, there was a
greater proportion of arrestees of color among the presumption cases
than in non-presumption cases.

Figure 42: Proportion of Cases Involving Arrestees of Color
Presumption case

Non-presumption case

0%

0%

100%

100%
Baltimore
Salt Lake City
Miami
Boston

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

jD.(SaltUtahLake City)

Presumption Cases

83%

67%

96%

71%

Non-Presumption
Cases

76%

64%

91%

64%

233

Figure 43 shows the categories of presumption cases we observed in
each district. Across all districts, the vast majority of presumption
cases were drug cases.

Figure 43: Types of Presumption Cases Observed

Drug Cases

Gun Cases
(924(c))

Trafficking

Minor Victim

0%

0%

0%

0%

100%

100%

100%

100%

Boston

--

234

I
I

I
■

I

■

Miami
Salt Lake City
Baltimore

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Drug Cases

69%

100%

92%

82%

Gun Cases (924(c))

26%

-

7%*

12%**

Trafficking

2%**

-

2%**

-

Terrorism

-

-

-

-

Minor Victim

17%

-

3%**

24%

B

Detention Rates in Presumption Cases
Figure 44 shows detention rates in presumption cases compared to
non-presumption cases in each district. Surprisingly, in Baltimore,
the detention rate in non-presumption cases was higher than in
presumption cases. In Miami, the detention rate for presumption
cases was far higher than the detention rate in non-presumption cases.

Figure 44: Detention Rates in Presumption
Cases vs Non-Presumption Cases

0%

100%

50%

•
•

•
•

• •
• •

Salt Lake City
Miami
Boston
Baltimore

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

Presumption Cases

48%

67%

79%

88%

Non-Presumption
Cases

67%

57%*

50%

71%

235

Figure 45 shows detention outcomes, by race, when the prosecutor
invoked the presumption at the Detention Hearing. Out of the
113 Detention Hearings in presumption cases, the prosecutor
explicitly invoked the presumption in only 35. Of those 35 hearings,
only one involved a white arrestee.

Figure 45: Detention Rates When Prosecutor Invoked
Presumption at Detention Hearing, By Race
100%

0%

0%

100%

0%

100%

0%

NA
NA
NA
NA

NA

Boston

NA

Baltimore

NA

Miami
Salt Lake City

NA

@m++

l

White

D. Md
(Baltimore)

D. Ma
(Boston)

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

73%*

100%**

71%*

-

Latino

-

100%**

100%*

100%**

Other

100%**

-

-

-

White

-

-

-

100%**

Black

236

100%

C

Judges’ Treatment of the Presumption
at Detention Hearings
In most presumption cases, either no Detention Hearing was held,
or the judge did not explicitly mention the presumption during the
hearing. Detention rates were generally high in presumption cases, but
they were even higher at Detention Hearings where a judge found that
the presumption applied and did not find the presumption rebutted
or did not discuss rebuttal. In 6 cases, however, the presumption was
found to be rebutted (5% of our sample: 6/113 Detention Hearings in
presumption cases). Judges released the arrestee in all of those cases.

Figure 46: Detention Outcomes, By Judges’ Treatment of the Presumption
at the Detention Hearing in Cases Where Presumption Applied by Law

Judge found presumption applied ...
... and arrestee
failed to rebut
100%

0%

... and did not
further discuss
rebuttal

... but found
presumption
rebutted

0%

0%

100%

100%

NA
NA
NA

NA
NA

j
Judge found presumption applied
and arrestee failed to rebut

D. Md
(Baltimore)

D. Ma
(Boston)

100%*

-

Judge did not mention
presumption

0%

100%

Salt Lake City

--

Miami

Boston
Baltimore

l

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

100%*

100%*

Judge found presumption applied 100%**
and did not further discuss rebuttal

100%**

100%**

100%**

Judge found presumption applied
but found presumption rebutted

-

-

-

-

Judge did not mention
presumption

50%**

70%*

75%

100%**
237

9

Financial
Conditions
In Baltimore and Salt Lake City, no financial conditions
of release were imposed. In Boston, 100% of arrestees
were able to meet the financial conditions and
were released. In Miami, however, 40% of arrestees
were unable to meet the financial conditions and
were ultimately detained pretrial. See Figure 25. We
identified these cases by checking PACER docket
reports for any release orders following the entry of a
financial condition in the case. If we found no release
order, we categorized the individual as detained
rather than released.
Of the districts we observed, corporate surety bonds (CSBs) were
only used in Miami. We observed 48 cases in Miami where CSBs
were imposed, representing 22% of the 220 cases we observed in that
district overall. The detention rate in cases where CSBs were imposed
was 92%, meaning 92% of arrestees were unable to satisfy the CSB
required for their release and were thus detained pretrial. Every
arrestee in those cases was a person of color. See Figure 24.

238

10

Noncitizen
Cases
Figure 47 shows the detention rate at the Initial
Appearance, at the Detention Hearing, and overall, for
identified noncitizens.
The overall detention rate for noncitizens was higher than for
citizens in each district we observed.

Figure 47: Detention Rates for Noncitizens vs. Citizens
Initial Appearance

Detention Hearing

Overall

0%

0%

0%

100%

•

100%

Non-citizen

•
• •

•
••
• •
• •

Citizen

D. Md
(Baltimore)

At IA
At DH

•
•

••
••
D. Ma
I(Boston)

100%

• •
• •

•

Baltimore

•

Salt Lake City
Miami
Boston

S.D. Fl
(Miami)

D. Utah
(Salt Lake City)

100%**

97%

100%**

Noncitizens

100%**

Citizens

62%

72%

75%

96%

Noncitizens

-

75%**

73%

100%**

Citizens

75%

79%

77%

81%

75%**

77%

100%**

60%

65%

79%

Overall Noncitizens 100%**
Citizens

52%

239

240

ENDNOTES
241

242

1

United States v. Salerno, 481 U.S. 739, 755 (1987).

2

The terms “jailing” and “detention” are used interchangeably in this Report to refer to the process by which an
arrestee is detained in jail pending trial. The terms “bond” and “bail” are used interchangeably as a shorthand
for the pretrial detention and release process more generally. We refer to the monetary aspects of pretrial
release as “financial conditions” generally, or name the specific type of monetary requirement imposed.

3

We have de-identified our interview subjects throughout this Report to ensure their anonymity, and therefore
will not be providing citations for interview quotes. To promote candor, we promised interviewees that
we would de-identify them, would not list their names in this Report, and would not attribute quotes to
them directly. The identity of the individual federal magistrate judge and Federal Defender who made each
statement quoted in this Report is on file with the authors.

4

Alison Siegler & Erica Zunkel, Rethinking Federal Bail Advocacy to Change the Culture of Detention, The Champion
46, 46 (July 2020), https://perma.cc/GA48-BY6Z.

5

18 U.S.C. § 3141 et seq.

6

The courtwatching component of our research was conducted and co-supervised by faculty and students at the
University of Maryland Law School’s Criminal Defense Clinic, Harvard Law School’s Criminal Justice Institute,
Boston University School of Law’s Criminal Law Clinical Program, and Campbell Law School’s Restorative
Justice Clinic.

7

We interviewed stakeholders from the following 11 federal courts of appeals: The First Circuit, Second Circuit,
Third Circuit, Fourth Circuit, Fifth Circuit, Sixth Circuit, Seventh Circuit, Eighth Circuit, Ninth Circuit, Tenth
Circuit, and Eleventh Circuit. Those stakeholders were located in the following 36 federal districts: the District
of Arizona, the Northern District of California, the Southern District of California, the District of Colorado, the
District of Delaware, the Southern District of Florida, the Central District of Illinois, the Northern District of
Illinois, the District of Maryland, the District of Massachusetts, the Eastern District of Michigan, the Northern
District of Mississippi, the Southern District of Mississippi, the Western District of Missouri, the Southern
District of New York, The Eastern District of New York, the District of New Mexico, the Eastern District of
North Carolina, the Western District of North Carolina, the District of North Dakota, the Northern District
of Ohio, the Southern District of Ohio, the Eastern District of Pennsylvania, the District of South Carolina,
the District of South Dakota, the Eastern District of Tennessee, the Western District of Texas, the Southern
District of Texas, the District of Utah, the District of the Virgin Islands, the Western District of Virginia,
the Eastern District of Virginia, the Western District of Washington, the Western District of Wisconsin, the
Eastern District of Wisconsin, and the District of Wyoming.

8

See Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 46 (coining the term “federal bail crisis”).

9

The Administration of Bail by State and Federal Courts: A Call for Reform, Hearing Before the Subcomm. on Crime,
Terrorism, and Homeland Security of the H. Comm. on the Judiciary, 116th Cong. 3 (Nov. 14, 2019) (written statement
of Alison Siegler, Dir. Federal Criminal Justice Clinic, University of Chicago Law School), https://www.congress.
gov/116/meeting/house/110194/witnesses/HHRG-116-JU08-Wstate-SieglerA-20191114.pdf [hereinafter Siegler
Bail Hearing written statement].

10 The data underlying this Report is on file with the authors. Given the relatively small size of our dataset,
our Report relies on summary statistics, rather than regression analyses. The arrestees in our study were
overwhelmingly people of color: Black individuals constituted 39% of our sample, Latino individuals 45%;
courtwatchers coded 2% of our sample “race unknown” (we assume, given the high percentage of people of
color in our sample overall, that these arrestees were most likely people of color of ambiguous ethnicity), and
1% “other.” Accordingly, our ability to estimate differences in the treatment of white arrestees versus nonwhite
arrestees is naturally somewhat limited from a statistical perspective. Moreover, our data do not allow us to
identify whether racial discrimination, in the purposeful sense that constitutional law focuses on, played a
role in outcomes. However, we can say with certainty that the vast majority of those bearing the brunt of the
federal bail crisis are Black and Latino. Some of the phenomena we documented were inflicted against only
people of color—for example, every single individual locked up for their inability to pay was Black or Latino.
People of color experienced other improper practices at higher rates than their representation in our sample.

11

admin. off. of U.S. CoUrTS, Caseload Statistics Data Tables, https://www.uscourts.gov/statistics-reports/caseloadstatistics-data-tables [https://perma.cc/FXS5-U8CY].

12

None of the AO’s publicly available H-Tables provides race information.

13

AO Table H-9. This annually updated table is not publicly available and can only be seen by insiders granted
access to the federal judiciary’s intranet system (the “J-Net”).

14

AO Table H-15. This annually updated table is not publicly available and can only be seen by insiders granted
access to the J-Net.

15

See, e.g., AO Table H-14, H-14A, H-3, H-3A.

16

18 U.S.C. §§ 3142(b)–(c) (mandating that the court “shall order . . . pretrial release . . . on personal recognizance,
or upon execution of an unsecured appearance bond . . . unless . . . such release will not reasonably assure
the appearance of the person as required or will endanger the safety of any other person or the community,”
in which case the court “shall order . . . pretrial release . . . subject to the least restrictive . . . condition, or
combination of conditions, that . . . will reasonably assure the appearance of the person as required and the
safety of . . . the community” (emphasis added)); see also United States v. Berrios-Berrios, 791 F.2d 246, 250
(2d Cir. 1986) (finding the Bail Reform Act of 1984 “codified . . . the traditional presumption favoring pretrial
release ‘for the majority of Federal defendants’” (quoting S. rep. no. 98-225, at 6–7 (1983), as reprinted in 1984
U.S.C.C.A.N. 3182, 3189))); United States v. Orta, 760 F.2d 887, 890 (8th Cir. 1985) (“[T]he statutory scheme of
18 U.S.C. § 3142 continues to favor release over pretrial detention.”).

17

18 U.S.C. § 3142(j).

18

481 U.S. 739 (1987).

19

Id. at 755.

20 United States v. Stone, 608 F.3d 939, 940 (6th Cir. 2010); see also United States v. Santos-Flores, 794 F.3d 1088,
1090 (9th Cir. 2015) (“Only in rare cases should release be denied, and doubts regarding the propriety of release
are to be resolved in favor of the defendant.”); United States v. Sabhnani, 493 F.3d 63, 75 (2d Cir. 2007) (“[T]he
law thus generally favors bail release.”); United States v. Byrd, 969 F.2d 106, 109 (5th Cir. 1992) (“There can be
no doubt that this Act clearly favors nondetention.”); United States v. Torres, 929 F.2d 291, 292 (7th Cir. 1991)
(deeming pretrial detention “an exceptional step”).

21

Pretrial Release and Detention: The Bail Reform Act of 1984, BUreaU of JUST. STaT. SpeCial rep., at 1 (Feb. 1988), https://
perma.cc/7A6U-S5XV [hereinafter BJS 1988 Report].

243

22 Id. (“The percent of defendants held on pretrial detention, that is, without bail, increased from less than 2%
before the Act to 19% after.”); see also id. at 2 tbl.1 & tbl.2 (10% of arrestees failed to meet bail conditions and
were held until trial).

23

See AO Table H-14 (2019); see also Alison Siegler & Kate M. Harris, How Did the ‘Worst of the Worst’ Become 3 out of
4?, n.Y. TimeS (Feb. 24, 2021), https://www.nytimes.com/2021/02/24/opinion/merrick-garland-bail-reform.html
[https://perma.cc/JN6N-LKVL]. We take the 2019 rate as our reference point because it represents the state of
the system before the COVID-19 pandemic hit.

24 AO Table H-14 (2020, 2021).
25

See infra Contextualizing the Culture of Detention—A Judge’s Decision to Jail Someone Pretrial Has Damaging
and Enduring Effects. Myriad studies demonstrate the devastating and long-lasting consequences of pretrial
detention. Pretrial detention adversely affects an individual’s physical and mental wellbeing, an individual’s
case outcome, and their economic stability, among other factors. See, e.g., Corruption, Abuse, and Misconduct at
U.S. Penitentiary Atlanta: Hearing Before the Homeland Sec. and Governmental Affairs Comm. Permanent Subcomm.
on Investigations, 117th Cong. 3–5 (July 26, 2022) (testimony of Rebecca Shepard, Assistant Federal Public
Defender, Northern District of Georgia), https://www.hsgac.senate.gov/subcommittees/investigations/
hearings/corruption-abuse-and-misconduct-at-us-penitentiary-atlanta [https://perma.cc/6KSB-93B2]
[hereinafter Shepard testimony] (listing unacceptable food and hygiene conditions and denials of mental
health treatment); Samuel R. Wiseman, Bail and Mass Incarceration, 53 Ga. l. rev. 235, 250 (“A group of studies
published in the past five years shows a compelling empirical connection between bail and convictions and
bail and guilty pleas, specifically.”); id. at 251–52 (“All of these studies controlled for factors not associated with
bail that could affect the likelihood of conviction and release, including, for example, race, age, gender, prior
offenses, and number of charged offenses . . . . [Pretrial detention is] a factor driving the higher conviction rates
for pretrial detainees.”); Alexander M. Holsinger & Kristi Holsinger, Analyzing Bond Supervision Survey Data: The
Effect of Pretrial Detention on Self-Reported Outcomes, 82 fed. proB. 39, 40–41 (2018), https://www.uscourts.gov/
sites/default/files/fedprobation-sept2018_0.pdf [https://perma.cc/2DZJ-Q6BE] (finding adverse and prolonged
effects on detainees’ economic and housing stability, as well as intergenerational effects for detainees who
have dependent children at the time of their jailing).

26 Thomas H. Cohen & Amaryllis Austin, Examining Federal Pretrial Release Trends over the Last Decade, 82(2) fed.
proB. 3, 10 (2018), https://www.uscourts.gov/sites/default/files/82_2_1_0.pdf [https://perma.cc/2SVY-A9TA];
see also id. at 6 (finding that among those not charged with immigration-related offenses, “the percentage
of defendants released pretrial has declined to a greater extent among defendants with less severe criminal
profiles than among defendants with more substantial criminal histories”).

27

Roughly 2% of federal cases involve violent offenses, as opposed to 25% of state felony cases. Compare U.S.
dep’T of JUST. BUreaU of JUST. STaT., Federal Justice Statistics 2015–2016, at 3, 11 (Jan. 2019), https://www.bjs.gov/
content/pub/pdf/fjs1516.pdf [https://perma.cc/X95K-QZT6], with U.S. dep’T of JUST. BUreaU of JUST. STaT., Felony
Defendants in Large Urban Counties, 2009, at 2 (Dec. 2013), https://www.bjs.gov/content/pub/pdf/fdluc09.pdf
[https://perma.cc/DJQ2-H6Y8].

28 In large urban counties, state systems only detained 38% of individuals charged with felonies in 2009.
See Felony Defendants in Large Urban Counties, 2009, supra note 27, at 15. The detention rate for violent state
felonies in 2009 was only 45%. Unfortunately, 2009 is the most recent year for which such nationwide
state-level pretrial detention data is available, preventing an exactly parallel comparison between the state
and federal numbers.

29 See BJS 1988 Report, supra note 21, at 5 tbl.10; AO Table H-9A (1997–2021).

244

30 AO Table H-9A (2021).
31

BJS 1988 Report, supra note 21, at 5 tbl.10 (showing an average of 53 days of pretrial detention in 1985).

32

See Memorandum, Costs of Community Supervision, Detention, and Imprisonment, admin. off. of U.S. CoUrTS
(Aug. 2021) (finding that the daily cost of supervision is $12, compared to $98 for pretrial detention). The cost
of incarceration rises annually. See, e.g., Incarceration Costs Significantly More than Supervision, admin. off. of
U.S. CoUrTS (Aug. 17, 2017) https://www.uscourts.gov/news/2017/08/17/incarceration-costs-significantlymore-supervision [https://perma.cc/Y4VU-8ZE9] (showing that in 2017, the cost was $31,842 for detention and
$4,026 for supervision).

33

See AO Table H-9, for the 12-Month Period Ending December 31, 2021 (2021) (showing that at the start of 2021,
there were 31,763 individuals detained pretrial, and by the end of the year, 32,004 individuals were jailed, for
an average of 31,884 individuals detained pretrial). At an average of $35,758 annually per person, federal pretrial
detention costs taxpayers upwards of $1.14 billion each year.

34 U.S. dep’T of JUST., fY 2021 BUdGeT reqUeST aT a GlanCe 3 (2021) https://www.justice.gov/doj/page/file/1246841/
download [https://perma.cc/E89P-JNY4] (showing costs of $9.2 billion in FY2021 towards “Prisons and
Detention.” This funding comprised nearly 30% of the Department of Justice’s (DOJ) discretionary
budget in FY2021).

35

Mark Motivans, Federal Justice Statistics, 2019, U.S. dep’T of JUST. BUreaU of JUST. STaT. SpeCial rep. at 8 tbl.5 (Oct.
2021), https://bjs.ojp.gov/content/pub/pdf/fjs19.pdf [https://perma.cc/J66J-6CV4] [hereinafter BJS 2019 Report].

36 See, e.g., Sonja B. Starr & M. Marit Rehavi, Racial Disparity in Federal Criminal Charging and Its Sentencing
Consequences, 122 J. pol. eCon 1320, 1349 (Dec. 2014) (“[B]lack male federal arrestees ultimately face longer
prison terms that whites arrested for the same offenses with the same prior records.”); id. at 1350 (“In the
federal system, more than half of the black-white sentence disparity that is unexplained by the arrest offense
and offenders’ prior traits can be explained by initial charge decisions, particularly the prosecutors’ decision
to file charges that carry ‘mandatory minimum’ sentences. Ceteris paribus, they do so 65 percent more often
against black defendants.”).

37

See Ad Hoc Comm. to Review the Criminal Justice Act, 2017 Report of the Ad Hoc Committee to Review the
Criminal Justice Act at xiv (2018), https://www.uscourts.gov/sites/default/files/2017_report_of_the_ad_hoc_
committee_to_review_the_criminal_justice_act-revised_2811.9.17.29_0.pdf [https://perma.cc/FM2Z-8GW3]
[hereinafter The Cardone Report] (finding that “90 percent of defendants in federal court cannot afford to hire
their own attorney”).

38

See, e.g., 18 U.S.C. § 3142(b), (c), (e), (g) (repeatedly highlighting that the goals of the BRA are to reasonably
assure “the appearance of the person as required” and “the safety of the community”); Guide to Judiciary
Policy, Vol. 8B, “Alternatives to Detention and Conditions of Release” (Monograph 110) at § 150 (finding that
“by imposing conditions of release and alternatives to detention, judicial officers are able to promote the
responsible use of public funding to protect the rights of defendants and to reasonably assure the appearance
of the defendant and the safety of the community as required”).

39 AO Table H-15 (2007–21). These tables show that federal failure-to-appear and rearrest rates have remained
consistently low over the past fifteen years, regardless of the detention rate. Although AO data regarding
failure-to-appear and rearrest rates are unavailable prior to 2007, some annual compendia from the Bureau of
Justice Statistics include these data. See, e.g., Laura Winterfield et al., Compendium of Federal Justice Statistics,
2004, U.S. BUreaU of JUST. STaT. at 53 tbl.3.8 (Dec. 2006), https://bjs.ojp.gov/library/publications/compendiumfederal-justice-statistics-2004 [https://perma.cc/4GN9-UHVF].

245

40 See AO Table H-15 (2007–21).
41

Id.; see also Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 47 & fig.1.

42 While release rates increased from 25% in 2019 to 29% in 2020 to 36% in 2021, the rates at which people
on pretrial release failed to appear in court or were rearrested remained extremely low. See AO Table H-14
(2019–21); see also AO Table H-15 (2019) (showing a failure-to-appear rate of 1.16% and a rearrest rate of 1.83%);
AO Table H-15 (2020) (showing a failure-to-appear rate of 1.17% and a rearrest rate of 1.94%); AO Table H-15
(2021) (showing a failure-to-appear rate of 1.32% and a rearrest rate of 2.25%). The federal Probation and
Pretrial Services Office recently emphasized that the slight increase in release rates during the pandemic
was not accompanied by any rise in the perennially low failure-to-appear and rearrest rates. See William
E. Hicks et al., Pretrial Work in a COVID-19 Environment, 85 fed. proB. 24, 29–30 (June 2021), https://www.
uscourts.gov/sites/default/files/usct10024-fedprobation-june2021_508.pdf [https://perma.cc/3YK9-NA4E]
(noting that “for the 12-month period ending in March of 2021, national failure rates remained low as follows:
failure to appear (1.7 percent), new criminal arrests (2.3 percent), and technical violations (4.1 percent),” and
that the increase in pretrial release rates during the COVID-19 pandemic came “without adverse effects on
pretrial supervision outcomes”).

43 See AO Table H-15 (2007–21) (listing “rearrest violations” for “felony,” “misdemeanor,” and “other”).
44 See, e.g., BJS 2019 Report, supra note 35, at 3 tbl.3, 6 tbl.10 (showing that there were fewer than half as many
federal convictions as arrests in 2019, at 78,254 and 206,630, respectively).

45

Paul Heaton, The Effects of Misdemeanor Bail Reform, qUaTTrone CTr. (Aug. 16, 2022), https://www.law.upenn.edu/
institutes/quattronecenter/reports/bailreform/#/lessons/298QqaqdYgFhhsKx7ei9zGKvT8ILGEVt [https://
perma.cc/BWK5-NRNQ].

46 d.C. preTrial ServiCeS aGenCY, ConGreSSional BUdGeT JUSTifiCaTion and performanCe BUdGeT reqUeST 41 (Mar.
24, 2022), https://www.psa.gov/sites/default/files/FY%202023%20PSA%20CBJ.pdf [https://perma.cc/X8R2DLNM].

47

Id. at 39.

48 Id. at 39, 41.
49 Heaton, supra note 45.
50 Id.

246

51

See John Pfaff, Op-Ed: The Never-Ending “Willie Horton Effect” Is Keeping Prisons Too Full for America’s Good, l.a.
TimeS (May 14, 2017), https://www.latimes.com/opinion/op-ed/la-oe-pfaff-why-prison-reform-isnt-working20170514-story.html [https://perma.cc/ZCZ9-JY6K] (“No matter how successful an ‘early release’ prison
program is, one single failure can impose huge political costs . . . [i]t didn’t matter that overall the reforms
appeared to be safely addressing the state’s mass incarceration problem; one murder derailed the effort.”); cf.
Note, Bail Reform and Risk Assessment: The Cautionary Tale of Federal Sentencing, 131 harv. l. rev. 1125, 1140 (2018)
(“Even after pretrial reforms, judges and prosecutors retain the incentive to be too cautious releasing people
pretrial because some released people will commit new crimes that will bring public scrutiny upon the judge or
district attorney . . . it’s a ‘covering your ass’ problem.”).

52

U.S. ConST. art. I; see also About Federal Judges, admin. off. of U.S. CoUrTS, https://www.uscourts.gov/judgesjudgeships/about-federal-judges [https://perma.cc/86QT-EWJG] (“Magistrate judges are judicial officers of the
U.S. district court appointed by the district judges of the court.”).

53

Erica Zunkel & Alison Siegler, The Federal Judiciary’s Role in Drug Law Reform in an Era of Congressional
Dysfunction, 18 ohio ST. J. Crim. 283, 306 (2020) (quoting Christopher L. Lowenkamp et al., The Hidden Costs
of Pretrial Detention, arnold foUnd. 19 (2013), https://craftmediabucket.s3.amazonaws.com/uploads/PDFs/
LJAF_Report_hidden-costs_FNL.pdf [https://perma.cc/FE39-BR7E]).

54

Judge James G. Carr, Why Pretrial Release Really Matters, 29 fed. SenT’G rep. 217, 220 (2017).

55

See, e.g., Holsinger & Holsinger, supra note 25, at 40–41.

56 Ellen A. Donnelly & John M. MacDonald, The Downstream Effects of Bail and Pretrial Detention on Racial
Disparities in Incarceration, 108 J. Crim. l. & CriminoloGY 775, 805 (2019).

57

Christopher Lowenkamp et al., Investigating the Impact of Pretrial Detention on Sentencing Outcomes, NaT’l inSTiT.
of Corr. 3 (Nov. 2013) (“Compared to defendants released at some point pending trial, defendants detained for
the entire pretrial period are more likely to be sentenced to jail or prison – and for longer periods of time.”).

58

Rachel E. Barkow, Categorical Mistakes: The Flawed Framework of the Armed Career Criminal Act and Mandatory
Minimum Sentencing, 133 harv. l. rev. 200, 221 (2019), https://harvardlawreview.org/wp-content/
uploads/2019/11/200-240_Online.pdf [https://perma.cc/Y3BS-M6A5].

59 Bush v. Gore, 531 U.S. 98, 106 (2000) (Stevens, J., dissenting).
60 The District of Massachusetts in the First Circuit, the District of Maryland in the Fourth Circuit, the District of
Utah in the Tenth Circuit, and the Southern District of Florida in the Eleventh Circuit.

61

We interviewed Chief Federal Public Defenders in all 4 of the districts where we courtwatched, yet we were able
to interview federal judges in only 3 of those districts (Baltimore, Boston, and Miami).

62 The detention eligibility net is a set of offenses and factors for which the BRA authorizes pretrial detention.
TimoThY r. SChnaCke, “model” Bail lawS: re-drawinG The line BeTween preTrial releaSe and deTenTion 78–80
(2017), https://perma.cc/8V4M-S9TF. This eligibility net “limits the type of federal criminal cases that are
eligible for pretrial detention.” Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48. The BRA’s
detention eligibility net is defined in § 3142(f). In Salerno, the Court emphasized that the BRA’s detention
eligibility net is quite narrow, and “carefully limits the circumstances under which detention may be sought to
the most serious of crimes.” 481 U.S. at 747; see also Bail, 50 Geo. l.J. ann. rev. Crim. proC. 394, 397–98 (2021) (“If
at the Initial Appearance the government requests that the defendant be detained, the judge is only authorized
to hold a detention hearing and detain the defendant if one of the factors in § 3142(f) is satisfied. If none of the
§ 3142(f) factors are satisfied, however, the judicial officer is prohibited from holding a detention hearing or
detaining the defendant pending trial.” (citing cases) (footnotes omitted)).

63 United States v. Ploof, 851 F.2d 7, 11 (1st Cir. 1988) (“Congress did not intend to authorize preventive
detention unless the judicial officer first finds that one of the § 3142(f) conditions for holding a
detention hearing exists.”).

64 18 U.S.C. § 3142(f)(1).
65 18 U.S.C. § 3142(f)(2).
66 We interviewed and surveyed stakeholders in 36 districts; 26 of those districts exhibited an access-to-counsel
problem at the Initial Appearance.

67 See The Cardone Report, supra note 37, at xiv.
68 18 U.S.C. § 3006A(c); see also fed. r. Crim. p. 44(a).
247

69 See 18 U.S.C. § 3142(e)(3) (“Subject to rebuttal by the person, it shall be presumed that no condition or
combination of conditions will reasonably assure the appearance of the person as required and the safety
of the community.”).

70 Amaryllis Austin, The Presumption for Detention Statute’s Relationship to Release Rates, 81 fed. proB. 52, 55 (2017),
https://www.uscourts.gov/sites/default/files/81_2_7_0.pdf [https://perma.cc/BL9D-D7LC].

71

See e.g., United States v. Wilks, 15 F.4th 842, 846 (7th Cir. 2021) (holding that the defense bears only a “light
burden of production” to rebut the presumption); United States v. Jessup, 757 F.2d 378, 380–84 (1st Cir. 1985)
(same); United States v. Dominguez, 783 F.2d 702, 707 (7th Cir. 1986) (same); United States v. Alatishe, 768 F.2d
364, 371 (D.C. Cir. 1985) (same); see also Bail, supra note 62, at 404 (2021) (collecting cases).

72

Siegler & Harris, supra note 23 (quoting Austin, supra note 70, at 61).

73

This analytic step was noted by the enacting Congress, which observed that all § 3142(g) factors must be
considered. See S. Rep. No. 98-225, at 24–25, as reprinted in 1984 U.S.C.C.A.N. at 3208–09 (“[A] court is expected to
weigh all the factors in the case before making its decision as to risk of flight and danger to the community.”).

74

18 U.S.C. § 3142(c)(2).

75

S. rep. no. 98-225, at 9, as reprinted in 1984 U.S.C.C.A.N. at 3192.

76 See 18 U.S.C. § 3142(c)(1)(B)(xii).
77

Salerno, 481 U.S. at 750.

78

Dominguez, 783 F.2d at 707 (emphasis added); see also supra note 71 and accompanying text.

79 See infra note 418 and accompanying text.
80 481 U.S. at 755.
81

admin. off. of U.S. CoUrTS, Caseload Statistics Data Tables, https://www.uscourts.gov/statistics-reports/caseloadstatistics-data-tables [https://perma.cc/FXS5-U8CY].

82 AO Table H-9A. After 2005, this annually updated table is not publicly available and can only be seen by
insiders granted access to the J-Net. Only data for the years 1997 through 2005 are publicly available.

83

AO Table H-14, H-14A, H-3, H-3A.

84 See 18 U.S.C. § 3142(f)(1).
85

See 18 U.S.C. § 3142(f)(2).

86 See 18 U.S.C. § 3142(e)(3). The BJS reports never provide any information about charging rates or detention rates
for the second-largest category of offenses that qualify for the presumption of detention: firearms offenses
charged under 18 U.S.C. § 924(c).

87

See, e.g., George E. Browne & Suzanne M. Strong, Pretrial Release and Misconduct in Federal District Courts, Fiscal
Years 2011–2018, BUreaU of JUST. STaT. at 1 (Mar. 2022), https://bjs.ojp.gov/content/pub/pdf/prmfdcfy1118.pdf
[https://perma.cc/C7CP-LYAN] [hereinafter BJS 2011–2018 Report] (combining 8 years of release rate statistics
into a single data point).

88 For example, the latest BJS report focusing on federal pretrial detention was published in March of 2022, see
generally id., and another report was published in October of 2021, see generally BJS 2019 Report, supra note 35.
But before that, the most recent BJS report on federal pretrial detention we could find was published in 2012.
248

See generally Thomas H. Cohen, Pretrial Release and Misconduct in Federal District Courts, 2008–2010, BUreaU
of JUST. STaT. (2012), https://www.bjs.gov/content/pub/pdf/prmfdc0810.pdf [https://perma.cc/3G9K-GLEF]
[hereinafter BJS 2008–2010 Report].

89 The annual BJS Federal Justice Statistics reports do not provide any information about detention outcomes by
race or detention outcomes by type of offense. See generally, e.g., BJS 2019 Report, supra note 35; Mark Motivans,
Federal Justice Statistics, 2017–2018, BUreaU of JUST. STaT. (Apr. 2021), https://bjs.ojp.gov/content/pub/pdf/fjs1718.
pdf [https://perma.cc/68BM-F4AF] [hereinafter BJS 2017–2018 Report]. The Compendium of Federal Justice
Statistics that BJS published annually through 2005 included detention outcomes by race and detention
outcomes by type of offense, but did not combine those datapoints to provide a picture of the rate at which
people of different races were detained pretrial, by type of offense.

90 BJS 2011–2018 Report, supra note 87, at 3 tbl.2.
91

BJS likewise does not disaggregate Latino and white arrestees in its earlier Pretrial Release and Misconduct in
Federal District Court reports (with the exception of the 2008–2010 report) or in its annual BJS Compendium of
Federal Justice Statistics (1992–2005).

92 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 47 & fig.2 (compiling studies before 2019).
93 Matthew G. Rowland, The Rising Federal Pretrial Detention Rate, in Context, admin. off. of The U.S. CoUrTS 14 fig.2
(Sept. 2018), https://www.uscourts.gov/sites/default/files/82_2_2_0.pdf [https://perma.cc/CPQ4-5GJR].

94 Stephanie Holmes Didwania, Discretion and Disparity in Federal Detention, 115 nw. U. l. rev. 1261, 1265 (2021),
https://scholarlycommons.law.northwestern.edu/nulr/vol115/iss5/1/ [https://perma.cc/3E3R-N4VN]; see also
id. at 1291 & fig.1 (finding in the raw data that a Black male arrestee was 20 percentage points more likely to be
detained than a white male arrestee, while a Latino male arrestee was almost 15 percentage points more likely
to be detained).

95 To obtain race data, courtwatchers recorded the demographics of everyone in the courtroom based on their
observations of each person’s physical appearance, and also recorded the basis for that assumption. Because
demographic information such as race and gender were not available from the docket sheet or any other public
source, we relied on courtwatchers’ perceptions and judgments.

96 In the Qualtrics survey where courtwatchers coded their results, we asked 3 questions about citizenship in
an effort to determine which arrestees in the study were not U.S. citizens: (1) “Was the client’s citizenship
mentioned in court or on the docket?” with answer choices Y/N; (2) “Is the client a citizen of the United
States?,” with answer choices reading “Y/N/Unsure (please explain your confusion)”; and (3) “What country is
the client a citizen of?,” with a blank for the courtwatcher to fill in.

97 The rate at which individuals were detained pretrial fell somewhat in 2020 and again in 2021. See AO Table
H-14 (2020, 2021). However, there is no evidence that the pandemic is to thank for that drop. It is entirely
possible that the critiques we and others began to levy against high federal pretrial detention rates starting in
2018 contributed to declining detention rates in the ensuing years. At the time of this Report’s publication, the
AO’s most recent public data ended in September 2021. See supra note 11.

98 AO Table H-9A (2019, 2021).
99 In that one case, however, defense counsel did not specifically argue that the arrestee was particularly
vulnerable to COVID-19. This arrestee was released at the Initial Appearance.

100 Nationwide, the rate at which prosecutors requested detention in pretrial hearings dropped by approximately
1% between 2019 and 2020 and by approximately 6% between 2020 and 2021. However, these changes were
249

not uniform across all federal courts. In fact, in Utah—one of the districts where we courtwatched—the
prosecutors’ detention request rate actually increased between 2019 and 2021 (from 93.3% to 94.5%). See AO Table
H-3 (2019, 2020, 2021).

101 Some studies about the consequences of pretrial detention were conducted in state systems rather than in
the federal system. Other studies use data which include both state and federal systems. Studies that are
not exclusively federal are nevertheless relevant to an understanding of the consequences of federal pretrial
detention; jail is jail, whether run by a state or the federal government. We expressly identify all federal and
joint federal/state studies discussed below; when a study is not so identified, it relied on state-level data only.

102 See, e.g., Charles E. Loeffler & Daniel S. Nagin, The Impact of Incarceration on Recidivism, 5 ann. rev. CriminoloGY
133, 143 (2021) (reviewing federal and state cases, researchers found, “[w]ith the combination of disruption
from temporary detention and the absence of programming or reentry resources, pretrial detention appears
unfavorable compared to less restrictive pretrial monitoring alternatives”); id. at 149 (“[T]he findings of studies
of pretrial incarceration are consistent—most find a deleterious effect on postrelease reoffending. This finding
adds to the growing literature on the disruptions that pretrial detentions have on the lives of those detained
even as many are not convicted or are convicted of minor offenses.” (citations omitted)); see also Lowenkamp
et al., supra note 57, at 19 (“Defendants detained pretrial were 1.3 times more likely to recidivate compared to
defendants who were released at some point pending trial.”); Austin, supra note 70, at 53 (citing Lowenkamp’s
2013 study for the proposition that “low-risk defendants who were held pretrial for two to three days were
almost 40 percent more likely to recidivate before trial compared to similarly situated low-risk defendants who
were detained for 24 hours or less,” those detained for 8 to 14 days “were 51 percent more likely to recidivate
within two years of their cases’ resolution, and when held for 30 or more days, defendants were 1.74 times more
likely to commit a new criminal offense”).

103 Diana D’Abruzzo, The Harmful Ripples of Pretrial Detention, arnold venTUreS (Mar. 24, 2022), https://www.
arnoldventures.org/stories/the-harmful-ripples-of-pretrial-detention [https://perma.cc/G5DM-5VEK] (citation
omitted).

104 Crystal S. Yang, Toward an Optimal Bail System, 92 N.Y.U. L. rev. 1399, 1426 (2017).
105 D’Abruzzo, supra note 103; see also generally Emily Leslie & Nolan G. Pope, The Unintended Impact of Pretrial
Detention on Case Outcomes: Evidence from New York City Arraignments, 60 J.l. & eCon. 529 (2017) (revealing that
pretrial jailing increases recidivism).

106 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 46 & n.3; see also supra note 27 and
accompanying text.

107 Will Dobbie et al., The Effects of Pretrial Detention on Conviction, Future Crime, and Employment: Evidence from
Randomly Assigned Judges, 108 am. eCon. rev. 201, 234–35 (2018).

108 Salerno, 481 U.S. at 747.
109 Gabriel A. Fuentes, Federal Detention and “Wild Facts” During the COVID-19 Pandemic, 110 J. Crim. l. & CriminoloGY
441, 448 (2020) (citing Hardeman v. Curran, 933 F.3d 816, 822–23 (7th Cir. 2019); Darnell v. Pineiro, 849 F.3d 17,
34–35 (2d Cir. 2017); Mays v. Dart, No. 20-C-2134, 456 F. Supp. 3d 966, 1002–03 (N.D. Ill. 2020)).

110 Laura M. Maruschak et al., Medical Problems of State and Federal Prisoners and Jail Inmates, 2011–12, U.S. BUreaU
of JUST. STaT. at 11 (Oct. 2016), https://bjs.ojp.gov/content/pub/pdf/mpsfpji1112.pdf [https://perma.cc/JM2EEV8Z].

111 Id.

250

112 Id.
113 Michael Balsamo & Michael R. Sisak, The Jail Where Jeffrey Epstein Killed Himself is Crumbling, ap newS (Sept. 23,
2021), https://apnews.com/article/health-prisons-new-york-manhattan-coronavirus-pandemic-5113a1e33a0c
3967787e04c0523e406b [https://perma.cc/2R4X-MHKS] (New York MCC); Alec Hamilton, Sentenced to Prison,
But Trapped in Jail, wnYC newS (Nov. 22, 2016), https://www.wnyc.org/story/sentenced-prison-trapped-jail/
[https://perma.cc/XG45-V282] (Brooklyn MDC); Annie Sweeney, MCC Can’t Escape Bad Reputation, Chi. TriB.
(Dec. 26, 2012), https://www.chicagotribune.com/news/ct-xpm-2012-12-26-ct-met-prison-escape-security-122320121226-story.html (Chicago MCC); Nick Chrastil, A Puerto Rican Federal Inmate’s Horrifying Account of What the
Prison Did After Hurricane Maria, Think proGreSS (Apr. 26, 2018), https://thinkprogress.org/puerto-rico-federaldetention-center-allegedly-abused-inmates-hurricane-maria-494c2a8306e8/ [https://perma.cc/KQK7-3GM2]
(Puerto Rico MCC).

114 Oversight Hearing on The Federal Bureau of Prisons and Implementation of the First Step Act of the House Subcomm.
on Crime, Terrorism, and Homeland Security, 116th Cong. 1 (Oct. 17, 2019) (statement of David E. Patton, Exec. Dir.,
Federal Defenders of New York), https://docs.house.gov/meetings/JU/JU08/20191017/110089/HHRG-116JU08-Wstate-PattonD-20191017.pdf [https://perma.cc/6LDQ-G7J6] [hereinafter Patton testimony].

115 Aviva Stahl, Prisoners Endure a Nightmare ‘Gulag’ in Lower Manhattan, Hidden in Plain Sight, The GoThamiST (June
19, 2018), https://gothamist.com/news/prisoners-endure-a-nightmare-gulag-in-lower-manhattan-hidden-inplain-sight; see also Alexander Nazaryan, Manhattan Jail Where Jeffrey Epstein Died Has Long History of Suicide,
Neglect, Yahoo (Aug. 14, 2019), https://news.yahoo.com/manhattan-prison-where-jeffrey-epstein-died-has-longhistory-of-suicide-neglect-172413714.html [https://perma.cc/5Z8V-SC5R].

116 Stahl, supra note 115.
117 T.A. DeFeo, Georgia’s Ossoff Says Atlanta Federal Prison Was ‘Extremely Dangerous and Insecure,’ CTr. SqUare
(July 27, 2022), https://www.thecentersquare.com/georgia/georgias-ossoff-says-atlanta-federal-prison-wasextremely-dangerous-and-insecure/article_f7045686-0db7-11ed-aaa0-2795af30132a.html [https://perma.
cc/8E85-8XE5].

118 Shepard testimony, supra note 25, at 3.
119 Joshua Ceballos & Alex Deluca, Employees at Understaffed Miami Prison Say Inmates, Guards, and Public Are at
Risk, mia. new TimeS (May 24, 2022), https://www.miaminewtimes.com/news/fdc-miami-staffing-shortagespurs-complaints-from-inmates-and-prison-guards-14521728 [https://perma.cc/E4MY-YCTM].

120 Shepard testimony, supra note 25, at 1, 9–11.
121 Id. at 2–5.
122 Id. at 1, 9–11.
123 See, e.g., id. at 5 (“Individuals [detained at USP Atlanta] are denied their right to effective assistance of
counsel. . . . The inevitable result of inadequate access to counsel is ineffective representation, and in some
cases, wrongful conviction.”); Patton testimony, supra note 114, at 5 (“[MCC Manhattan] was originally built
without rooms for attorney visitation even though it is a pretrial detention facility. The limited number of
attorney visitation rooms create expensive and aggravating delays.”).

124 Shepard testimony, supra note 25, at 5–9.
125 Id. at 4.

251

126 Sexual Victimization Reported by Adult Correctional Authorities, 2016–2018, BUreaU of JUST. STaT. at 1 (June 2021),
https://bjs.ojp.gov/content/pub/pdf/svraca1618.pdf [https://perma.cc/R64X-TLYL] [hereinafter BJS 2016–2018
Report]. The last year for which such statistics are available is 2018.

127 Id. at 5 tbl.2.
128 Id. at 1.
129 Id. at 23 app. tbl.6.
130 Id.
131 Emily Widra, No Escape: The Trauma of Witnessing Violence in Prison, priSon pol’Y inST. (Dec. 2, 2020), https://
www.prisonpolicy.org/blog/2020/12/02/witnessing-prison-violence/ [https://perma.cc/839X-SAVV]. While
Widra discusses both state and federal incarceration interchangeably, the statistics referenced here appear to
use exclusively federal data.

132 Id.
133 John Broman, What I Learned from 13 Years of Witnessing Violence in Federal Prisons, viCe (Sept. 25, 2015), https://
www.vice.com/en/article/nn9qeq/what-i-learned-from-13-years-of-witnessing-violence-in-federal-prisons-925.

134 Brandon L. Garrett & Lee Kovarsky, Viral Injustice, 110 Cal. l. rev. 117, 124 (2022).
135 Id.
136 Brandon Saloner et al., COVID-19 Cases and Deaths in Federal and State Prisons, 324 J. am. med. aSS’n 602,
602–03 (2020).

137 Jon Seidel, COVID-19 Outbreak Among Inmates at MCC in Chicago Among Largest Across the System, Chi. SUn
TimeS (May 10, 2020), https://chicago.suntimes.com/2020/5/10/21252571/covid-19-outbreak-inmates-mccchicago-among-largest-in-nation [https://perma.cc/6AHM-L7UG]. Given inadequate testing, the actual rate
of infection at federal detention centers is likely higher. See, e.g., Nick Pinto, Detainees at a Federal Jail Said Their
Coronavirus Symptoms Were Ignored. The Government is Fighting to Keep the Records Secret, The inTerCepT (May 15,
2020), https://theintercept.com/2020/05/15/mdc-brooklyn-coronavirus-hearing/ [https://perma.cc/WC9V4U97] (finding that by May 2020, only 15 people in the entire Brooklyn Metropolitan Detention Center had
been tested for COVID-19 despite many more complaining about symptoms).

138 Alison Siegler & Erica Zunkel, Commentary: Don’t Let Chicago’s Federal Jail Become the next Coronavirus Hot Spot,
Chi. TriB. (Apr. 24, 2020), https://www.chicagotribune.com/opinion/commentary/ct-opinion-coronavirus-jailcook-county-mcc-20200424-zagv2nvjyzcrvknxbfasusx63a-story.html.

139 Kelly Davis, Coronavirus in Jails and Prisons, The appeal (Sept. 30, 2020), https://theappeal.org/coronavirus-injails-and-prisons-61/ [https://perma.cc/3K5K-5PQ7].

140 COVID-19 Cases, fed. BUreaU of priSonS, https://www.bop.gov/coronavirus/ [https://perma.cc/X87V-CZMQ].
141 Hearing on Oversight of the Federal Bureau of Prisons: Hearing Before the Senate Judiciary Committee, 117th Cong.
8–9 (Apr. 15, 2021) (written testimony of Anita Aboagye-Agyeman, Assistant Federal Public Defender, District
of New Jersey & Alison K. Guernsey, Clinical Associate Professor, University of Iowa College of Law), https://
law.uiowa.edu/sites/law.uiowa.edu/files/2021-04/Congressional%20Testimony%20-%204-13-21-%20Final.pdf
[https://perma.cc/L66J-XV7C].

252

142 Compassionate Release, Univ. of iowa Coll. of law, fed. Crim. def. CliniC, https://law.uiowa.edu/compassionaterelease [https://perma.cc/J9XQ-YLPP] (finding that 302 people detained at BOP-managed or private prisons
with federal contracts had died from COVID-19 as of January 31, 2022).

143 Ayyab Rahim Abdullah, COVID-19 Experience at a Federal Detention Center, priSon JoUrnaliSm proJeCT (Mar. 25,
2021), https://prisonjournalismproject.org/2021/03/25/covid19-exerience-at-a-federal-detention-center/
[https://perma.cc/469H-L79C].

144 Robert Barton, Frequent Prison Lockdowns Backfire. I Know From Experience, poliTiCo (Feb. 2, 2022), https://
www.politico.com/news/agenda/2022/02/10/frequent-prison-lockdowns-backfire-00007797 [https://perma.
cc/3PVK-S63W].

145 How it Spreads, CTr. for diSeaSe ConTrol (July 2022), https://www.cdc.gov/poxvirus/monkeypox/transmission.
html [https://perma.cc/77RU-A28A].

146 Homer Venters, CDC Must Act to Prevent Monkeypox Explosion in Prisons, The hill (July 2022), https://thehill.
com/opinion/healthcare/3576465-cdc-must-act-to-prevent-monkeypox-explosion-in-prisons/ [https://perma.
cc/77RU-A28A].

147 Blair Paddock, As Monkeypox Spreads, Attention Turns Toward Jails, wTTw newS (Aug. 2022), https://news.wttw.
com/2022/08/02/monkeypox-spreads-attention-turns-toward-jails?mkt_tok=MjUwLUNRSC05MzYAAAGGCI
ZqRQhZ8PD-R4nkUCkKWw4XH-P7wR_MXvLBlZmGqvYUn18k02V9NUcb4_0rVluSQuH3zjIAi9Rk3eBz9ztSIF
UTEKfPLWR77xx2U-f9kqrP [https://perma.cc/8545-W3W9].

148 Jed S. rakoff, Why the Innocent Plead GuIlty and the GuIlty Go Free 23–24 (2021).
149 Donnelly & MacDonald, supra note 56, at 805.
150 Id. at 804; see also Wiseman, supra note 25, at 251–52 (“All of these studies controlled for factors not associated
with bail that could affect the likelihood of conviction and release, including, for example, race, age, gender,
prior offenses, and number of charged offenses . . . . [Pretrial detention is] a factor driving the higher conviction
rates for pretrial detainees.”).

151 Dobbie et al., supra note 107, at 225.
152 Stephanie Holmes Didwania, The Immediate Effects of Pretrial Detention, 22 am. l. & eCon. rev. 24, 26 (2020); see
also James C. Oleson et al., The Sentencing Consequences of Federal Pretrial Supervision, 63 Crime & delinqUenCY
313, 325 (2017) (finding “being released on pretrial services supervision was associated with a decrease
in the likelihood of being sentenced to prison, and a decrease in sentence length”); Lowenkamp et al.,
supra note 57, at 4:
Defendants who are detained for the entire pretrial period are much more likely
to be sentenced to jail and prison. Low-risk defendants who are detained for the
entire pretrial period are 5.41 times more likely to be sentenced to jail and 3.76 times
more likely to be sentenced to prison when compared to low-risk defendants who
are released at some point before trial or case disposition. Moderate and high-risk
defendants who are detained for the entire pretrial period are approximately
3 times more likely to be incarcerated than similar defendants who are released
at some point.

153 Didwania, Immediate Effects of Pretrial Detention, supra note 152, at 26 (“Pretrial release appears to decrease
sentence length, increase the probability of receiving a below-Guidelines sentence, and decrease the
probability of receiving a mandatory minimum.”).
253

154 Id.
155 Carr, supra note 54, at 218.
156 In our study, 89% of those detained at their Initial Appearance hearing were people of color.
157 Dobbie et al., supra note 107, at 204.
158 Id. at 227.
159 Holsinger & Holsinger, supra note 25, at 40.
160 Id. at 41.
161 Id. at 40.
162 While there are some differences between the state and federal systems, there is no reason to believe
that people detained federally would experience more salubrious long-term outcomes than their peers in
the state system.

163 Terry-Ann Craigie et al., Conviction, Imprisonment, and Lost Earnings, Brennan CTr. for JUST. 6 (Sept. 15, 2020),
https://www.brennancenter.org/our-work/research-reports/conviction-imprisonment-and-lost-earnings-howinvolvement-criminal [https://perma.cc/36AD-XARQ]. This study considered individuals who were imprisoned
post-conviction, rather than jailed pending trial. As the researchers explain, “[r]eliable information on the jail
population is hard to find, but the jail system’s sprawling size makes the issue an important area for future
research.” Id. at 7.

164 Id.; see also id. at 19 fig.4, 20 fig.5 (finding that, on an annual basis, Black people without any criminal record
earn less than socioeconomically similar white people with a criminal record).

165 Holsinger & Holsinger, supra note 25, at 40.
166 Id.
167 Id. (emphasis added).
168 See, e.g., Mark Halsey & Melissa de Vel-Palumbo, GeneraTionS ThroUGh priSon: experienCeS of inTerGeneraTional
inCarCeraTion (Jan. 2020) (finding that approximately one in five incarcerated individuals had a parent who
was or is currently incarcerated).

169 Holsinger & Holsinger, supra note 25, 40–41.
170 Id.
171 CharleS ColSon TaSk forCe on fed. Corr., TranSforminG priSonS, reSTorinG liveS 15 (Jan. 2016).
172 Scott Hechinger, We Must Follow Facts, Not Fear: Bail Reform Not the Reason for Rise in Some Crimes in NYC, dailY
newS (Mar. 6, 2022), https://www.nydailynews.com/opinion/ny-oped-nyc-follow-facts-not-fear-20220306e56f7dyxx5frzia5tdsrugaoiu-story.html [https://perma.cc/QZC8-F9YE].

173 Lauren Jones et al., The Cost of Incarceration in New York State at 15, vera inST. of JUST. (Jan. 2021), https://www.
vera.org/publications/the-cost-of-incarceration-in-new-york-state [https://perma.cc/EU36-9Y8W].

174 Ames Grawert & Noah Kim, The Facts on Bail Reform and Crime Rates in New York State, Brennan CTr. for JUST.
(Mar. 22, 2022), https://www.brennancenter.org/our-work/research-reports/facts-bail-reform-and-crime-ratesnew-york-state [https://perma.cc/FP5J-WC29]; see also Hechinger, supra note 172; Michael R. Sisak, Don’t Blame
254

Bail Reform for Higher Crime, NYC Watchdog Says, ap newS (Mar. 22, 2022), https://apnews.com/article/kathyhochul-covid-health-new-york-violence-aa42caf25b4fc333e4c864346f28d42c [https://perma.cc/9XBC-FR4L].

175 Heaton, supra note 45.
176 S. rep. no. 98-225, at 6, as reprinted in 1984 U.S.C.C.A.N. at 3188.
177 Id. at 7, as reprinted in 1984 U.S.C.C.A.N. at 3189.
178 Id. at 8–10, as reprinted in 1984 U.S.C.C.A.N. at 3190–93.
179 Salerno, 481 U.S. at 750.
180 Id. at 747.
181 Id. at 752.
182 See supra note 16 and accompanying text.
183 The legal standard that applies at the Initial Appearance is found in 18 U.S.C. § 3142(f).
184 18 U.S.C. § 3142(e).
185 18 U.S.C. § 3142(c)(2).
186 See 18 U.S.C. § 3142(f)(1)–(2). The factors listed in § 3142(f)(1) are offense-specific (such as many drug and gun
crimes or crimes involving a minor victim). The § 3142(f)(2) factors are subjective and allow a judge to schedule
a Detention Hearing in a case involving “a serious risk that [the] person will flee” (18 U.S.C. § 3142(f)(2)(A)
(emphasis added)) or “a serious risk that such person will obstruct or attempt to obstruct justice.” 18 U.S.C.
§ 3142(f)(2)(B).

187 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 49; Siegler Bail hearing written statement,
supra note 9, at 5 (“Caselaw furthers supports § 3142(f)’s role as a gatekeeper. . . . [E]very court of appeals to
address the issue agrees that it is illegal to detain someone—or even hold a Detention Hearing—unless the
government affirmative invokes one of the § 3142(f) factors.”); see also id. at 5 n.10 (collecting cases).

188 Bail, supra note 62, at 400 (“If there is no legal basis for detention at the Initial Appearance, [a judge] may
release a defendant on personal recognizance, on unsecured appearance bond, or subject to conditions.”).

189 See SChnaCke, supra note 62, at 78–80 (discussing the federal detention eligibility net); see also Pretrial Release:
Detention, naT’l Conf. of STaTe leGiSlaTUreS (June 20, 2022), https://www.ncsl.org/research/civil-and-criminaljustice/pretrial-detention.aspx [https://perma.cc/67L2-7WS2] (defining the “detention eligibility net” as “the
charges and circumstances that make a defendant eligible for detention”).

190 18 U.S.C. § 3142(f)(1).
191 18 U.S.C. § 3142(f)(2)(A).
192 18 U.S.C. § 3142(f)(2)(B).
193 18 U.S.C. § 3142(e)(3).
194 See 18 U.S.C. § 3142(e)(3) (no mention of crimes of violence or gun possession offenses under 18 U.S.C. § 922(g)).
It is irrelevant that some of the offenses listed in § 3142(f)(1) also qualify for a presumption of detention during
the Detention Hearing (such as drug offenses and § 924(c) gun offenses). Even in such cases, the presumption
of detention does not kick in until the Detention Hearing.
255

195 18 U.S.C. § 3006A (emphasis added).
196 S. Rep. No. 91-790, at 4 (1970); H.R. Rep. No. 91-1546, at 7 (1970) (emphasis added); see also fed. r. Crim.
p. 44 advisory committee’s note to 1966 amendment (“The phrase ‘from his initial appearance before the
commissioner or court’ is intended to require the assignment of counsel as promptly as possible after it
appears that the defendant is unable to obtain counsel.”); fed. r. Crim. p. 5(d)(2) (stating that “[t]he judge must
allow the defendant reasonable opportunity to consult with counsel” during the Initial Appearance).

197 United States v. Barber, 291 F. Supp. 38, 42 (D. Neb. 1968).
198 18 U.S.C. § 3142(f).
199 18 U.S.C. § 3142(f).
200 18 U.S.C. § 3142(f).
201 Compare 18 U.S.C. § 3142(b) (“The judicial officer shall order the pretrial release of the person . . . unless the
judicial officer determines that such release will not reasonably assure the appearance of the person as
required or will endanger the safety of any other person or the community.” (emphasis added)), with 18 U.S.C.
§ 3142(e)(3) (“Subject to rebuttal by the person, it shall be presumed that no condition or combination of
conditions will reasonably assure the appearance of the person as required and the safety of the community.”);
see also Bail, supra note 62, at 403.

202 See Bail, supra note 62, at 404 (citing 18 U.S.C. § 3142(e)(1), (f)) (“Even when a presumption of detention applies,
the government continues to bear the ultimate burden of proving that no conditions of release will reasonably
assure the defendant’s appearance and the safety of the community.”).

203 18 U.S.C. § 3142(e)(3) (a rebuttable presumption of detention applies at the Detention Hearing when the
charged offense is: (A) a drug crime carrying a maximum penalty of 10 or more years; (B) a crime listed in 18
U.S.C. § 924(c), § 956(a), or § 2332b; (C) a crime listed in 18 U.S.C. § 2332b(g)(5)(B) carrying a maximum penalty
of 10 years or more; (D) a crime listed in Title 18, Chapter 77, carrying a maximum penalty of 20 years or more;
or (E) a crime involving a minor victim).

204 18 U.S.C. § 3142(e)(2).
205 See, e.g., Dominguez, 783 F.2d at 707; see also Bail, supra note 62, at 404.
206 See, e.g., Dominguez, 783 F.2d at 707; Jessup, 757 F.2d at 384 (holding that a rebutted presumption is “but one
factor among many” and that a judge “is free to . . . release the defendant” in a presumption case).

207 See infra note 418 and accompanying text.
208 Wilks, 15 F.4th at 847.
209 18 U.S.C. § 3142(b).
210 18 U.S.C. § 3142(c)(1).
211 18 U.S.C. § 3142(c)(1).
212 18 U.S.C. § 3142(c)(1)(B)(i).
213 18 U.S.C. § 3142(c)(1)(B)(xiv) (allowing a judge to impose “any other condition that is reasonably necessary
to assure the appearance of the person as required and to assure the safety of any other person
and the community”).

256

214 18 U.S.C. § 3142(c)(2) (emphasis added).
215 As evinced by our courtwatching data, federal prosecutors often request detention for reasons that Congress
has deemed inappropriate or otherwise impermissible under the BRA. And because judges often do little to
curb these illegal requests, defense attorneys are reticent to push back—as their efforts would be futile.

216 See Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48 (“Prosecutors only provided a valid
basis for detention in 5 percent of cases and provided evidence to support the request in one case [among
those watched by the FCJC in Chicago.”); id. (“[I]n nearly 10 percent of the Phase 1 cases, there was no statutory
basis for detention whatsoever, rendering [the judge’s order of] detention illegal.”).

217 See id. at 46–47 (highlighting that “[t]he BRA was supposed to authorize detention for a narrow set of people:
those who are highly dangerous or pose a high risk of absconding”).

218 Bail, supra note 62, at 400; see also supra note 16 and accompanying text.
219 18 U.S.C. § 3142(b); 18 U.S.C. § 3142(c)(1).
220 18 U.S.C. § 3142(f).
221 Bail, supra note 62, at 398.
222 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 49.
223 Salerno, 481 U.S. at 747.
224 Siegler Bail Hearing written statement, supra note 9, at 5 (“A key reason the Supreme Court upheld the Bail
Reform Act as constitutional in United States v. Salerno was because the statute only authorizes detention at
the Initial Appearance under certain limited circumstances.”).

225 Salerno, 481 U.S. at 747 (emphasis added); see also S. rep. no. 98-225, at 20, as reprinted in 1984 U.S.C.C.A.N. at
3204 (“Because detention may be ordered under section 3142(e) only after a detention hearing pursuant to
subsection (f), the requisite circumstances for invoking a detention hearing in effect serve to limit the types
of cases in which detention may be ordered prior to trial.”); United States v. Watkins, 940 F.3d 152, 158 (2d Cir.
2019) (“Section 3142(f)(1) thus performs a gate-keeping function by ‘limit[ing] the circumstances under which
[pretrial] detention may be sought to the most serious of crimes.’” (citing Salerno, 481 U.S. at 747)).

226 See Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 49 (“When no § 3142(f) factor is met, the
judge is flatly prohibited from holding a Detention Hearing; the [arrestee] must be released.”); Siegler Bail
Hearing written statement, supra note 9, at 5 (“Caselaw furthers supports § 3142(f)’s role as a gatekeeper. . . .
[E]very court of appeals to address the issue agrees that it is illegal to detain someone—or even hold a
Detention Hearing—unless the government affirmative invokes one of the § 3142(f) factors.”); Ploof, 851 F.2d
at 11 (“Congress did not intend to authorize preventive detention unless the judicial officer first finds that
one of the § 3142(f) conditions for holding a detention hearing exists.”); United States v. Friedman, 837 F.2d 48,
49 (2d Cir. 1988); United States v. Himler, 797 F.2d 156, 160 (3d Cir. 1986); Byrd, 969 F.2d at 109; United States v.
Twine, 344 F.3d 987, 987 (9th Cir. 2003); United States v. Singleton, 182 F.3d 7, 9 (D.C. Cir. 1999) (“First, a [judge]
must find one of [seven] circumstances triggering a detention hearing . . . [under] § 3142(f). Absent one of these
circumstances, detention is not an option.”).

227 Watkins, 940 F.3d at 158 (2d Cir. 2019).
228 Ploof, 851 F.2d at 11.

257

229 U.S. Dep’t of Just., Just. Manual § 26 (Title 9-6100), https://www.justice.gov/archives/jm/criminalresource-manual-26-release-and-detention-pending-judicial-proceedings-18-usc-3141-et [https://perma.
cc/7R8V-YUXY] (citations omitted) (enumerating 6 (f) factors by counting § 3142(f)(2)(A) and (B) as a single
ground for detention).

230 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48–49.
231 See 18 U.S.C. § 3142(f)(1)(A)–(E) (“The judicial officer shall hold a [detention] hearing . . . upon motion of the
attorney for the Government, in a case that involves” one of five categories of offenses (emphasis added)).

232 Section 3142(f)(1) does not authorize the judge to hold a Detention Hearing on their own motion. Compare 18
U.S.C. § 3142(f)(1) (“upon motion of the attorney for the Government”) with 18 U.S.C. § 3142(f)(2) (“upon motion
of the attorney for the Government or upon the judicial officer’s own motion”).

233 See 18 U.S.C. § 3142(f) (“During a continuance, such person shall be detained.”).
234 Siegler Bail Hearing written statement, supra note 9, at 3.
235 See Starr & Rehavi, supra note 36, at 48.
236 Barkow, supra note 58, at 210 (“In the Bail Reform Act, one part of the [Comprehensive Crime Control Act],
Congress expanded the availability of pretrial detention.”); see also Statement of Sakira Cook on the Civil Rights
Implications of Cash Bail, leaderShip Conf. on Civ. & hUm. rTS. (Feb. 22, 2021), https://civilrights.org/resource/
statement-of-sakira-cook-senior-director-justice-reform-program-the-leadership-conference-on-civil-andhuman-rights-u-s-commission-on-civil-rights-hearing-on-the-civil-rights-implications/# [https://perma.cc/
ME94-X29S] [hereinafter Cook statement].

237 See Cook statement, supra note 236.
238 BJS 2019 Report, supra note 35, at 15 fig.7 (showing that the racial composition of people sentenced in drug, gun,
and crime of violence cases where the § 3142(f)(1) factors applied was 41% Black, 31% Hispanic, 4% other, and
24% white). In our dataset, people of color comprised approximately 88% of those who qualified for automatic
detention at the Initiation Appearance. The racial breakdown was: 41% Black, 44% Latino, 12% white, 1% other,
and 2% unknown.

239 See generally Austin, supra note 70.
240 Id. at 52.
241 Lauryn P. Gouldin, Defining Flight Risk, 85 U. Chi. l. rev. 677, 705 (2018).
242 Id.
243 See, e.g., Friedman, 837 F.2d at 49 (holding that a judge “must first determine by a preponderance of the evidence”
that § 3142(f)(1) or (2) is met); Himler, 797 F.2d at 160 (holding that an arrestee “may be detained [at the Initial
Appearance] only if the record supports a finding that he presents a serious risk of flight” under § 3142(f)(2)
(A) (emphasis added)). In addition, “there is a constitutional argument that the standard for flight risk
should be clear and convincing evidence” because “the Constitution requires a higher standard of proof than
preponderance of the evidence for deprivations of liberty.” Siegler & Zunkel, Rethinking Federal Bail Advocacy,
supra note 4, at 51 n.102; see also generally Jaden M. Lessnick, Comment, Pretrial Detention by a Preponderance:
The Constitutional and Interpretive Shortcomings of the Flight Risk Standard, 89 U. Chi. l. rev. 1245 (2022)
(contending that using a preponderance of the evidence standard to jail someone as a flight risk violates Due
Process and misinterprets the BRA).

258

244 18 U.S.C. § 3142(f)(2)(B) (on motion of the government or on the judge’s own motion).
245 Ploof, 851 F.2d at 11; see also United States v. Dillard, 214 F.3d 88, 96 (2d Cir. 2000) (“The question whether
the defendant poses a danger to the safety of the community . . . cannot be considered unless the defendant
is found to be eligible for detention under subsection 3142(f). A defendant who is not eligible must be released,
notwithstanding alleged dangerousness.” (emphasis added)); Friedman, 837 F.2d at 49 (“[T]he [BRA] does not
permit detention on the basis of dangerousness.”); Byrd, 969 F.2d at 110 (“[A] defendant who clearly may pose
a danger to society cannot be detained on that basis alone.”); Himler, 797 F.2d at 160 (detention at the Initial
Appearance is not authorized “upon proof of danger to the community other than from those offenses which
will support a motion for detention” under § 3142(f)(1)).

246 Twine, 344 F.3d at 987.
247 Bail, supra note 62, at 398.
248 Id. at 400.
249 Id. at 399.
250 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48 (“[T]he statute and case law make clear
that neither ‘danger to the community’ nor ordinary ‘risk of flight’ [as opposed to serious risk of flight] is a
legitimate basis for detention at the Initial Appearance Hearing.”).

251 Bail, supra note 62, at 398–99; see also Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 50.
252 Himler, 797 F.2d at 160 (emphasis added).
253 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 50.
254 None of these offenses is listed under § 3142(f)(1).
255 See, e.g., United States v. Taylor, 142 S. Ct. 2015 (2022) (holding that attempted Hobbs Act robbery does not
qualify as a “crime of violence” under 18 U.S.C. § 924(c)(3)(A) because no element of the offense requires proof
that the defendant used, attempted to use, or threatened to use force).

256 BJS 2011–2018 Report, supra note 87, at 2 tbl.1. Specifically, nearly 60% of federal cases charged from 2011–18
involved fraud, public order, or immigration offenses (as opposed to weapons offenses, drug offenses, or
property offenses that did not involve fraud). The detention rates at Initial Appearances for non-(f)(1) offenses
are surprisingly high: 37% for fraud arrestees, 57% for public order arrestees, and 91% for immigration arrestees.
See id.

257 See, e.g., United States v. Gibson, 384 F. Supp. 3d 955, 963 (N.D. Ind. 2019) (releasing defendant in a fraud case
and holding, “Any reading of the [BRA] that allows danger to the community as the sole ground for detaining
a defendant where detention was moved for only under (f)(2)(A) runs the risk of undercutting one of the
rationales that led the Salerno Court to uphold the statute as constitutional”); United States v. Morgan, No.
14-CR-10043, 2014 WL 3375028, at *1–6 (C.D. Ill. July 9, 2014) (concluding that financial dangerousness is not a
legitimate ground for detention at the Initial Appearance and denying the prosecutor’s detention request in an
access device fraud case).

258 BJS 2011–2018 Report, supra note 87, at 2 tbl.1.
259 Verbatim Notes Form, United States v. Dupin, No. 21-MJ-2225 (S.D. Fla. Feb. 5, 2021) (on file with the FCJC).
The 3-day detention in this particular case was not itself unlawful, as there was a legitimate statutory basis
for holding a Detention Hearing under § 3142(f)(1). But when no one in the courtroom mentions the BRA, a
problematic feedback loop develops that fuels unlawful detentions.
259

260 BJS 2011–2018 Report, supra note 87, at 2 (showing a 22% release rate at Initial Appearances nationwide).
261 The results of our 2019 Chicago pilot study were even more striking: the prosecutor cited an improper statutory
basis in 95% of the cases that we observed, citing danger to the community in 56% of cases and ordinary risk of
flight in 60% of cases. See Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48.

262 BJS 2011–2018 Report, supra note 87, at 2 tbl.1.
263 Verbatim Notes Form of Initial Appearance, United States v. Watson, No. 21-MJ-02322 (S.D. Fla. Feb. 19, 2021)
(on file with the FCJC).

264 Id.
265 Verbatim Notes Form of Detention Hearing, United States v. Watson, No. 21-MJ-02322 (S.D. Fla. Feb. 24, 2021)
(on file with the FCJC).

266 Verbatim Notes Form of Detention Hearing, United States v. Muniz, No. 21-MJ-02279 (S.D. Fla. Feb. 17, 2021)
(on file with the FCJC).

267 Id.
268 Id.
269 Id.
270 Verbatim Notes Form of Initial Appearance, United States v. Sterling, No. 19-CR-20736 (S.D. Fla. Mar. 2, 2021)
(on file with the FCJC).

271 Verbatim Notes Form of Initial Appearance, United States v. Patino, No. 11-CR-20409 (S.D. Fla. Feb. 19, 2021)
(on file with the FCJC).

272 Transcript of Initial Appearance at 8, United States v. Cedric Desmond Smith, Jr., No. 21-MJ-02264 (S.D. Fla.
Feb. 10, 2021) (on file with the FCJC).

273 Id.
274 Transcript of Initial Appearance at 5, United States v. Oviedo, No. 20-CR-20033 (S.D. Fla. Nov. 16, 2020),
ECF No. 43.

275 Salerno, 481 U.S. at 747.
276 BJS 2011–2018 Report, supra note 87, at 2 tbl.1.
277 We use the language “agreed to detention” to mean that the defense attorney submitted to detention, waived
the right to seek release, agreed to detention without prejudice, waived the Detention Hearing, or otherwise did
not contest detention at a given hearing.

278 See supra note 96.
279 18 U.S.C. § 3142(d).
280 See, e.g., United States v. Ailon-Ailon, 875 F.3d 1334, 1338 (10th Cir. 2017) (“[A] risk of involuntary removal
does not establish a serious risk that [the defendant] will flee.”); Santos-Flores, 794 F.3d at 1091 (“[T]he risk of
nonappearance referenced in 18 U.S.C. § 3142 must involve an element of volition.”); United States v. VillatoroVentura, 330 F. Supp. 3d 1118, 1135–36 (N.D. Iowa 2018); United States v. Suastegui, No. 3:18-MJ-00018, 2018 WL
3715765, at *4 (W.D. Va. Aug. 3, 2018); United States v. Marinez-Patino, 2011 WL 902466 (N.D. Ill. Mar. 14, 2011).
260

281 BJS 2008–2010 Report, supra note 88, at 15 tbl.13.
282 See id.
283 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 50.
284 Himler, 797 F.2d at 160.
285 Zunkel & Siegler, The Federal Judiciary’s Role in Drug Law Reform, supra note 53, at 314–15.
286 See id. at 315.
287 Salerno, 481 U.S. at 755.
288 We interviewed and surveyed stakeholders in 36 districts; 26 of those districts exhibited an access-to-counsel
problem at the Initial Appearance.

289 See BJS 2011–2018 Report, supra note 87, at 2 tbl.1 (documenting the high rate of detention at the
Initial Appearance).

290 See generally Danielle Soto & Mark Lipkin, Representation at Arraignment: The Impact of “Smart Defense” on Due
Process and Justice in Alameda County, impaCT JUST. (Dec. 2018), https://impactjustice.org/wp-content/uploads/
Smart-Defense-Report-2019.pdf [https://perma.cc/G82D-KT82]; Access to Counsel at First Appearance: A Key
Component of Pretrial Justice, naT’l leGal aid and def. aSS’n (Feb. 2020), https://www.nlada.org/sites/default/
files/NLADA%20CAFA.pdf [https://perma.cc/TU4G-NNEH].

291 See The Cardone Report, supra note 37, at xiv.
292 We use the term “indigent arrestee” as a shorthand for people who are unable to afford counsel to represent
them in federal court, while recognizing that someone does not need to be living in poverty to qualify for
appointed counsel under the Criminal Justice Act of 1964 (CJA), 18 U.S.C. § 3006A. The purpose and legislative
history of the CJA shows that Congress intended federal judges to appoint counsel for every arrestee unable
to afford one, not just poor arrestees. See Statement by Attorney General Robert F. Kennedy Before S. Comm. on the
Judiciary Regarding S. 1057, the Proposed Criminal Justice Act, 88th Cong. 7 (1963) (written statement of Robert F.
Kennedy, Attorney General of the United States) [hereinafter Kennedy CJA Testimony].

293 fed. r. Crim. p. 44 (“A defendant who is unable to obtain counsel is entitled to have counsel appointed to
represent the defendant at every stage of the proceeding from initial appearance through appeal, unless the
defendant waives this right.” (emphasis added)); see also 18 U.S.C. § 3006A(c) (“A person for whom counsel is
appointed shall be represented at every stage of the proceedings from his initial appearance before the United
States magistrate judge or the court through appeal.” (emphasis added)).

294 See United States v. Williams, 411 F. Supp. 854, 856 (S.D.N.Y. 1976) (“The federal courts have the duty to
implement the policies embodied in the Criminal Justice Act, as well as to ensure the general fairness of a
criminal proceeding. This duty extends beyond the standards required by the constitution and must prevent
even the possibility of unfairness.”).

295 18 U.S.C. § 3006A(c).
296 See From, n.2(a), oxford enG. diCTionarY online (Dec. 2021) (“Indicating the starting-point or the first considered
of two boundaries adopted in defining a given extent in space.”); Union P. R. Co. v. Hall, 91 U.S. 343, 348 (1875)
(“The words ‘from,’ ‘to,’ and ‘at,’ are taken inclusively, according to the subject-matter.”).

261

297 See Charlie Gerstein, Note, Plea Bargaining and the Right to Counsel at Bail Hearings, 111 miCh. l. rev. 1513, 1516
(2013) (“The right to the presence of counsel [at a bail hearing] and the right to her effective assistance are
coterminous—if you get one, you get the other.”).

298 S. rep. no. 88-346, at 12–13 (1963); h.r. rep. no. 88-864, at 7 (1963).
299 S. rep. no. 91-790, at 4; h.r. rep. no. 91-1546, at 7.
300 Barber, 291 F. Supp. at 42.
301 See Kennedy CJA Testimony, supra note 292, at 1 (“[The CJA] seeks to guarantee competent legal representation
and services to every accused person whose lack of funds prevents him from providing for his own defense.”).

302 Id. at 7.
303 Guide to Judiciary Policy, Vol. 7A, Ch. 2, Appx. 2A: Model Plan for Implementation and Administration of the
Criminal Justice Act 10, https://www.uscourts.gov/sites/default/files/vol07a-ch02-appx2a.pdf [https://perma.cc/
E6AY-YPYJ] [hereinafter Model Plan].

304 fed. r. Crim. p. 44 (emphasis added).
305 See fed. r. Crim. p. 44 advisory committee’s note to 1966 amendment.
306 fed. r. Crim. p. 5(d)(2).
307 See supra The Law: The Bail Reform Act Carefully Limits the Cases Eligible for Detention.
308 This problem is not limited to judges at the trial court level. Many appellate judges also fail to appreciate the
important liberty interest at stake during the Initial Appearance. Laboring under this misconception, these
judges write opinions that undercut the rights of arrestees and contribute to the culture of detention. See,
e.g., United States v. Perez, 776 F.2d 797, 800 (9th Cir. 1985) (characterizing an Initial Appearance as a benign
proceeding “at which the indictment is read, the name of the defendant asked, the defendant is apprised of his
Miranda rights, and counsel is appointed,” omitting entirely that the hearing also involves a pretrial detention
determination); United States v. Mendoza-Cecelia, 963 F.2d 1467, 1473 (11th Cir. 1992) (calling the Initial
Appearance “largely administrative” and trivializing the detention determination as “not a trial on the merits”).

309 See Gideon v. Wainwright, 372 U.S. 335, 344 (1963) (Not only . . . precedent[] but also reason and reflection require
us to recognize that in our adversary system of criminal justice, any person haled into court who is too poor for
a lawyer, cannot be assured a fair trial unless counsel is provided for him.”).

310 See supra Findings & Recommendations—The Practice: At the Initial Appearance Hearing, Federal Judges Jail
People Unlawfully.

311 Johnson v. Zerbst, 304 U.S. 458, 462–63 (1938) (emphasis added).
312 372 U.S. 335 (1963).
313 See id. at 339–45.
314 554 U.S. 191 (2008).
315 Id. at 212; Brewer v. Williams, 430 U.S. 387, 398 (1977) (“[A] person is entitled to the help of a lawyer at or after the
time that judicial proceedings have been initiated against him – ‘whether by way of formal charge, preliminary
hearing, indictment, information, or arraignment.’” (emphasis added) (quoting Kirby v. Illinois, 406 U.S. 682,
689 (1972))).

262

316 See, e.g., United States v. Wade, 388 U.S. 218, 224 (1967); Rothgery, 554 U.S. at 212 (“Once attachment occurs,
the accused at least is entitled to the presence of appointed counsel during any ‘critical stage’ of the
postattachment proceedings.”).

317 In Rothgery, the Court did not decide whether the Initial Appearance in Texas state court constitutes a
“critical stage,” holding only that the right to counsel “attaches” at such a hearing. 554 U.S. at 212; see also
id. at 216 (“[The critical stage analysis] lies beyond our reach, petitioner having never sought our review of
it.” (Alito, J., concurring)).

318 Id. at 198 (quoting Kirby, 406 U.S. at 689).
319 Bell v. Cone, 535 U.S. 685, 696 (2002) (defining “critical stage”); see also Hamilton v. State of Ala., 368 U.S. 52, 54
(1961) (holding arraignments in Alabama to be critical stages because what happens there affects the whole
trial, including the loss of certain defenses); Coleman v. Alabama, 399 U.S. 1, 7 (1970) (noting that pretrial stages
“where certain rights may be sacrificed or lost” have been held to be critical stages entitling someone to a
right to counsel).

320 Rothgery, 544 U.S. at 212.
321 Id. at 212 n.16 (quoting United States v. Ash, 413 U.S. 300, 312–13 (1973)).
322 Id. (quoting Wade, 388 U.S. at 226).
323 Wade, 388 U.S. at 227; see also United States v. Leonti, 326 F.3d 1111, 1117 (9th Cir. 2003) (“[T]he essence of a
‘critical stage’ is not its formal resemblance to a trial, but the adversary nature of the proceeding, combined
with the possibility that a defendant will be prejudiced in some significant way by the absence of counsel.”).

324 See, e.g., Salerno, 481 U.S. at 759.
325 See, e.g., Ditch v. Grace, 479 F.3d 249, 253 (3d Cir. 2007); Hurrell-Harring v. State, 930 N.E.2d 217, 223 (N.Y. 2010);
Gonzalez v. Comm’r of Corr., 68 A.3d 624, 641–43 (Conn. 2013) (Palmer, J., concurring), cert. denied 134 S. Ct. 639
(2013); Booth v. Galveston Cnty., 2019 WL 3714455, at *11–15 (S.D. Tex. Aug. 7, 2019); see also Paul S. Heaton,
Enhanced Public Defense Improves Pretrial Outcomes and Reduces Racial Disparities, 96 ind. l.J. 701, 741 (2021)
(“The results here provide clear support for the notion that the early bail settings should be considered a
critical stage, and therefore fall under the Sixth Amendment’s ambit.”).

326 In Philadelphia, for example, the likelihood of conviction increased by 13% when an arrestee was detained
pretrial, while in New York, conviction rates jumped from nearly half of all non-felony cases to 92%, and from
59% to 85% in felony cases, when individuals were subject to pretrial detention. See Megan Stevenson, Distortion
of Justice: How the Inability to Pay Bail Affects Case Outcomes, 34 J.l. eCon. & orG. 511, 512 (2018); Mary T. Phillips,
a deCade of Bail reSearCh in new York CiTY 116 (2012).

327 See Didwania, Immediate Effects of Pretrial Detention, supra note 152, at 26, 44–46 & tbl.4.
328 Ian A. Mance, Covid-19 Jail Restrictions and Access to Counsel, admin. of JUST. BUll., Univ. of n.C. SCh. of Gov’T (Oct.
2020); see also United States v. Davis, 449 F. Supp. 3d 532 (D. Md. 2020) (ordering a federal arrestee released
pretrial because COVID-19 restrictions at the jail would impede their access to counsel); State v. Hill, 277 N.C.
547, 552 (1971) (holding that the “right to communicate with counsel . . . necessarily includes the right of access
to them”); Geders v. United States, 425 U.S. 80, 91 (1976) (holding Sixth Amendment implicated where accused
was prevented “from consulting with his attorney . . . when an accused would normally confer with counsel”);
Maine v. Moulton, 474 U.S. 159, 170 (1985) (“[T]o deprive a person of counsel during the period prior to trial may
be more damaging than denial of counsel during the trial itself.”); Heaton, supra note 325, at 705 (“[T]he quality
of representation provided at the preliminary arraignment directly affects the outcome of the case and the
defendants’ likelihood of future involvement in the criminal justice system.”).
263

329 See supra Contextualizing the Culture of Detention—A Judge’s Decision to Jail Someone Pretrial Has
Damaging and Enduring Effects (illuminating the many pernicious consequences that accrue from pretrial
detention); see also State v. Fann, 239 N.J. Super. 507, 519 (Law. Div. 1990) (“The effect on family relationships
and reputation is extremely damaging. Failure of pretrial release causes serious financial hardship in most
cases. Jobs and therefore income are lost. The immediate consequence of the absence of bail or the inability to
make bail-deprivation of freedom-standing alone, is critically consequential.”).

330 Ash, 413 U.S. at 307 (“A concern of more lasting importance was the recognition and awareness that an unaided
layman had little skill in arguing the law or in coping with an intricate procedural system.”).

331 Id.; see also Booth, 2019 WL 3714455, at *11 (“Unrepresented defendants . . . are in no position at an initial bail
hearing to present the best, most persuasive case on why they should be released pending trial.”).

332 776 F.2d 797 (9th Cir. 1985).
333 Id. at 800.
334 Id.
335 See supra Contextualizing the Culture of Detention—Federal Pretrial Detention Leads to Longer Sentences
(“Strikingly, one study of federal and state arrestees found that ‘[p]retrial detention increases a defendant’s
likelihood of conviction by 55%,’ even when controlling for confounding factors.” (citations omitted)).

336 963 F.2d 1467 (11th Cir. 1992).
337 Id. at 1473.
338 See Gov’t of Canal Zone v. Peach, 602 F.2d 101, 104 (5th Cir. 1979); Ditch, 479 F.3d at 253; United States v.
Cookston, 379 F. Supp. 487, 487 n.1 (W.D. Tex. 1974); Caliste v. Cantrell, 329 F. Supp. 3d 296, 313–14 (E.D. La. 2018),
aff’d, 937 F.3d 525 (5th Cir. 2019); Booth, 2019 WL 3714455, at *16.

339 Swarthout v. Cooke, 562 U.S. 216, 219 (2011); see also Mathews v. Eldridge, 424 U.S. 319, 334–35 (1976).
340 See Caliste, 329 F. Supp. 3d at 312; Schultz v. State, 330 F. Supp. 3d 1344, 1374 (N.D. Ala. 2018). But see Booth, 2019
WL 3714455, at *8–9 (holding that a failure to provide counsel at a state Initial Appearance violates the Sixth
Amendment but not the Fifth Amendment).

341 The “duty attorney” is the appointed lawyer responsible for handling new arrests on a given day and may be an
Assistant Federal Public Defender who works for the federal defender’s office, or a Criminal Justice Act Panel
Attorney (CJA attorney) who is appointed by the court to represent indigent arrestees but is not employed by
the federal defender’s office.

342 fed. r. Crim. p. 44 advisory committee’s note to 1979 amendment.
343 The Defender elaborated: “[T]here was a point system where, if you declined a case, you dropped to the bottom
of the list.” “Now, they wouldn’t wait for people to call them back. If they called and they got voicemail or
whatever, they just keep going down the list until they find a CJA lawyer able to take the case.”

344 Specifically, we recorded an individual as “partially represented by counsel” or “represented during part of the
Initial Appearance” if they received representation after substantive questioning by a judge, e.g., about his
financial conditions or other elements of his case. In contrast, we recorded an individual as “fully represented”
or “represented during the entire Initial Appearance” only if a defense attorney was present for the entire
hearing and appeared to be representing the arrestee from the beginning of the hearing. We tracked this
distinction because of the potential for arrestees to inadvertently incriminate themselves before they receive
appointed counsel—a potential that was realized in multiple cases that we observed.
264

345 In addition, information about the presence or absence of an attorney at the Initial Appearance was missing
from 3% of our data set.

346 Verbatim Notes Form of Initial Appearance, United States v. Vallejo, No. 18-CR-20796 (S.D. Fla. Feb. 22, 2021)
(on file with the FCJC). While the deprivation of counsel violated the law, the 3-day detention in this particular
case was not itself unlawful, as there was a legitimate statutory basis for holding a Detention Hearing under
§ 3142(f)(1). However, when there is no valid statutory basis for holding a Detention Hearing or detaining an
arrestee at the Initial Appearance under § 3142(f), an unrepresented arrestee has no chance of vindicating their
rights or opposing a prosecutor’s detention request.

347 Revised Plan for Furnishing Representation Pursuant to the Criminal Justice Act 18 U.S.C. § 3006A, V: Timely
Appointment of Counsel 11, https://www.flsd.uscourts.gov/sites/flsd/files/CJA%20Plan%202021.pdf [https://
perma.cc/967T-PHLK].

348 There are 94 federal judicial districts. See 28 U.S.C. § 81–132; admin off. of The U.S. CoUrTS., Court Role and
Structure, https://www.uscourts.gov/about-federal-courts/court-role-and-structure [https://perma.cc/WX74S4YM]. The denial of counsel at the Initial Appearance is a confirmed problem in at least one division of at
least 16 of those districts: the Southern District of Florida, and the 15 additional named districts confirmed in
our research and named below. These 16 districts account for 17% of the federal district courts. As discussed,
the problem may also exist in 10 additional districts, bringing the total to 26. These 26 districts account for
28% (over one-quarter) of the federal district courts.

349 The 16 named districts alone span the Third Circuit, Fourth Circuit, Fifth Circuit, Sixth Circuit, Seventh
Circuit, Eighth Circuit, Ninth Circuit, Tenth Circuit, and Eleventh Circuit.

350 In at least one division of every district named here, some arrestees are not represented by appointed counsel
during the Initial Appearance. The many permutations of the problem are described in more detail below.

351 The information about these 10 additional unnamed districts was obtained from interviews and surveys
conducted across the country. We conducted some of these surveys in 2020–2021 and others in 2022. If a
district recently altered its representation practices, our survey results may not capture those changes.

352 This Report’s authors have closely reviewed the CJA Plans for each of the 26 districts. To the degree that those
plans differ from the Model Plan, none authorizes courts to wait until after the Initial Appearance hearing to
provide counsel—with one notable exception. The Central District of Illinois’s CJA Plan removes the language
that counsel must be appointed “as soon as feasible in the following circumstances, whichever occurs earliest.”
By deleting that clause, the Plan leaves the timing of appointment of counsel entirely to the judge’s discretion,
does not require appointment of counsel “when [the person] first appear[s]” in court, and enables the judge to
delay appointing counsel until such time as the “judicial officer otherwise considers appointment of counsel
appropriate under the CJA and related statutes.” See Criminal Justice Act Plan, U.S. diST. CT. CenT. diST. ill.
2–6 (Nov. 2019), https://www.ilcd.uscourts.gov/sites/ilcd/files/general-ordes/Nov.%202019%20REVISED%20
CJA%20PLAN%20Final.pdf [https://perma.cc/U928-JMC3].

353 In the District of New Mexico, for example, there is universal representation during the Initial Appearance in
Albuquerque and Las Cruces, but there is a remote magistrate court in Farmington where they are still in the
process of assuring representation at the Initial Appearance.

354 These districts changed their practices regarding representation by counsel at the Initial Appearance within
the last decade. In some of these districts, the change occurred in the last one to three years.

355 This 33% figure was reached by adding the 5 districts discussed here to the 26 districts discussed above
(31 districts) and dividing by the 94 total districts.
265

356 See supra Findings & Recommendations—The Practice: At the Initial Appearance Hearing, Federal Judges Jail
People Unlawfully.

357 Two of these arrestees were unlawfully detained, as there was no statutory basis under § 3142(f)(1) for a
Detention Hearing.

358 Transcript of Initial Appearance at 6, United States v. Kramer, No. 21-CR-20069 (S.D. Fla. Feb. 11, 2021),
ECF No. 12.

359 Transcript of Initial Appearance at 4–5, Oviedo, No. 20-CR-20033, ECF No. 43; see also id. at 2, 6 (showing the
only 4 sentences spoken by the prosecutor during the hearing).

360 Black and Latino arrestees made up 87% of our sample in total.
361 Among the 134 Black arrestees in our study, 46 were not provided lawyers until partway through their Initial
Appearance; 49 of the 153 Latino arrestees in our study likewise were not represented until partway through
their Initial Appearance.

362 Transcript of Record at 4, United States v. Lampe, No. 21-MJ-02111 (S.D. Fla. Jan. 22, 2021), ECF No. 3.
363 See Kennedy CJA Testimony, supra note 292, at 1.
364 model rUleS of pro. CondUCT R. 3.8 (am. Bar aSS’n, Comment 2020), https://www.americanbar.org/groups/
professional_responsibility/publications/model_rules_of_professional_conduct/rule_3_8_special_
responsibilities_of_a_prosecutor/comment_on_rule_3_8/ [https://perma.cc/7R8V-YUXY].

365 The promise of innocent until proven guilty is even codified in the BRA itself. See 18 U.S.C. § 3142(j) (“Nothing
in this section shall be construed as modifying or limiting the presumption of innocence.”).

366 Didwania, Discretion and Disparity, supra note 94, at 1278.
367 18 U.S.C. § 3142(e)(3). There is a second presumption in 18 U.S.C. § 3142(e)(2), but it is very narrow and rarely
applies. “Note that [under § 3142(e)(2)] there is a separate presumption of detention if the person is charged
with a detention eligible crime under § 3142(f), has been convicted of a similar offense, was on release when
the prior offense was committed, and not more than five years have elapsed since conviction or release for that
prior offense. . . . This § 3142(e)(2) presumption is extraordinarily rare.” Siegler & Zunkel, Rethinking Federal Bail
Advocacy, supra note 4, at 58 n.81.

368 Siegler & Harris, supra note 23.
369 Zunkel & Siegler, Federal Judiciary’s Role in Drug Law Reform, supra note 53, at 292.
370 Austin, supra note 70, at 55.
371 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 50.
372 Siegler & Harris, supra note 23.
373 Id.
374 18 U.S.C. § 3142(e)(3). At the Initial Appearance, individuals charged in presumption-of-detention cases are
subject to automatic temporary detention until a Detention Hearing takes place. See 18 U.S.C. § 3142(f)(1).
Importantly, the § 3142(f)(1) factors and the presumption are not coextensive. But while (f)(1) is more expansive,
the crimes captured by the presumption are also included in the (f)(1) factors.

375 The presumption for drug and gun cases applies in approximately 45% of all federal cases, while the
presumption for drug cases in particular applies in 93% of all federal drug cases. See Austin, supra note 70, at 55.
266

376 Siegler & Harris, supra note 23.
377 Zunkel & Siegler, Federal Judiciary’s Role in Drug Law Reform, supra note 53, at 293 (quotation marks omitted).
378 See id. at 283, 290. The original presumption did not include terrorism or crimes involving minor victims. These
crimes were added under the presumption in later iterations of the BRA. See Antiterrorism and Effective Death
Penalty Act of 1996, Pub. L. No. 104-132 (1996); Prosecutorial Remedies and Tools Against the Exploitation of
Children Today Act of 2003 (PROTECT Act), Pub L. No. 108-21 (2003); see also Barkow, supra note 58, at 210.

379 See generally Mark W. Bennett, A Slow Motion Lynching? The War on Drugs, Mass Incarceration, Doing Kimbrough
Justice, and a Response to Two Third Circuit Judges, 66 rUTGerS l. rev. 873 (2014) (showing how the “War on Drugs”
and associated criminal laws disproportionately affect people of color).

380 Zunkel & Siegler, Federal Judiciary’s Role in Drug Law Reform, supra note 53, at 292.
381 Id. at 293.
382 Id. at 294 (quotation marks omitted).
383 Id.
384 Federal Judicial Caseload Statistics 2020, admin. off. of The U.S. CoUrTS (2021), https://www.uscourts.gov/
statistics-reports/federal-judicial-caseload-statistics-2020 [https://perma.cc/AZX3-LCNT].

385 John Scalia, Federal Drug Offenders, 1999 with Trends 1984–99, BUreaU of JUST. STaT. at 1 (Aug. 2001), https://www.
csdp.org/research/fdo99.pdf [https://perma.cc/47FY-FZZ6] [hereinafter BJS 1984–1999 Report].

386 Didwania, Discretion and Disparity, supra note 94, at 1322 (comparing presumption cases with
non-presumption cases).

387 Id. at 1306 (ultimately concluding that the presumption “does not appear to explain race-based disparity
in detention”).

388 For example, in 2019, 75% of people sentenced for drug trafficking—the most common presumption offense—
were people of color. See U.S. SeNt’g Comm’N, 2019 ANNUAl RepoRt ANd SoURCebook of fedeRAl SeNteNCiNg StAtiStiCS
48 tbl.5, 110 tbl.D-2 (2020), https://www.ussc.gov/sites/default/files/pdf/research-and-publications/annualreports-and-sourcebooks/2019/2019-Annual-Report-and-Sourcebook.pdf [https://perma.cc/XMX7-JDYF]
[hereinafter Sent’g Comm’n 2019 Report]. Assuming that sentencing data roughly tracks with charging data,
people of color are disproportionately subject to the presumption of detention. Federal Bureau of Prisons data
points to the same conclusion: 76% of people incarcerated for federal drug offenses in 2019 were people of
color. See Mark Motivans, Federal Justice Statistics, 2015–2016, BUreaU of JUST. STaT. at 15 fig.7 (Jan. 2019), https://
www.bjs.gov/content/pub/pdf/fjs1516.pdf [https://perma.cc/XVP8-VA6W] [hereinafter BJS 2015–2016 Report].

389 JUd. Conf. of The U.S., reporT of The proCeedinGS of The JUdiCial ConferenCe of The UniTed STaTeS 10–11 (2017),
https://www.uscourts.gov/sites/default/files/17-sep_final_0.pdf [https://perma.cc/2GFW-LZUA].

390 Austin, supra note 70, at 57.
391 Id. at 53.
392 Id. at 60. “High-risk . . . cases” involve arrestees who fall in Pretrial Risk Assessment (PTRA) categories 4 and 5,
the highest categories in the system. Id. at 57. But arrestees in presumption cases had similar—if not lower—
rearrest and failure-to-appear (FTA) rates as arrestees in non-presumption cases across every PTRA category.
See id. at 56 tbl.2.

267

393 See id. (showing lower failure-to-appear rates in presumption cases across 3 of the 5 PTRA categories).
394 For a comprehensive discussion of the operation of the presumption as a matter of law, see Alison Siegler,
Guest Posts on Big Seventh Circuit Wilks Decision on Bail Reform Act’s “Presumption of Detention,” SenT’G l. &
pol’Y (Jan. 19, 2022), https://sentencing.typepad.com/sentencing_law_and_policy/2022/01/guest-postson-big-seventh-circuit-wilks-decision-on-bail-reform-acts-presumption-of-detention.html [https://perma.
cc/4M7T-YCWV].

395 See United States v. Moore, 607 F. Supp. 489, 499 (N.D. Cal. 1985):
In sum, the court finds that the rebuttable presumption of the BRA places only a burden
of production on defendant and requires the government to rebut it with clear and convincing evidence. The government may not rely on the presumption alone to satisfy the
clear and convincing evidence standard. This construction affords the defendant an opportunity to show that the presumption does not apply to him so as to avoid any Eighth
Amendment or Fifth Amendment problems stemming from “conclusive” presumptions.

396 Foucha v. Louisiana, 504 U.S. 71, 80 (1992).
397 Reno v. Flores, 507 U.S. 292, 302 (1993) (citing Salerno, 481 U.S. at 746).
398 United States v. Scott, 450 F.3d 863, 874 (9th Cir. 2006); see also Lopez-Valenzuela v. Arpaio, 770 F.3d 772, 791
(9th Cir. 2014) (“[T]he individualized determination of flight risk or dangerousness . . . [was something] Salerno
deemed essential” in upholding the BRA’s constitutionality.).

399 Lopez-Valenzuela, 770 F.3d at 791.
400 Id.
401 Wilks, 15 F.4th at 846; see also Stone, 608 F.3d at 945; United States v. Stricklin, 932 F.2d 1353, 1354–55 (10th Cir.
1991); United States v. Moss, 887 F.2d 333, 338 (1st Cir. 1989); United States v. Hare, 873 F.2d 796, 798 (5th Cir.
1989); United States v. Martir, 782 F.2d 1141, 1144 (2d Cir. 1986); United States v. Perry, 788 F.2d 100, 115 (3d Cir.
1986); Alatishe, 768 F.2d at 371 n.14; United States v. Chimurenga, 760 F.2d 400, 405 (2d Cir. 1985); United States
v. Hurtado, 779 F.2d 1467, 1470 n.4 (11th Cir. 1985); Orta, 760 F.2d at 891 n.17.

402 Dominguez, 783 F.2d at 707 (emphasis added); see also Bail, supra note 62, at 404 (“Once defendants ‘[come]
forward with some evidence that [they] will not flee or endanger the community if released,’ the presumption
of flight risk and dangerousness is rebutted.” (quoting Dominguez, 783 F.2d at 707)); Jessup, 757 F.2d at 380–84;
Alatishe, 768 F.2d at 371.

403 Dominguez, 783 F.2d at 707 (“Any evidence favorable to a defendant that comes within a category listed in
§ 3142(g) can affect the operation of one or both of the presumptions, including evidence of their marital,
family and employment status, ties to and role in the community, clean criminal record and other types of
evidence encompassed in § 3142(g)(3).”).

404 Id.
405 United States v. Jackson, 845 F.2d 1262, 1266 (5th Cir. 1988) (emphasis added).
406 Wilks, 15 F.4th at 844.
407 Id.
408 Id. at 847; see also Siegler, supra note 394 (“Even if the defense does not carry its light burden of production
. . . [, under Wilks,] that lack of rebuttal does not, standing alone, authorize detention in a presumption case.”);
268

State v. Mascareno-Haidle, 2022 WL 2351632, at *8 (N.M. June 30, 2022) (“[P]retrial detention or release
decisions cannot be made to turn on any single factor, be it the nature and circumstances of the charged
offense(s) or otherwise.”).

409 See Wilks, 15 F.4th at 847; Jackson, 845 F.2d at 1266.
410 Jackson, 845 F.2d at 1266 (quoting 18 U.S.C. § 3142(g)(2)).
411 Wilks, 15 F.4th at 847; see also S. rep. no. 98-225, at 23, as reprinted in 1984 U.S.C.C.A.N. at 3206:
Subsection (g) enumerates the factors that are to be considered by the judicial officer
in determining whether there are conditions of release that will reasonably assure
the appearance of the person and the safety of any other person and the community.
Since this determination is made whenever a person is to be released or detained under
this chapter, consideration of these factors is required . . . [and] a court is expected
to weigh all the factors in the case before making its decision as to risk of flight and
danger to the community. (emphasis added).

412 Wilks, 15 F.4th at 846–47 (emphasizing that the burden of persuasion never shifts to the defense); Bail, supra
note 62, at 404 (“Even when a presumption of detention applies, the government continues to bear the
ultimate burden of proving that no conditions of release will reasonably assure the defendant’s appearance
and the safety of the community.”).

413 See, e.g., Orta, 760 F.2d at 891 n.17 (“In this case, the district court erred in interpreting the ‘reasonably assure’
standard set forth in the statute as a requirement that release conditions ‘guarantee’ community safety and
the defendant’s appearance. Such an interpretation contradicts both the framework and the intent of the
pretrial release and detention provision of the 1984 Act.”).

414 18 U.S.C. § 3142(f), (e)(1).
415 Salerno, 481 U.S. at 751 (emphasis added); see also United States v. Munchel, 991 F.3d 1273, 1280 (D.C. Cir.),
judgment entered, 844 F. App’x 373 (D.C. Cir. 2021) (citing this passage in reversing district court’s detention
order). The New Mexico Supreme Court recently reflected as follows on the clear and convincing evidence
standard in the pretrial detention context:
In keeping with the presumption of innocence that attaches to all defendants prior
to conviction . . . [,] [p]roof by clear and convincing evidence represents that standard,
one satisfied only by evidence that instantly tilts the scales in the affirmative when
weighed against the evidence in opposition and the fact finder’s mind is left with an
abiding conviction that the evidence is true.
Mascareno-Haidle, 2022 WL 2351632, at *6 (citations and internal quotation marks omitted).

416 Dominguez, 783 F.2d at 706.
417 See, e.g., Bail, supra note 62, at 404 (quoting Dominguez, 783 F.2d at 707).
418 Courts generally treat even a rebutted presumption as relevant to a detention determination. See Dominguez,
783 F.2d at 707 (holding that a rebutted presumption “remains in the case as an evidentiary finding militating
against release, to be weighed along with the other evidence relevant to factors in § 3142(g)”); United States
v. Hir, 517 F.3d 1081, 1086 (9th Cir. 2008); United States v. Rodriguez, 950 F.2d 85, 88 (2d Cir. 1991); Hare, 873
F.2d at 788–89; Jessup, 757 F.2d at 384; United States v. Dillon, 938 F.2d 1412, 1416 (1st Cir. 1991); Martir, 782 F.2d
at 1144. These cases are incorrect and should be reversed or limited. The non-bursting bubble presumption
is likely unconstitutional and is a creation of the courts—not Congress. The BRA offers no textual support
for the proposition that a judge can lawfully consider a rebutted presumption of detention in determining
269

whether detention is warranted. The statute does not say that a rebutted presumption remains relevant,
and the presumption is not among the factors in § 3142(g). The presumption is likely unconstitutional since
it impermissibly lowers the prosecution’s evidentiary burden and substitutes for specific evidence of an
individual arrestee’s dangerousness and flight risk.

419 18 U.S.C. § 3142(e)(1); Salerno, 481 U.S. at 751.
420 See supra note 418 and accompanying text.
421 See supra note 386 and accompanying text.
422 The team searched for all the written pretrial release decisions available on LexisNexis, filtered the decisions
for those involving the presumption, and catalogued whether the decision ordered detention or release.

423 It should be noted that this research does not capture all of the pretrial Detention Hearings in the First
Circuit. It captures only written opinions that appear on LexisNexis, which may differ from the totality of bail
decisions in unknown ways. The written opinions likely understate the detention rate in presumption cases,
as they do not capture any of the cases in which an arrestee stipulated to detention. Regardless, the statistics
remain illustrative and striking.

424 For example, a recent article made multiple misstatements about the legal standard in presumption cases,
erroneously claiming (1) that the defense bears the burden of persuasion, and (2) that the arrestee must
affirmatively and definitively prove the absence of danger and flight risk. See Christine S. Scott-Hayward &
Connie Ireland, Reducing the Federal Prison Population: The Role of Pretrial Community Supervision, 34 fed. SenT’G
rep. 327, 327 (2022) (“[T]he statute sets out two situations where the presumption [of release] is reversed and,
instead, defendants must demonstrate that they pose no risk to public safety and are not at risk of failing to
appear.” (emphasis added)).

425 See, e.g., Jessup, 757 F.2d at 380–84; Dominguez, 783 F.2d at 707; Alatishe, 768 F.2d at 371; see also Bail, supra
note 62, at 404 (collecting cases).

426 Dominguez, 783 F.2d at 707.
427 The FCJC’s legal research found similar results. Looking at the set of written detention opinions from the First
Circuit referenced above, judges found the presumption rebutted in only 12% of presumption cases. They either
found that the presumption was not rebutted or made no finding in the remaining 88% of cases.

428 Verbatim Notes Form, United States v. Ernesto Morales Bacallao, No. 21-MJ-02179, (S.D. Fla. Feb. 4, 2021) (on
file with the FCJC).

429 In approximately 75% of the Detention Hearings where judges did not affirmatively find the presumption
of detention rebutted, defense attorneys contested detention—they did not waive the Detention Hearing or
otherwise agree to detention.

430 Prosecutors erroneously invoked the presumption of detention at the Detention Hearing in two cases, one
of which the judge affirmed at the hearing. Judges misapplied the presumption at detention at another two
Detention Hearings.

431 See supra note 52 and accompanying text.
432 Note, supra note 51, at 1140; see also id. (“Judges can be expected to take advantage of loopholes in
pretrial procedures.”).

433 18 U.S.C. § 3142(e)(3)(A).
434 Zunkel & Siegler, Federal Judiciary’s Role in Drug Law Reform, supra note 53, at 291.
270

435 At the Initial Appearance, the median bond imposed for presumption and non-presumption cases was
$250,000 and $100,000, respectively.

436 See supra note 388 and accompanying text.
437 Of the 198 presumption cases we observed, 176 individuals were people of color. This is a slightly higher
percentage of people of color than in the broader dataset.

438 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 50.
439 Id.
440 For example, Senator Durbin recently introduced the Smarter Pretrial Detention for Drug Charges Act, which
eliminates the presumption of detention in drug cases. Smarter Pretrial Detention for Drug Charges Act of
2021, S. 309, 117th Cong. (2021), https://www.congress.gov/bill/117th-congress/senate-bill/309/text [https://
perma.cc/B9C3-X6LL]. The bill has bipartisan sponsorship—in addition to Senator Durbin, Senators Coons,
Lee, and Wicker cosponsor the legislation. Although broader reform to the presumption is needed, this
bill takes a significant step in the right direction by addressing perhaps the most deleterious aspect of the
presumption as it currently exists.

441 Wilks, 15 F.4th at 844.
442 See, e.g., Bail, supra note 62, at 404 (quoting Dominguez, 783 F.2d at 707).
443 See supra note 418 and accompanying text.
444 18 U.S.C. § 3142(c)(2).
445 18 U.S.C. § 3142(c).
446 See, e.g., United States v. Price, 773 F.2d 1526, 1528 (11th Cir. 1985) (“The policy of [the BRA] is to permit release
under the least restrictive condition compatible with assuring the future appearance of the defendant.”).

447 United States v. Fidler, 419 F.3d 1026, 1028 (9th Cir. 2005).
448 18 U.S.C. § 3142(c)(1)(B)(xii).
449 See, e.g., S. Rep. No. 98-225, at 11, as reprinted in 1984 U.S.C.C.A.N. 3194 (acknowledging that financial conditions
such as bail bonds and solvent surety requirements pose “the potential for . . . abuse”); id. at 9, as reprinted in
1984 U.S.C.C.A.N. at 3191–92 (stating that financial conditions of release are appropriate in a narrower subclass
of cases than other conditions of release).

450 Siegler Bail Hearing written statement, supra note 9, at 10–11.
451 See, e.g., S. Rep. No. 98-225, at 16, as reprinted in 1984 U.S.C.C.A.N. at 3199 (“The purpose of this provision is to
preclude the sub rosa use of money bond to detain dangerous defendants.”).

452 Id. at 5, as reprinted in 1984 U.S.C.C.A.N. at 3187–88.
453 Id. at 9, as reprinted in 1984 U.S.C.C.A.N. at 3191–92.
454 Id. at 9, as reprinted in 1984 U.S.C.C.A.N. at 3192; see also Siegler Bail Hearing written statement, supra note 9,
at 10–11.

455 S. Rep. No. 98-225, at 10, as reprinted in 1984 U.S.C.C.A.N. at 3193. The 1966 Act did not attempt “to deal with
evaluating defendants’ dangerousness during the bail inquiry.” Mani S. Walia, Putting the Mandatory Back in the
Mandatory Detention Act, 85 ST. John’S l. rev. 177, 192 (2011); see also h.r. rep. no. 89-1541, at 3 (1966), as reprinted
271

in 1966 U.S.C.C.A.N. 2293, 2296 (“[P]retrial bail may not be used as a device to protect society from the possible
commission of additional crimes by the accused.”).

456 S. Rep. No. 98-225, at 11, as reprinted in 1984 U.S.C.C.A.N. at 3194.
457 Id.
458 Orta, 760 F.2d at 890.
459 See, e.g., United States v. Edwards, 960 F.2d 278, 283 (2d Cir. 1992) (“We also note that the [BRA’s provisions]
primarily were drafted to deemphasize the use of money bonds as a condition of bail, while simultaneously
permitting a court to consider the likelihood of flight and community safety in setting release conditions”
(citations omitted)); United States v. Holloway, 781 F.2d 124, 127 (8th Cir. 1986) (“[B]ail is not to be set at a level
that the defendant cannot make, so as to result in detention.”).

460 Holloway, 781 F.2d at 127.
461 See, e.g., Resolution: NAACP Calls for Major Improvements to Bail Bonds, naaCp (2016), https://naacp.org/
resources/naacp-calls-major-improvements-bail-bonds [https://perma.cc/5JRG-BKS4] (“[T]he [NAACP] urges
each state and municipality to adopt the Federal Bail System, to include various pretrial services . . . in lieu
of money bail.”); Ames Grawert, Expert Brief: How to Fix the Federal Criminal Justice System (in Part), Brennan CTr.
for JUST. (Jan. 2, 2020), https://www.brennancenter.org/our-work/research-reports/how-fix-federal-criminaljustice-system-part [https://perma.cc/H6DH-Z8ZZ] (“Federal pretrial release isn’t perfect, but it’s well ahead of
where many of the states are today. It also offers ongoing proof that cash bail isn’t necessary to preserve public
safety.”); daTa for proGreSS & JUST. CollaBoraTive, The end moneY Bail aCT 5 (2020), https://www.filesforprogress.
org/memos/money-bail-memo.pdf [https://perma.cc/SBD9-VQAF] (comparing “enormously successful”
state reforms to the federal system, which does not “allow a person to be held based only on an inability
to pay money bail”).

462 342 U.S. 1 (1951).
463 Id. at 5.
464 481 U.S. at 754.
465 Some circuits, however, have disagreed, finding that “when faced with a risk of flight, [a] judge is entitled to
set bail at a level he finds reasonably necessary; if defendant cannot afford bail, and must be detained pending
trial, it is not because he cannot raise the money, but because without the money the risk of flight is too great.”
United States v. Mantecon-Zayas, 949 F.2d 548, 550 (1st Cir. 1991) (quotation marks omitted) (quoting Jessup, 757
F.2d at 388–89); see also United States v. McConnell, 842 F.2d 105, 108–10 (5th Cir. 1988); United States v. WongAlvarez, 779 F.2d 583, 584 (11th Cir. 1985). This flawed reasoning demonstrates that courts are circumventing
the BRA’s protections against excessive financial conditions. If a considerable amount of money is required to
reasonably assure a person’s appearance at trial, perhaps conditions of release are insufficient to secure their
appearance and detention would be appropriate.

466 Courts are ill-equipped to distinguish between honestly imposed bonds and sub rosa detention orders.
Allowing financial conditions that result in pretrial detention, as some courts do, permits judges to bypass the
many procedural requirements of a Detention Hearing, including the right to cross examine witnesses, testify,
and present information by proffer. See 18 U.S.C. § 3142(f).

467 18 U.S.C. § 3142(b).
468 18 U.S.C. § 3142(c)(1)(B).
272

469 18 U.S.C. § 3142(c)(1)(B)(xi)–(xii).
470 See, e.g., 18 U.S.C. § 3142(c)(B)(i)–(xiv) (allowing judges to impose, inter alia, curfew, employment requirements,
and electronic monitoring); Holloway, 781 F.2d at 125:
The statute, § 3142(a), provides the judicial officer with a broad range of pre-trial
release options. These options are to be considered sequentially, in order of severity,
and the judicial officer is directed to select the option which is the least restrictive of
the defendant but which will adequately assure his appearance for further judicial
proceedings and will also protect the safety of the community.

471 Siegler Bail Hearing written statement, supra note 9, at 10.
472 S. Rep. No. 98-225, at 13–14, as reprinted in 1984 U.S.C.C.A.N. at 3196–97.
473 United States v. Spiltoro, 786 F.2d 808, 816 (8th Cir. 1986); see also United States v. Irizarry, No. 22-3028, 2022
WL 2284298, at *2 (D.C. Cir. June 24, 2022) (holding that a court erred in making “global judgments about all
defendants charged with offenses related to January 6, rather than on an individualized assessment of safety
concerns or flight risks”); United States v. Tortora, 922 F.2d 880, 888 (1st Cir. 1990) (“Detention determinations
must be made individually and . . . must be based on the evidence which is before the court regarding the
particular defendant.”).

474 John l. weinBerG & evelYn J. fUrSe, federal Bail and deTenTion handBook 29–30, praCTiSinG l. inST. (2021) (“Many
courts have a ‘standard set’ of conditions which are part of every defendant’s appearance bond. It seems likely
these include some conditions which are not necessary for specific defendants.”); see also Timothy P. Cadigan
& Christopher T. Lowenkamp, Implementing Risk Assessment in the Federal Pretrial Services System, 75 fed. pRob.
46, 47 (Sept. 2011), https://www.uscourts.gov/sites/default/files/75_2_5_0.pdf [https://perma.cc/2D9J-34EN]
(finding that unless the imposition of further conditions can “be demonstrated to increase the likelihood that
the defendant will appear in court as required and/or reduce new offenses committed . . . , then the significant
investment of pretrial release agencies and courts in these conditions and their enforcement is ineffective
and unwise.”). The overuse of conditions can also have adverse effects, particularly among low-risk arrestees.
See, e.g., Lowenkamp et al., The Development of a Pretrial Screening Tool, 72 fed. pRob. 2, 3 (Dec. 2008), https://
www.uscourts.gov/sites/default/files/72_3_1_0.pdf [https://perma.cc/V5B7-PWSS] (“[P]roviding intensive
services and supervision to low-risk offenders does little to change their likelihood of recidivism and, worse,
occasionally increases it.”).

475 See generally Michael R. Jones, Unsecured Bonds: The As Effective and Most Efficient Pretrial Release Option, preTrial
JUST. inST. (2013), http://web.archive.org/web/20160204155044/http://www.pretrial.org/download/research/
Unsecured Bonds, The As Effective and Most Efficient Pretrial Release Option - Jones 2013.pdf (finding that
unsecured bonds are just as effective at achieving public safety and court appearance as secured bonds).

476 In our study, “financial conditions” include both secured bonds and unsecured bonds.
477 BJS 2008–2010 Report, supra note 88, at 5.
478 How Federal Bail Bonds Work in the United States, how Bail BondS work (Nov. 13, 2021), https://
howbailbondswork.com/how-federal-bail-bonds-work [https://perma.cc/VJJ6-GAGZ].

479 Id.
480 Id.
481 Color of ChanGe & aClU’S CampaiGn for SmarT JUST., SellinG off oUr freedom: how inSUranCe CorporaTion have
Taken over oUr Bail SYSTem 14 (May 2017), https://www.aclu.org/sites/default/files/field_document/059_bail_
report_2_1.pdf [https://perma.cc/XJ9K-SN84].
273

482 Who Really Makes Money Off of Bail Bonds?, The aTlanTiC (May 12, 2017), https://www.theatlantic.com/business/
archive/2017/05/bail-bonds/526542/ [https://perma.cc/3V54-F46Q].

483 Allie Preston et al., Fact Sheet: Profit Over People: Inside the Commercial Bail Bond Industry Fueling America’s Cash
Bail Systems, CTr. for am. proGreSS (Jul. 6, 2022), https://www.americanprogress.org/article/fact-sheet-profitover-people/ [https://perma.cc/383U-RR74].

484 See, e.g., Joshua Page, I Worked as a Bail Bond Agent. Here’s What I Learned., The appeal (Apr. 4, 2019), https://
theappeal.org/i-worked-as-a-bail-bond-agent-heres-what-i-learned/ [https://perma.cc/9GD5-JKLT].

485 United States v. Nebbia, 357 F.2d 303 (2d Cir. 1966).
486 In this district, as well as in Miami, judges frequently used cash or corporate surety bonds in cases involving
noncitizen arrestees. Bonds are set at rates that typically prevent arrestees from securing release, thus acting
as de facto detention orders.

487 This requirement prevents those who are self-employed from acting as solvent sureties.
488 Ava Kofman, Digital Jail: How Electronic Monitoring Drives Defendants into Debt, propUBliCa (July 3, 2019), https://
www.propublica.org/article/digital-jail-how-electronic-monitoring-drives-defendants-into-debt [https://perma.
cc/A6C7-8M5W].

489 In our study, we were unable to determine how often judges required arrestees to cover the cost of their own
electronic monitoring.

490 See, e.g., Angela Hanks et al., Systemic Inequality: How America’s Structural Racism Helped Create the
Black-White Wealth Gap, CTr. for am. proGreSS (Feb. 21, 2018), https://www.americanprogress.org/issues/race/
reports/2018/02/21/447051/systematic-inequality/ [https://perma.cc/FHF7-AQSB].

491 See, e.g., id.
492 Id.
493 Id. tbl.5 (showing that Black and white households make nearly the same debt payments relative
to their incomes).

494 John H. Doyle et al., Report of the Working Committees to the Second Circuit Task Force on Gender, Racial and Ethnic
Fairness in the Courts, 1997 ann. SUrv. am. l. 117, 317–18 (1997).

495 In the secured bond cases we observed, 24% of arrestees were Black, 68% were Latino, and only 5% were
white. The arrestee’s race was marked as unknown in 3% of secured bond cases; we assume, given the high
percentage of people of color in our sample overall, that these arrestees were most likely people of color
of ambiguous ethnicity.

496 In the CSB cases we observed, 19% of arrestees were Black, and 79% were Latino. One arrestee’s race was marked
as unknown; we assume, given the high percentage of people of color in our sample overall, that this arrestee
was most likely a person of color of ambiguous ethnicity.

497 Transcript of Detention Hearing at 7–8, United States v. Ramos, No. 21-MJ-02141 (S.D. Fla. Feb. 11, 2021),
ECF No. 21.

498 18 U.S.C. § 3142(c)(2).
499 Id.

274

500 Transcript of Detention Hearing at 8, United States v. Ramos, No. 21-MJ-02141 (S.D. Fla. Feb. 11, 2021),
ECF No. 21.

501 The Cardone Report, supra note 37, at xiv.
502 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 48 (In Phase 1 of our Chicago pilot study,
“[p]rosecutors only provided a valid basis for detention in 5 percent of [Initial Appearances] and only provided
evidence to support the request in one case. . . . [D]uring Phase 2, prosecutors either explicitly cited the statute
or used the words ‘serious risk of flight’ in 82 percent of the Initial Appearances in which clients were detained
without conceding detention.”).

503 Id. at 51 (“FCJC’s courtwatching revealed that the defense waived Detention Hearings 35 percent of the
time. But after FCJC held its training [for the defense bar], the defense waived the hearing just 22 percent
of the time.”).

504 Siegler & Zunkel, Rethinking Federal Bail Advocacy, supra note 4, at 46.
505 Id.
506 Id.
507 Id.
508 Id. at 47–48.
509 Id. at 48.
510 Id. at 47.
511 Id.
512 The Administration of Bail by State and Federal Courts: A Call for Reform, Hearing Before the Subcomm. on Crime,
Terrorism, and Homeland Security of the H. Comm. On the Judiciary, 116th Cong. 3 (Nov. 14, 2019) (testimony of
Alison Siegler, Dir. Federal Criminal Justice Clinic, University of Chicago Law School).

513 Id.
514 Siegler Bail Hearing written statement, supra note 9.
515 Salerno, 481 U.S. at 755 (“In our society liberty is the norm, and detention prior to trial or without trial is the
carefully limited exception.”).

516 Federal Bail Reform Act of 2020, H.R. 9065, 116th Cong. (2020).
517 JUSTiCe roUndTaBle, TranSformaTive JUSTiCe: JUSTiCe roUndTaBle’S reCommendaTionS for The new adminiSTraTion and
117tH CoNgReSS 22–23 (2020), https://www.sentencingproject.org/wp-content/uploads/2020/11/TransformativeJustice.pdf [https://perma.cc/NF4F-C7TD].

518 Memorandum from Alison Siegler, Dir. Federal Criminal Justice Clinic, University of Chicago Law School,
to the Biden-Harris Administration, Federal Bail Priorities for the Biden-Harris Administration: Executive Branch
Policies (Dec. 7, 2020), https://perma.cc/UY3B-KEXY.

519 Siegler & Harris, supra note 23.
520 Id.
521 The students in the 2020–2021 FCJC cohort were Miami Courtwatching Site Managers Sam Bonafede, Kate
Harris, and Stephen Ferro, as well as Michael Belko, Clare Downing, Mikaila Smith, and Adam (Psi) Simon.
275

522 See supra note 96.
523 Coase-Sandor is a research institute at the University of Chicago Law School that employs economic and
quantitative approaches to the law.

524 Hearings were marked as unobserved if interns were unable to locate a courtwatcher Verbatim Notes Form for
them. For the purposes of our project, we were most interested in the final Detention Hearing, as this allowed
us to ascertain the ultimate determination over the course of the entire pretrial detention/release process.

525 The students in the 2021–2022 FCJC cohort were Team Leaders and Primary Authors Clare Downing, Stephen
Ferro, Angela Chang, and Jaden Lessnick; Project Managers Krysta Kilinski and Jacqueline Lewittes; Interview
Team Leaders Alessanrdo Clark-Ansani and Kilinski; and FCJC students Mikaila Smith and Paige Petrashko.

526 See supra Executive Summary for details. For a list of districts and circuits in which we conducted interviews,
see supra note 7.

527 Percentages may not add up to 100% due to rounding.
528 * Indicates fewer than 10 cases observed in the category. ** Indicates fewer than 5 cases observed
in the category.

529 See supra note 10 and accompanying text.
530 See supra note 35 and accompanying text. Racial disparities in law enforcement and prosecution may
themselves be a result of conscious or unconscious bias. See supra note 36 and accompanying text.

531 Transcript of Detention Hearing at 7, United States v. Ramos, No. 21-MJ-02141 (S.D. Fla. Feb. 11, 2021),
ECF No. 21.

532 When CSB cases are included in the rate of prosecutors’ detention requests, the rate at Initial Appearance
increases to 75%, the rate at Detention Hearing increases to 83%, and the overall percentage increases to 67%.

533 2020 Federal Sentencing Statistics, U.S. SenT’G Comm’n (May 2021), https://www.ussc.gov/research/data-reports/
geography/2020-federal-sentencing-statistics [https://perma.cc/K4EU-3QDZ]. The percentages include 7
categories of cases: antitrust, bribery/corruption, forgery/counterfeit/copyright, fraud/theft/embezzlement,
immigration, money laundering, and tax.

534 We did not observe any non-(f)(1) cases in Salt Lake City.
535 Because we did not observe any non-(f)(1) cases in Salt Lake City, there was a valid statutory basis for detention
at the Initial Appearance in every case we observed in that district.

536 Because we did not observe any non-(f)(1) cases in Salt Lake City, there was a valid statutory basis for detention
at the Initial Appearance in every case we observed in that district.

537 We recorded an individual as “partially represented by counsel” or “represented during part of the Initial
Appearance” if they received representation after substantive questioning by a judge about financial conditions
or other elements of their case. In contrast, we recorded an individual as “fully represented” or “represented
during the entire Initial Appearance” only if a defense attorney was present for the entire hearing and appeared
to be representing the arrestee from the beginning of the hearing. We tracked this distinction because of the
potential for arrestees to inadvertently incriminate themselves before they receive appointed counsel—a
potential that was realized in multiple cases that we observed.

276

277

This Report reveals a fractured and
freewheeling federal pretrial detention
system that has strayed far from
the norm of pretrial liberty.
To begin to remedy the many harms of
mass pretrial detention and support the
positive outcomes of pretrial release,
federal judges must shift the culture from
one prioritizing pretrial detention to one
prioritizing pretrial release.
It is incumbent upon judges to act boldly
and to be guided by data, not institutional
pressures. Ultimately, federal judges
have the power to uphold the rule of law,
to make detention prior to trial the rare
exception, and to be champions of liberty.