Skip navigation

Moving Beyond Money: A Primer on Bail Reform, Criminal Justice Policy Program, Harvard Law School, 2016

Download original document:
Brief thumbnail
This text is machine-read, and may contain errors. Check the original document to verify accuracy.

Moving Beyond Money:
A Primer on Bail Reform


Moving Beyond Money: A Primer on Bail Reform was prepared by the
Criminal Justice Policy Program (CJPP) at Harvard Law School. Substantial
research and drafting were contributed by Harvard Law School students
William Ahee, Micaela Alvarez, Jevhon Rivers, and Grace Signorelli,
who participated in the 2015-2016 Criminal Justice Fellows Seminar.
The drafting of this primer was overseen by CJPP’s executive director,
Larry Schwartztol; faculty co-directors, Prof. Carol Steiker and Prof. Alex
Whiting; and legal fellow, Anna Kastner. CJPP is grateful for generous
insights and feedback from Cherise Fanno Burdeen, Brandon Buskey, Paul
Heaton, Alec Karakatsanis, Sandra Mayson, and Timothy R. Schnacke.



The Criminal Justice Policy Program (CJPP) at Harvard
Law School conducts research and advocacy to support
criminal justice reform. It generates legal and policy
analysis designed to serve advocates and policymakers throughout the country, convenes
diverse stakeholders to diagnose problems and chart concrete reforms, and collaborates with
government agencies to pilot and implement policy initiatives.


Introduction to Bail	



A.	 Bail Basics	



B.	 Pathologies of Money Bail and the Growing Movement for Reform	



C.	 Core Legal Principles	



1.	 Federal Constitutional Principles	



2.	 Basics of State Law	



Crucial Safeguards for Pretrial Systems that Use Money Bail	



A.	 Guaranteeing Ability to Pay Determinations	



B.	 Individualizing Bail Determinations and Eliminating Bail Schedules	



C.	 Regulating or Prohibiting Compensated Sureties	


III.	Moving Beyond Money: Practical, Legal, and Policy Considerations	


Surrounding Risk-Based Systems for Pretrial Justice


A.	 Pretrial Services Agencies and Conditions of Release	




1.	 Court Date Notification	



2.	 Pretrial Supervision	






4.	 Drug Testing	


Electronic Monitoring	


B.	 Actuarial Risk Assessment	



1.	 Policy Considerations	



2.	 Constitutional Considerations	



C.	 Preventive Detention	



1.	 Constitutional Requirements	



2.	 Vital Procedural Protections	



Moving Forward	


V.	Endnotes	






being declared unconstitutional, destabilizing wellestablished practices and focusing the attention
of policymakers on the problems resulting from
money bail.1 Increasing media attention to the unjust
consequences of money bail has intensified scrutiny of
existing practice.2 All of this builds on sustained attention
from experts and advocacy groups who have long called
for fundamental reform of cash bail.3 As policymakers
across the political spectrum seek to end the era of
mass incarceration,4 reforming pretrial administration
has emerged as a critical way to slow down the flow of
people into the criminal justice system.

Bail reform presents a historic challenge – and also an
opportunity. Bail is historically a tool meant to allow
courts to minimize the intrusion on a defendant’s liberty
while helping to assure appearance at trial. It is one
mechanism available to administer the pretrial process.
Yet in courtrooms around the country, judges use the
blunt instrument of secured money bail to ensure that
certain defendants are detained prior to their trial.
Money bail prevents many indigent defendants from
leaving jail while their cases are pending. In many
jurisdictions, this has led to an indefensible state of
affairs: too many people jailed unnecessarily, with their
economic status often defining pretrial outcomes.

This primer on bail reform seeks to guide policymakers
and advocates in identifying reforms and tailoring
those reforms to their jurisdiction. In this introductory
section, it outlines the basic legal architecture of pretrial
decision-making, including constitutional principles that
structure how bail may operate. Section II describes
some of the critical safeguards that should be in place in
jurisdictions that maintain a role for money bail. Where
money bail is part of a jurisdiction’s pretrial system, it
must be incorporated into a framework that seeks to
minimize pretrial detention, ensures that people are
not detained because they are too poor to afford a
cash bond amount, allows for individualized pretrial
determinations, and effectively regulates the commercial
bail bond industry.

Money bail is often imposed arbitrarily and can result in
unjustified inequalities. When pretrial detention depends
on whether someone can afford to pay a cash bond,
two otherwise similar pretrial defendants will face vastly
different outcomes based merely on their wealth. These
disparities can have spiraling consequences since even
short periods of pretrial detention can upend a person’s
employment, housing, or child custody. Being jailed
pretrial can also undercut a defendant’s ability to mount
an effective defense. As these outcomes accumulate in
individual cases, improper use of money bail
can accelerate unnecessarily high rates of incarceration
and deepen disparities based on wealth and race
throughout the criminal justice system. Detaining
unconvicted defendants because they lack the wealth to
afford a cash bond also violates the Constitution.

Section III addresses the legal and policy considerations
relevant to eliminating the use of money bail. It describes
leading reform strategies, highlights competing policy
considerations implicated by these strategies, and
elaborates constitutional principles that should guide
policy reform. It focuses on a set of reforms that many

A recent wave of advocacy has created national
momentum for fundamentally rethinking how pretrial
decision-making operates. Litigation across the country
has resulted in the bail systems of several jurisdictions



that should structure a reform strategy. It aims to help
translate growing momentum for bail reform into on-theground change by providing policymakers and advocates
with guidance on what alternatives are available and how
they might be implemented.

Bail reform presents a
historic challenge – and also
an opportunity.

When a person is arrested, the court must determine
whether the person will be unconditionally released
pending trial, released subject to a condition or
combination of conditions, or held in jail during the
pretrial process. Any outcome other than unconditional
release must be justified by a finding of a significant
risk that the defendant will not appear at future
court appearances or will commit a serious crime in
the community during the pretrial period.5 In some
very rare instances, a judge will determine that there
is no condition or combination of conditions that can
adequately address those risks; in those instances, a
judge is deciding that the person is non-bailable and
should be subject to pretrial detention.

advocates have advanced as a way to move to a “riskbased” system of pretrial decision-making. In particular, it
focuses on three aspects of such a system: the expanded
use of pretrial services agencies and the tools those
agencies employ to supervise pretrial defendants in the
community; actuarial risk assessment instruments, which
provide judges with a quantitative model for forecasting
the risk that particular defendants will fail to appear for
trial or will commit a serious crime during the pretrial
period; and the limited use of preventive detention. This
primer does not prescribe a one-size-fits-all package of
pretrial reforms. Indeed, some of the potential reforms
raise knotty legal and policy questions. Answering
those questions will require jurisdictions to assess
local circumstances and needs and make fundamental
judgments among competing policy values in order to
craft appropriate policies. While this primer does not
propose a uniform model of bail reform, it can guide
advocates and policymakers through the considerations

If, however, the judge decides that the person may be
released prior to their court date, then the person is
bailable and several options are available. The judge can
release the person on their own personal recognizance,
meaning that the person promises to reappear for







M O N E TA R Y B O N D /


scheduled court dates in the future. Alternatively, the
judge may conditionally release the person such that
their continued freedom is subject to certain nonmonetary conditions, such as pretrial supervision or
enrolling in a substance treatment program.

Jailing people on the basis
of what amounts to a
wealth-based distinction
violates well-established
norms of fairness as well as
constitutional principles.

The court can also conditionally release the person
by imposing a secured or unsecured bond. A secured
bond typically allows a defendant to be released only
after he pays the monetary amount set by the court,
though a bond may also be secured by the defendant’s
property (such as a house). When bond is unsecured, the
defendant will owe the unsecured bond amount if he
fails to appear in court.
When secured money bonds are used, the amount of
money set by the court that a person is obligated to pay
as a condition of his release is that person’s cash bail or
money bail.6 The person may be released upon posting
a bond, or in some cases 10 percent of the total bond
amount. Sometimes the person may be able to make
that 10 percent payment directly to the court, which will
often return the bond payment if the defendant makes
all required pretrial appearances. But in many instances,
if the person does not have enough money to pay the
money bail set by the court, a bail bonds agent, also
known as a surety, may make the payment for them
via a surety bail bond. If the person cannot make the
payment, either personally or through a surety, they will
remain incarcerated based on their inability to pay the
money bail.

jail pretrial solely because they are unable to pay a cash
bond, and most of these people are among the poorest
third of Americans.10 National data from local jails in 2011
showed that 60% of jail inmates were pretrial detainees
and that 75% of those detainees were charged with
property, drug or other nonviolent offenses.11 In fiscal
year 2014 alone, local jails admitted 11.4 million people
and the nationwide average daily population included
467,500 pretrial defendants.12
The core critique of money bail is that it causes
individuals to be jailed simply because they lack the
financial means to post a bail payment. Jailing people on
the basis of what amounts to a wealth-based distinction
violates well-established norms of fairness as well as
constitutional principles. It can also lead to significant
levels of unnecessary jailing, which imposes intensely
negative consequences on individuals, communities, and
the justice system.


Unnecessary pretrial jailing carries significant human
costs. The experience of even short terms of pretrial
detention can be devastating for an individual. Although
“jail operations vary considerably, from local detention
facilities in rural America that hold three or four inmates
to the jail systems of Chicago, Los Angeles, or New
York that hold upwards of 20,000 inmates...regardless
of facility size, a consistent theme in the extant
literature is that jails have always been characterized by
overcrowding, resource limitations, litigation, suicide and
violence.”13 Jails “collect and concentrate individuals at
high risk of violence, substance abuse, mental illness, and
infectious disease.”14 The living and sleeping conditions
expose inmates to unsafe and unsanitary conditions.
A former jail inmate in Baltimore described conditions
including “people that are getting skin bacterial

Reliance on money bail has been shown to unfairly
disadvantage impoverished defendants and to
undermine community safety. The money bail system
results in presumptively innocent people, who have been
determined eligible for release, remaining incarcerated
simply because they do not have enough money to
afford the cash bond. For instance, a 2013 review of New
York City’s jail system showed that “more than 50% of
jail inmates held until case disposition remained in jail
because they couldn’t afford bail of $2,500 or less.”7
Most of these people were charged with misdemeanors.8
Of these non-felony defendants, thirty-one percent
remained incarcerated on monetary bail amounts of
$500 or less.9 Nationwide, 34% of defendants are kept in



less likely to be able to post money bail as a condition of
release.26 Because pretrial detention has such a profound
effect on later-in-the-case outcomes, racial disparities in
the application of cash bail may reinforce or exacerbate
larger inequalities in rates of incarceration.

diseases...they have measles, scabies, lice, fleas.”15 Jails,
traditionally designed for short periods of detention,
often provide inadequate healthcare, activities, and
programming.16 Serious mental illness affects jail
inmates at rates “four to six times higher than in the
general population,” yet “83 percent of jail inmates
with mental illness did not receive mental health care
after admission.”17 According to the Bureau of Justice
Statistics, suicide has been the leading cause of death in
jails every year since 2000.18

Unnecessary jailing also undermines community safety.
Statistical studies have shown that similarly situated,
low-risk individuals who are detained pretrial, even for
short periods, are actually more likely to commit new
crimes following release.27 This seemingly counterintuitive
outcome reflects the profoundly destabilizing effects of
even short durations of pretrial detention. Further, the
inability to post money bail may induce innocent people
accused of relatively low-level crimes to plead guilty,
simply so they can be released.28 In the case of certain
offenses, this endangers communities, as the person
actually responsible for committing the crime remains
free, yet law enforcement is no longer investigating
them.29 Unnecessary detention is also counterproductive
from the perspective of guaranteeing appearance at trial.
Studies show that those who remain in pretrial detention
for longer than 24 hours and are then released are less
likely to reappear as required than otherwise similar
defendants who are detained for less than 24 hours.30

Pretrial detention also impacts many aspects of an
individual’s life, including the outcome of his criminal
case. Even a short period of pretrial detention can have
cascading effects on an individual. Pretrial detention can
threaten a person’s employment, housing stability, child
custody, and access to health care.19 These destabilizing
effects may explain the negative impact that pretrial
detention has on the prospects of a defendant’s case.
Defendants who are detained for the entire pretrial
period are “over four times more likely to be sentenced
to jail and over three times more likely to be sentenced
to prison than defendants who were released at some
point pending trial.”20 In addition to a greater likelihood
of receiving a jail or prison sentence, defendants who are
detained pretrial face longer sentences once convicted.
The sentences of those who are detained pretrial are
“significantly longer – almost three times as long for
defendants sentenced to jail, and more than twice as
long for those sentenced to prison.”21 Recent studies have
identified a causal link between pretrial detention and
adverse case outcomes.22 One of those studies analyzed
over 375,000 misdemeanor cases filed between 2008
and 2013 in Harris County, Texas, and concluded that
“misdemeanor pretrial detention causally affects case
outcomes.”23 The study included a regression analysis
that controlled for “a wide range of confounding factors”
including demographics, criminal history, and wealth, and
found that “detained defendants are 25% more likely
than similarly situated releasees to plead guilty.”24

Policymakers in many states around the country
have embraced the call for bail reform. For instance,
in 2013, Colorado overhauled its bail statutes to
discourage the use of money bail and to encourage the
use of risk assessment tools when determining which
defendants should be released subject to supervision
by a pretrial services agency.31 In August 2014, New
Jersey passed legislation to shift from a money-based to
a risk-based system.32 Connecticut’s governor recently
announced a proposal for bail reform which included a
prohibition on setting money bail for anyone charged
with a misdemeanor.33
Other jurisdictions have been motivated to take
legislative action based on court rulings. In November
2016, New Mexico voters will decide on a constitutional
amendment that would authorize limited preventive
detention and permit those held on a cash bond to
petition the court for relief when they cannot afford
bail.34 The amendment was proposed in response to a
2014 New Mexico Supreme Court opinion, which held
that a trial judge erred in using a high bond amount to
detain a murder defendant prior to his trial when less
restrictive conditions of release would protect public

The current money bail system also exacerbates racial
disparities in the criminal justice system. Money bail
inherently discriminates against poor defendants, who
are by definition less likely to be able to cover bond. Due
to well-established linkages between wealth and race,25
money bail will often result in increased rates of pretrial
detention for Black and Latino defendants. Studies have
shown that Black and Hispanic defendants are more
likely to be detained pretrial than white defendants and



amounts (or other conditions of release43) to the amount
needed to incentivize particular defendants to appear
at court proceedings. In practice, however, the courts
have not applied this Eighth Amendment principle in a
way that has meaningfully constrained the use of bail.
The Supreme Court has not substantially addressed these
principles since deciding Stack v. Boyle in 1951.

safety.35 Across the country, a recent wave of civil rights
lawsuits filed in federal court have led localities to reform
their practices by ending the use of secured money bail in
certain situations for arrestees who are unable to pay.36


Although the Eighth Amendment is the only
constitutional provision to explicitly address bail, due
process and equal protection principles also apply to
the pretrial deprivation of liberty. Due process principles
govern the circumstances under which any person can
be deprived of their liberty, including through pretrial
detention. The Supreme Court has emphasized that
“[i]n our society liberty is the norm, and detention
prior to trial or without trial is the carefully limited
exception.”44 Due process has a substantive component
and a procedural one. Substantive due process “forbids
the government to infringe certain ‘fundamental’
liberty interests at all, no matter what process is
provided, unless the infringement is narrowly tailored
to serve a compelling state interest.”45 This means that,
as a threshold requirement, any system providing for
pretrial detention must be narrowly tailored to the
compelling government interest put forward to justify
detention. Where that substantive requirement is met,
a deprivation of liberty must also reflect procedural
safeguards designed to balance public and private
interests and to minimize the risk of error.46 The contours
of these due process requirements are discussed in more
detail in Section III.C.

A starting point for effectively reforming money bail
is understanding the existing legal frameworks that
govern pretrial decision-making. This section begins by
describing some of the baseline federal constitutional
requirements relevant to bail. Next, it describes the
role that state constitutions play in defining how bail
operates. Finally, this section discusses some of the basic
elements of state statutory law and suggests resources
for assessing whether a particular state’s laws are
consistent with best practices.
1. Federal Constitutional Principles
Several constitutional provisions establish basic
protections in the pretrial setting. As a threshold
matter, the Fourth Amendment’s protection against
unreasonable seizures guarantees that an arrestee
receive a probable cause determination by a neutral
magistrate within 48 hours of being arrested.37
The Eighth Amendment prohibits the use of “excessive
bail,”38 but it does not define what “excessive” means or
specify when bail should be granted.39 In Stack v. Boyle,
the Supreme Court provided some guidance in assessing
whether bail is excessive. Starting from the premise
that the “traditional right to freedom before conviction
permits the unhampered preparation of a defense,
and serves to prevent the infliction of punishment
prior to conviction,” the Court defined “excessive” as
bail “set at a figure higher than an amount reasonably
calculated” to “assure the presence of the accused.”40
Significantly, the Court tied the question of whether a
bail determination is excessive to the purpose of bail.
As the Court explained, the purpose of bail is to help
assure the presence of that defendant at subsequent
proceedings.41 “Since the function of bail is limited, the
fixing of bail for any individual defendant must be based
upon standards relevant to the purpose of assuring the
presence of that defendant.”42 This functional analysis
of bail suggests that the Eighth Amendment imposes a
sliding scale, linking constitutionally permissible bond


The use of money bail also implicates equal protection
principles, which forbid courts to impose jail or other
adverse consequences on the basis of a defendant’s
indigence. The Supreme Court has repeatedly reaffirmed
that “[t]here can be no equal justice where the trial a
man gets depends on the amount of money he has.”47
In Bearden v. Georgia, the Supreme Court invalidated
the automatic revocation of an indigent defendant’s
probation on the basis of non-payment of a fine,
explaining that to “deprive [a] probationer of his
conditional freedom simply because, through no fault
of his own he cannot pay [a] fine...would be contrary to
the fundamental fairness required by the Fourteenth
Amendment.”48 Lower courts have applied this principle
to the bail context.49


may present challenges to reforming or eliminating
money bail.62 On the other hand, a statute outlining a
robust pretrial services program,63 or limiting the influx of
arrestees by encouraging citations in lieu of arrest,64 may
prove useful in reducing a state’s reliance on money bail.

2. Basics of State Law
a. State Constitutional Provisions
Most state constitutions fall into one of two categories:
•• Right to bail: Most state constitutions include a
provisions guaranteeing a right to bail. A typical rightto-bail provision states: “all persons shall be bailable
by sufficient sureties, unless for capital offenses,
where the proof is evident, or the presumption
great.” This common formulation, however, has been
subject to varied interpretations.50 In states where
courts have interpreted the word “shall” to require
an absolute right to bail, all defendants (except in
capital cases) are eligible for release and defendants
are only detained in practice if they are unable to
pay the monetary bond amount set.51 In other states,
despite employing the same or substantially similar
language, the word “bailable” and the “sufficient
sureties” clause have been interpreted to preserve the
court’s discretion in extending bail.52 In these states,
non-capital defendants are eligible for bail but the
court may always deny bail if it determines that no
amount of surety can prevent a defendant’s flight or
dangerousness to the community.53 In a few states,
this interpretation has been codified in the state
constitution.54 Additionally, in at least one state, the
court has interpreted the constitution to mean that
the court can revoke the right to bail if a defendant
violates a condition of release.55

There are resources available to advocates or
policymakers seeking a comprehensive overview of the
terrain that state law should cover in the pretrial context.
The American Bar Association’s Standards for Criminal
Justice: Pretrial Release (“ABA Standards”) provides
guidance on the core principles that should structure
a state’s pretrial justice framework.65 An extensive
treatment of the legal considerations and historical
background surrounding pretrial issues is available in
Timothy R. Schnacke’s “Fundamentals of Bail: A Guide
for Pretrial Practitioners and a Framework for American
Pretrial Reform” which was published by the U.S. Justice
Department in 2014.66

•• No explicit right to bail: Nine state constitutions
mirror the language of the U.S. Constitution and only
prohibit the use of excessive bail.56
b. State Statutory Provisions
In most states, provisions governing bail appear in the
statutory code, the rules of criminal procedure,57 or court
rules.58 In some states, there is a specific chapter of the
code devoted to bail,59 while in other states, relevant
provisions are scattered throughout the code.60 For
instance, the penal code itself may specify minimum bail
amounts for certain offenses.61
Certain features of a state’s law of bail can entrench the
use of money bail and impede reform, while others may
facilitate change. For example, a statute encouraging the
use of an offense-based bail schedule or bail minimums





There are a variety of ways that states can limit the
harms of money bail or eliminate the use of money bail
almost entirely. This section describes strategies for
mitigating the harmful effects of money bail. Examples
of such reforms include guaranteeing meaningful ability
to pay determinations, eliminating bail schedules, and
regulating commercial sureties. The reforms outlined
in this section are each powerful tools for addressing
some of the worst harms of money bail; however, they
all rest on the premise that money bail is being used at
least in some circumstances. Any reforms should reflect
the principle that pretrial detention should be reduced
except where strictly necessary.

determining factor in whether any particular defendant
is released or detained.
The Supreme Court has provided some guidance on
what an ability-to-pay determination should entail. In
Turner v. Rogers, a case involving unpaid child support
obligations, the Court held that jailing a defendant
without inquiring into his financial status “violated the
Due Process Clause.”69 In reaching its holding, the Court
noted certain procedures that, taken together, create
“safeguards” that can “significantly reduce the risk of
an erroneous deprivation of liberty” in the nonpayment
context.70 These safeguards included:
(1) notice to the defendant that his “ability
to pay” is a critical issue in the contempt
proceeding; (2) the use of a form (or the
equivalent) to elicit relevant financial
information; (3) an opportunity at the hearing
for the defendant to respond to statements
and questions about his financial status, (e.g.,
those triggered by his responses on the form);
and (4) an express finding by the court that the
defendant has the ability to pay.71

If jurisdictions intend to impose money bail as a condition
of release, it is critical to ensure that courts inquire into the
defendant’s ability to pay any monetary sum imposed. The
Supreme Court has held that a person may not be jailed
based on his inability to make a monetary payment unless
the court has made an inquiry into the person’s ability to
pay and determined that non-payment was willful or that
no other alternative measure will serve the government’s
legitimate interests.67 Though elemental, this principle is
violated routinely in jurisdictions all over the country.68
While there are undoubtedly complex questions about
how to structure pretrial decision-making, a clear first
principle should be that wealth should not be the


In the bail context, an ability-to-pay determination with
substantially similar safeguards would ensure that people
are not held in jail solely as a result of their inability to
pay money bail. Although the Supreme Court has not
stated exactly what procedures are required, an abilityto-pay determination during a bail hearing should
include the following procedures:


•• Notice to the defendant that bail determinations
must be individualized. A defendant should be
notified that his ability to pay may be a critical
consideration in setting the amount of bail.

would ensure that money bail is only used where it can
facilitate release by realistically incentivizing appearance.


•• Use of a standard form. Courts should use a standard
form setting out a defendant’s income, assets,
financial obligations, and receipt of public benefits, or
other financial information relevant to gauging ability
to pay.72

Jurisdictions throughout the country use bail schedules
to determine the amount of money bail that will be
applied to certain categories of offenses. Generally,
a bail schedule will list particular offenses or offense
types (e.g., various classes of misdemeanor or felony)
and assign a specific dollar amount or dollar range.
Jurisdictions may embrace bail schedules as a tool of
efficiency or because they provide uniformity along
certain dimensions (that is, defendants accused of the
same offense will have the same bond amount applied
to them). Bail schedules present another benefit: by
creating a rigid framework for bail determinations, they
prevent decision-makers from directly discriminating
on the basis of suspect characteristics, like race.
But by setting out a simple matrix of offenses and
corresponding dollar amounts, bail schedules do not
allow for meaningfully individualized considerations of a
defendant’s circumstances. Bail schedules are often used
to set cash bond prior to a defendant’s first appearance
before a judge or magistrate, precluding judges from
determining a defendant’s ability to pay or tailoring the
amount of the money bond to the defendant’s risk of
failing to appear.78

•• Presumptions about indigence or inability to pay
money bail. At a certain threshold, a defendant
should be presumed indigent and therefore unable
to pay money bail as a condition of release. Such
presumption may be appropriate where, for example,
a defendant’s income is below a certain threshold,
such as income at or below 125% of the Federal
Poverty Level.73
•• Clearly articulated standards and operative terms.
Terms such as “ability to pay” or “indigence” should
be clearly defined by court rules or statute.
•• The right to counsel. The right to counsel at the bail
determination is necessary to ensure that defendants
are not unnecessarily detained prior to trial.74
According to a 2011 national survey, “only ten states
guarantee representation at the initial assessment of
bail at an initial appearance.”75
•• A hearing on the record. A bail hearing on the
record will ensure proper procedures are met and
give the defendant an opportunity to contest a
bail determination.76

Bail schedules may be mandatory or advisory and may
be set at the state or local level.79 Once bail schedules
are in place, however, they often become de facto law
even if they are not formally mandatory. For example,
in Alabama, the bail statute states that “[t]he amount
of bail shall be set in the amount that the judicial officer
feels, in his or her discretion, is sufficient to guarantee the
appearance of the defendant.”80 But judges also have the
option of using a bail schedule that the Alabama Supreme
Court or the local judge has prescribed.81 Although the
bail schedule adopted by the Alabama Supreme Court
notes that “courts should exercise discretion in setting
bail above or below the scheduled amounts,”82 in practice
this has not always occurred. In a lawsuit challenging bail
practices in the City of Clanton, Alabama, a federal judge
found that the Clanton Municipal Court did not deviate
from a generic bail schedule and that indigent defendants
who could not post bail were forced to wait up to a week
until they received an individualized bail determination.83

•• Right to prompt review. The right to promptly seek
review of a bail determination will also ensure that
defendants who are unable to pay money bail are not
unnecessarily jailed.77
Much of the information about a defendant’s ability to
pay may already be collected when the court determines
whether the person qualifies for court-appointed counsel.
Such financial information is routinely obtained within
minutes from arrestees under penalty of perjury. Drawing
on that information-collecting process will be crucial in
order to allow prompt ability-to-pay determinations to
take place. Having an ability-to-pay determination with
these safeguards would ensure that judges set money
bond only in an amount that a defendant can afford. This



By setting out a simple matrix of offenses and corresponding dollar
amounts, bail schedules do not allow for meaningfully individualized
considerations of a defendant’s circumstances.

conditions and the defendant’s flight risk,
and should never be set by reference to a
predetermined schedule of amounts fixed
according to the nature of the charge.89

Some states, rather than require or authorize the
creation of bail schedules, will set minimum bail amounts
for certain offenses by statute. Statutory bail minimums
also preclude judges from making individualized bail
determinations. For example, in Alaska, a judge must
impose a minimum cash bond of $250,000 for persons
charged with offenses involving methamphetamines
who have been previously convicted of possession,
manufacture, or delivery of the drug.84 The judge can
reduce this amount only if the defendant demonstrates
that he or she did not stand to gain financially from the
methamphetamine involvement and only participated
as an aider or abettor.85 These standard amounts have
no relation either to the amount necessary to ensure
appearance or the individual defendant’s ability to pay.

Individualized determinations of appropriate bail
amounts should be seen as a baseline precondition in
any system using money bail. It reflects best practices as
well as foundational constitutional requirements.90

Commercial sureties play a central role in the pretrial
procedures of many jurisdictions. A commercial surety,
or bail bond agent, purports to guarantee a defendant’s
appearance by promising to pay the financial condition
of a bond if the individual does not appear for court. Bail
bond agents are usually licensed by a state and the bonds
are underwritten by an insurance company. Bond agents
not only charge a non-refundable fee for their service,
but usually require the defendant or his friends or family
to provide collateral for the full amount of the financial
condition. Between 1994 and 2004, the percentage of
defendants released on commercial sureties increased
from 24% to 42%.91 In some circumstances, the existence
of commercial sureties will act as a safety valve against
unnecessary detention by enabling some defendants
who could not afford a full bond amount to avoid pretrial

Bail schedules are fundamentally inconsistent with
individualized decision-making. Money bail may serve
only one legitimate role: to incentivize someone to
return to court as required.86 To do that, it must be
individualized to the defendant. Just as a fixed bail
amount may be too high for a poor defendant to
post (and therefore will have the effect of imposing
pretrial detention), that same bail amount may be
so inconsequential to a wealthy defendant that
the prospect of forfeiting bail will not function as
a meaningful incentive to appear for trial. The ABA
Standards emphasize the importance of properly
individualized determinations when setting money bail.
Under those standards, money bail may be “imposed
only when no other less restrictive condition of release
will reasonably ensure the defendant’s appearance
in court.”87 Cash bonds “should not be set to prevent
future criminal conduct during the pretrial period.”88
Significantly, the ABA Standards state:

But commercial sureties have also been subject to
strong criticisms. Commercial sureties can deepen the
pathologies of money bail by devolving pretrial decisions
from courts to private companies. For many defendants,
pretrial release or detention will depend on whether
a commercial surety posts their bond. Ironically, some
bail bond agents will not post bail for defendants with
low money bail amounts because it is less lucrative for
the bail bond company than posting bail for defendants

Financial conditions should be the result of
an individualized decision taking into account
the special circumstances of each defendant,
the defendant’s ability to meet the financial



with high cash bonds.92 The effect of those incentives may
be that defendants with lower bond amounts – typically
defendants a court perceives to present lower pretrial risk –
remain detained because they cannot pay a cash bond and
commercial sureties do not view them as worthwhile clients.
Moreover, commercial surety companies face frequent
criticism for inadequate training and aggressive pricing
practices.93 Private sureties are also notorious for physically
and economically coercive practices and exacerbating the
potential for violence, bribery, and corruption in the bail
context.94 The prominence of compensated sureties is,
from a global perspective, an outlier – outside the U.S.,
only the Philippines allows the operation of a commercial
surety industry.95
Some states, such as Kentucky and Illinois, have passed
legislation to ban the bail bonds industry entirely.96
States can also pass legislation that reduces the role of
compensated sureties by allowing defendants to pay
deposits directly to the court, instead of bond agents. For
example, in Massachusetts, trial court judges now routinely
set a money bail amount as a percentage of the surety
required so that defendants can pay a 10% deposit directly
to the court, rather than a bond agent, and have the
deposit returned at the resolution of their case – a practice
that effectively eliminated the bail bonds industry.97




R I S K- B A S E D S Y S T E M S F O R

The reforms described above assume the continued
use of money bail and propose safeguards to help
mitigate the worst harms that flow from that system.
An alternative approach is to re-conceptualize the
pretrial process in a way that replaces money bail with
tools better suited to further the legitimate purposes
of pretrial decision-making. If cash bonds serve to
incentivize defendants to appear for trial, are there
alternative practices that more effectively and fairly
reduce the risk of pretrial flight? Similarly, to the extent
that some judges use high cash bonds as a sub rosa
means of detaining pretrial defendants whom
they consider dangerous, are there mechanisms that
promote community safety in a more equitable and
transparent way?

2.	 Quantitative risk assessment determinations that use
algorithms to assign a risk category that judges can
incorporate into pretrial decision-making

One model for displacing the role of money bail is a riskbased approach to pretrial justice. A risk-based model
proceeds from the presumption that pretrial defendants
should be released. When that presumption is overcome
by a significant risk that the defendant will fail to appear
or commit a serious crime, a court should impose the
minimally invasive condition necessary to address that
risk.98 Many champions of bail reform have called for riskbased system composed of three elements:


3.	 Limited use of preventive detention
This section discusses each of those elements in turn,
addressing practical, legal, and policy questions. While
this primer takes no position on whether jurisdictions
should adopt those elements, it does seek to highlight
some of the important considerations that a jurisdiction
ought to consider in weighing potential approaches
to bail reform. The discussion below seeks to bring the
relevant considerations to the surface.

A key element of a risk-based model is the strategic,
evidence-based use of pretrial services. Pretrial
services can take many forms, but it generally refers
to the bundle of interventions that will ensure that
an individual defendant appears at trial and is not
rearrested during the pretrial period. Pretrial services
are thus an indispensable element of a system that
replaces money bail. Instead of relying on cash bonds
and pretrial detention, pretrial services offer an array

1.	 Pretrial service agencies that use a variety of nondetention-based interventions to ensure appearance
at trial and promote community safety



Washington, DC offers an example of a busy and
complex court system that has virtually eliminated
money bail and maintained positive pretrial
outcomes. The city has a high-functioning pretrial
services agency that facilitates pretrial release
and detention decisions and provides appropriate
levels of supervision and treatment for released
defendants that do not rely on money bail.99 Nearly
88% of defendants in Washington D.C. are released
with non-financial conditions.100 This nearly cashless bail system has proven successful in maintaining
public safety and the integrity of the court system.
Between 2007 and 2012, 90% of released defendants
have made all scheduled court appearances and over
91% were not rearrested while in the community
before trial.101 Ninety-nine percent of released
defendants were not rearrested on a violent crime
while in the community.102 At the same time, the

D.C. bail system has allowed defendants awaiting
trial to remain in their communities for the entirety
of their pretrial period; 88% of released defendants
remained in the community while their cases
were pending without a revocation of release or
supervision.103 Of course, the DC system has certain
unique characteristics: all of its judges operate in a
single courthouse, which may reinforce a culture of
pretrial release; it has an extremely high-functioning
public defender system, which helps ensure proper
representation at pretrial detention hearings; and
its pretrial services agency receives funding from
the federal government. Still, the D.C. bail system
demonstrates that, with alternative methods to
manage risk, money can be virtually eliminated from
the bail process without negatively affecting court
appearance rates or public safety.

to reduce its pretrial jail population by 27% without
negative consequences for public safety.109

of less restrictive tools that are likely to produce better
outcomes for the jurisdictions in which they operate. For
these reasons, expanded pretrial services have been an
important component of recent state-based efforts at
bail reform.

While some jurisdictions attach pretrial services
to probation or other supervisory departments or
include it in the role of the courts, the best practice
is to create a separate agency to administer pretrial
service. The National Association of Pretrial Services
Agencies (NAPSA) has emphasized the importance of
independence to the critical role of the pretrial services
agency, especially in light of the “unique mission and
role of pretrial services, which in some instances may not
be congruent with the mission of the host entity” if the
agency is housed within another department.110 NAPSA’s
Standards on Pretrial Release reiterates that “although
a pretrial services program may be organizationally
housed within a probation department, sheriff’s office,
or local corrections department, it should function as an
independent entity.”111

Some states already authorize the creation of a pretrial
services agency that is empowered to screen defendants,
make recommendations regarding detention or bail,
and provide services such as treatment for mental health
conditions and substance use disorder.104 In recent years,
six states – Colorado, Hawaii, Nevada, New Jersey,
Vermont, and West Virginia – have passed legislation to
create or bolster pretrial services agencies.105 In Colorado,
for example, pretrial services are authorized by state
statute and administered at the county level.106 The Mesa
County, Colorado Pretrial Services Agency has been held
up as a national model. The agency uses risk assessment
tools to determine a defendant’s risk of failure to
appear or re-arrest and supervises defendants who are
released prior to trial.107 The lowest level defendants
receive phone calls reminding them of their court dates,
while other defendants may be required to meet with
their pretrial services officer as often as once a week.108
From July 2013 to December 2014, the county was able

Any state seeking to mandate the use of pretrial service
agencies, of course, must contend with the budgetary
implications of establishing or expanding a freestanding
criminal justice agency. The costs involved will vary
depending on the needs of particular jurisdictions. Though
it is not possible to forecast those costs for all jurisdictions,



of their upcoming court date improved appearance
rates.115 These studies highlight how notifications had
varying degrees of effectiveness depending on the
type of contact. The different approaches included:
(1) having people call the defendants; (2) using an
automated calling system to contact defendants; (3)
sending letters or postcards; and (4) a combination
of the above. While it is not possible to make a direct
comparison between the approaches because the studies
employed different methodologies, the results indicate
that all effectively reduced failures to appear in court. In
Multnomah County, Oregon, simply calling defendants
dramatically decreased rates of failure to appear. The use
of automated telephone call reminders, referred to as
“Court Appearance Notification System,” was associated
with a 41 percent decrease in failures to appear among
defendants who successfully received a phone call.116
Similarly, Coconino County, California117 and Jefferson
County, Colorado118 reduced their failure to appear rates
significantly through phone calls by volunteers.119

Electronic monitoring should
only be used as an alternative
to incarceration, not as a way
to monitor low or medium-risk
defendants whose detention
would clearly not be justified.

in many instances those start-up and operational costs
may be counterbalanced by the savings that flow from
decreased detention and improved pretrial outcomes,
including fewer new crimes being committed.112
Pretrial services may employ an array of interventions to
ensure appearance and protect public safety. Most of
these interventions operate on a continuum of liberty
restrictions from the most minor, such as monthly
phone calls with a pretrial services agency, to the most
restrictive, such as electronic monitoring or house arrest.
As discussed below, more restrictive interventions
may raise significant constitutional considerations.113
Depending on how they are implemented, pretrial
conditions of supervision may implicate the prohibitions
against unreasonable searches, deprivations of liberty
without due process, or excessive bail.114 For both legal
and policy reasons, it is crucial that the least restrictive
alternatives to detention be imposed in order to ensure
a defendant returns to court or avoids re-arrest during
the pretrial period.

2. Pretrial Supervision
Pretrial supervision refers to the practice of maintaining
regular contact with defendants, often to facilitate,
support, and monitor their compliance with their
pretrial release conditions. There is no consensus
definition of what pretrial supervision entails, and
the requirements and practices referred to as pretrial
supervision vary widely.120 The primary mechanisms
used to supervise pretrial defendants include in-person
contact, home contact, telephone contact, contact
with those knowledgeable about the defendant’s
situation, regular criminal history checks, and also court
date reminders.121 The most recent studies that focus
on regular communication suggest that it may reduce
rates of failure to appear and re-arrest compared
with defendants released without supervision. A 2006
study in Philadelphia found that regular supervision
substantially reduced rates of re-arrest and failure to
appear,122 and a study by the Laura and John Arnold
Foundation also found that moderate-to-high-risk
defendants who were regularly supervised were
more likely to appear in court and less likely to be rearrested.123 Controlling for relevant variables, moderaterisk defendants who were supervised missed court dates
38% less than unsupervised defendants. Supervised
high-risk defendants missed court appearances 33% less
often.124 The study found that supervision decreased rearrest rates for medium and high risk defendants.125

Jurisdictions may consider a broad range of potential
conditions. Without attempting to exhaustively
catalogue every condition a jurisdiction may employ,
each of the following sub-sections describes a
potential pretrial intervention, highlighting practical
and constitutional considerations that should inform
decisions about whether to deploy those interventions.
1. Court Date Notification
The least invasive tool to ensure that defendants show
up to court is also one that has been shown to be quite
effective: reminders. Studies over the past three decades
have demonstrated that simply reminding defendants



a neutral restriction that should simply be imposed as
a matter of course; it restricts liberty in profound but
sometime subtle ways. Electronic monitoring can be
intrusive and deleterious to a defendant’s relationships
and employment.131 In a survey of probation officers
and convicted people who were given an electronic
monitoring device in Florida, both groups described a
negative impact on the individual’s relationships and
employment.132 Those who had to wear the electronic
monitoring device told researchers that the device gave
them a “sense of shame” and a feeling of being “unfairly
stigmatized.”133 Forty-three percent of those who wore it
believe that electronic monitoring had a negative impact
on their partners because of the inconvenience it created.134
Probation officers and those who wore the devices were
unanimous in their belief that wearing an electronic
monitoring device made it difficult to hold a job.135

3. Electronic Monitoring
Electronic monitoring is a tool to track a defendant’s
movements in order to deter him from absconding or
committing a serious offense. Electronic monitoring has
been used for the past twenty years and its popularity
is growing.126 From 2000 to 2014 the use of electronic
monitoring grew by 32 percent.127
Existing research on the efficacy of electronic monitoring
has documented mixed results. This is probably
because increased monitoring also increases the rate
at which violations are detected, and because of the
comparatively high-risk population that currently
receives electronic monitoring.128 Electronic monitoring
as a condition of pretrial release has not been shown to
reduce pretrial failure.129 However, there are significant
limitations to the studies, which examined programs that
may have already been using electronic monitoring for
more high-risk defendants – defendants who may not
otherwise have been released if not for the availability
of this alternative to detention. Electronic monitoring
may have potential to reduce unnecessary detention
for higher risk defendants with an acceptable level of
risk. Electronic monitoring may be a powerful tool for
ensuring pretrial success while reducing or minimizing
the need for detention.

Jurisdictions considering electronic monitoring must
also tailor such programs to ensure that they comply
with constitutional requirements. For example, several
federal courts have ruled that it is unconstitutional
to impose electronic monitoring as a mandatory
condition for certain categories of offenses. Because
electronic monitoring constitutes a significant
deprivation of liberty,136 these courts have found
that imposing it categorically – without an inquiry
into whether it serves legitimate pretrial needs in
particular cases – may violate the Constitution.137
And in light of the growing understanding that GPS
empowers the government to invade constitutionallyprotected privacy in unique ways,138 courts may

Electronic monitoring should only be used as an
alternative to incarceration, not as a way to monitor
low or medium-risk defendants whose detention would
clearly not be justified.130 Electronic monitoring is not

Moving away from a money bail system that
penalizes the poor is a good thing, but policymakers
and reformers should be wary of a new hazard
that may emerge: “offender-funded” supervision.
For example, in all states except Hawaii and the
District of Columbia, defendants are charged a
fee for electronic monitoring.147 Defendants may
also be charged a monthly fee for pretrial services
supervision, drug or alcohol testing, or participation
in counseling or anger management classes.148 In
some cases, a defendant who is ordered released
with conditions like electronic monitoring may be
forced to wait in jail until he can pay a fee to setup

the GPS monitoring, or may be sent back to jail if he
cannot continue paying fees.
These onerous conditions of release may create
harms that mirror the injustices associated with
money bail. Jurisdictions should avoid charging fees
for pretrial supervision. Any jurisdiction that charges
fees pretrial should ensure that defendants receive
an ability-to-pay hearing and provide judges the
option of fee waiver. If fees are imposed on pretrial
defendants, it is critical that defendants not be
detained because of their inability to pay such fees.



increasingly subject electronic monitoring of pretrial
defendants to probing Fourth Amendment scrutiny.
Ultimately, the invasiveness of electronic monitoring
will almost always be less severe than detention, so
these constitutional considerations should not lead
jurisdictions to conclude that electronic monitoring is
unavailable as an alternative to incarceration. But as
a general matter, these constitutional considerations
counsel in favor of procedures that require courts to
engage in individualized decision-making to determine
whether electronic monitoring will significantly advance
the purposes of pretrial supervision in light of the
circumstances of particular defendants.

further in the criminal justice system. Mandatory drug
testing also raises well-established Fourth Amendment
considerations,145 and for court-ordered drug testing
to survive Fourth Amendment scrutiny a jurisdiction
utilizing drug testing on pretrial defendants will need to
ensure that it has adequate empirical evidence justifying
the use of drug testing to further the legitimate aims
of pretrial supervision.146 Because defendants seem to
fail to abide by drug testing conditions regardless of
the sanctions imposed, programs that use drug testing
and impose sanctions for noncompliance are setting
defendants up to fail.

Electronic monitoring can also be expensive for
defendants, many of whom are required to pay fees in
order to be subject to electronic monitoring.139 One recent
news report documented the experience of a man in
Richland County, South Carolina who was charged with
driving without a license and required to pay a $179.50
setup fee and $300 a month fee to be on electronic
monitoring as a condition of his release – if he stopped
making payments, he would be detained prior to his
trial.140 The unnecessary use of expensive electronic
monitoring could potentially replicate the same economic
injustices that exist in a money bail system. For that
reason, jurisdictions should eliminate or minimize fees
imposed on pretrial defendants, and any fees imposed
should be conditioned on a judicial finding that the
defendant has a reasonable ability to pay such fees.

“Risk assessment” is a broad term that encompasses
a range of procedures for predicting criminal justice
outcomes, and risk assessment tools are used widely
beyond the pretrial context. In the pretrial context, risk
assessment instruments are typically used to gauge the
risk of failing to appear for court proceedings or being
arrested while awaiting trial. Pretrial decision-making
is always, at bottom, a process of risk assessment.
Whether applying categorical criteria, exercising
unfettered judicial discretion, or implementing chargebased schedules, pretrial decisions represent a forwardlooking appraisal of what interventions (if any) are
needed to prevent a defendant from failing to appear
or committing a serious crime while his case is pending.
When reformers or scholars refer to “risk assessment
tools” or “risk assessment instruments,” they generally
refer to a formalized system for incorporating those
kinds of forward-looking assessments into the pretrial
decision-making process.

4. Drug Testing
Drug testing is a widely used condition of release that is
counterproductive in the pretrial supervision context.141
Drug testing has increased considerably as a condition of
release since its inception in the 1980s, despite the fact
that no empirical studies have found solid evidence that
it is effective at reducing pretrial failure. The number
of pretrial services agencies offering drug testing as a
pretrial release condition has grown from 75 percent in
2001 to 90 percent in 2009.142 Yet the studies examining
the effectiveness of drug testing have all found that
drug testing fails to improve pretrial outcomes.143 Drug
testing is simply ineffective in reducing pretrial failure,
even when the court subjects defendants to increasingly
severe sanctions for noncompliance.144 Moreover, a
program that adopts drug testing as a condition of
pretrial release may not only be less effective at reducing
pretrial failure rates but could entrench a defendant even


Broadly, pretrial risk assessment tools will fall into two
categories: clinical tools, which rely on specialists within
the court system (typically pretrial services workers)
to exercise judgment, and actuarial risk assessment
instruments, which generate risk scores based on
statistical analysis. The discussion in this primer focuses
on actuarial tools, often referred to as Actuarial Pretrial
Risk Assessment Instruments (APRAIs).149
Building an APRAI requires not only the expertise of
statisticians, but also access to and maintenance of a
high-quality pretrial database. An APRAI assesses the risk
that a defendant presents on the basis of “risk factors”
incorporated into a statistical formula that uses existing


to accurately predict the outcomes it purports to track.151
After an APRAI is in use, ongoing validation of the tool
is required to ensure its continued efficacy, particularly
in light of changes to a jurisdiction’s population or other
conditions.152 This validation process consists of applying
an instrument to an existing dataset and comparing
risk scores to results.153 Validation studies may include
not only the examination of actual re-arrest or failure
to appear rates, but also racial disparities or other
unwarranted disparities that cannot be justified by risk
differences.154 This validation process may be costly and
complicated. Indeed, once an APRAI is implemented
within a jurisdiction, it becomes increasingly difficult to
validate the accuracy of its results because there may no
longer be a comparison group available. For example, if
a tool designates certain offenders as “high risk,” and
almost all of those “high risk” defendants are detained,
it becomes impossible to test whether individuals who
receive that designation actually have high rates of
pretrial failure.

data to estimate future outcomes.150 Some factors may
reflect information that is immediately available from
mining a defendant’s criminal history and current charge.
Other factors, like employment, history of substance
abuse, and residency status, will require interviewing
the defendant. The complexity of the risk-factor scheme
presents a set of trade-offs: more factors may allow an
instrument to achieve greater accuracy, but collecting
more extensive information may add administrative
costs to or slow-down the application of the instrument,
which may result in some defendants remaining in jail
during that information-gathering process. Once the risk
factors are entered into an APRAI’s statistical algorithm,
the judge considers the resulting “risk score” in setting
conditions of release.
It is not enough for a jurisdiction to proclaim that it will
use a quantitative risk assessment tool – jurisdictions
must ensure the tool’s validity. A valid tool is one that
has been shown (and can be shown on an ongoing basis)

defendant’s existing case and prior criminal history.165
From this dataset they constructed the Public Safety
Assessment-Court (PSA-Court), which produces three
separate risks scores for each defendant, on a scale of
one to six.166 The three axes on which defendants are
scored are risk of “failure to appear,” “new crime”,
and ”new violence.”167

Increasingly, individual jurisdictions or entire states
may consider deploying nationally applicable risk
assessment instruments.159 Much of that change
is being driven by a national APRAI developed by
the Laura and John Arnold Foundation (LJAF).160 It
has developed an APRAI it describes as an “entirely
objective risk assessment score” based solely on
factors related to criminal history, current charge,
and age.161 The tool was piloted in Kentucky, and
one Arnold Foundation-funded study found that the
predictive power of the APRAI was not diminished by
the elimination of the interview-dependent factors,
which had previously made the assessment difficult
to administer.162 After deploying the tool, Kentucky
was able to reduce re-arrests among defendants on
pretrial release while increasing the percentage of
defendants who are released before trial.163 These
findings led the foundation to the second phase
of its project, in which the researchers amassed a
database comprised of over 1.5 million cases drawn
from over 300 jurisdictions.164 Researchers analyzed
the predictive power of and relationship between
hundreds of factors, both interview and noninterview dependent. They identified the nine most
predictive factors, all of which were drawn from a

A report published in June of 2014 summarizing the
results of the first six months of Kentucky’s use of
the PSA-Court revealed that 70% of defendants were
released, which represented only a slight increase
in the rate of release, which had averaged 68% in
the four years prior.168 The rate of pretrial arrest was
reduced by close to 15%.169 Using a control group
to test the usefulness of the third category of risk
(new violent crime), the summary reports that the
PSA-Court predicted this risk with a “high degree of
accuracy.”170 Specifically, those flagged as posing an
increased risk of violent crime were arrested for a
violent offense at a rate 17 times that of defendants
who were not flagged (8.6% versus 0.5%).171 The
PSA-Court has been adopted in jurisdictions around
the country, including across Arizona, New Jersey
and in several major cities.172



Actuarial pretrial risk assessment tools are in use around
the country. They are currently employed statewide in
Virginia, Kentucky, and Ohio, in at least one county of
several states (Arizona, Illinois, Minnesota, New York,
Pennsylvania, and Texas), in Washington, D.C., and for
certain defendants in the federal system.155 Although
risk assessment may be used in a cash-based bail system,
states aiming to reduce their reliance on money bail,
including New Mexico and New Jersey, have relied on
risk assessment as a key feature of reform.156 APRAIs may
be developed on behalf or specific state agencies, by
non-profit groups, or by for-profit corporations.157

1.	 Policy Considerations
a.	 Potential Benefits of Risk Assessment
Several policy considerations may counsel in favor of
using actuarial risk assessment as one factor during bail
determinations. Risk assessment tools may transform
some of the worst pathologies of the pretrial process
by replacing arbitrary or discriminatory decision-making
with a more systematic method grounded in evidence.
As noted earlier, there are only two legitimate bases
for restricting a pretrial defendant’s liberty: preventing
failure to appear at trial and protecting the community
from serious crime. Both of those justifications are, at
bottom, inescapably about assessing risk. The promise of
risk assessment tools is that they allow judges to consider
risk based on sophisticated analysis of data, as opposed
to a more intuitive or amorphous kind of risk prediction
grounded in an individual decision-maker’s experiences
or analysis.173 While no quantitative instrument can
perfectly predict the outcome in a particular defendant’s
case, proponents of risk assessment argue that it is far
superior to a judge’s unguided discretion, which may
reflect stereotypes and other biases or otherwise fail to
engage in individualized consideration of a defendant.174
Indeed, researchers have found that actuarial predictions
are in many contexts more predictive than clinical
assessments of dangerousness and risk of re-offense.175

Actuarial risk assessment tools have been embraced by
many reformers seeking to ensure greater fairness and
efficacy in pretrial justice. Instead of setting bail using
offense-based bail schedules or a judge’s hunch, these
tools give judges an evidence-based framework to set
appropriate conditions of release, reducing the risk that
a defendant will fail to appear in court or be a danger
to the public in the pretrial period. When used properly,
risk assessment tools may offer great promise as a way
to replace money bail with an alternative grounded in
statistical assessments of pretrial outcomes.
At the same time, the use of risk assessment tools in
the pretrial context raises very serious concerns and
has attracted considerable criticism. Even the strongest
arguments in favor of risk assessment recognize
that a tool must be properly calibrated to reflect a
jurisdiction’s specific population, which means that the
potential benefits turn on complicated and potentially
costly determinations about which instrument to
use.158 Moreover, even the best risk assessment tools
may generate serious disparities along racial or other
demographic lines. Without being considered in a
broader context, quantitative risk assessment scores may
also displace other potentially relevant considerations,
resulting in mechanical application of pretrial outcomes
that may be poorly suited to the circumstances and
needs of individual defendants.

In addition to improving individual outcomes, risk
assessment tools may decrease the overall rates of
pretrial detention. A 2012 study, which looked at a
dataset of 116,000 defendants from 1990 to 2006, found
that if judges chose to release all defendants with less
than a 30 percent chance of being rearrested for any
crime during the pretrial period, 85 percent of pretrial
defendants would have been released, significantly
more than the number of defendants who were actually
released during that period.176 Risk assessment tools
may supply courts with an objective basis to release
low-risk defendants on their own recognizance or with
limited pretrial conditions. Reducing the jail population
serves many important interests: it spares individuals
from the serious infringement on liberty and collateral
consequences (such as exposure to violence or job
loss) that can follow even a short period of pretrial
detention,177 and it spares defendants’ families the
destabilizing effects that may follow from loss of income,
housing, or child custody. Reduction of detention at a
sufficiently significant scale also lowers the economic
costs associated with administering jails.

Risk assessment tools, in other words, present complex
considerations. This primer does not attempt to provide
a standard prescription for every jurisdiction. Instead, the
following discussion outlines some of the policy and legal
considerations that should guide the decision-making
about whether to utilize quantitative risk assessment
tools in any particular jurisdiction.



Risk assessment tools may also counteract unfair
disparities in current bail practices, particularly along
racial and socioeconomic lines. Actuarial predictions may
help ameliorate these disparities in several ways. First,
simply by helping to displace money bail, risk assessment
tools may substantially cure racial or other unwarranted
disparities. As noted earlier, entrenched linkages
between race and wealth will result in patterns of racial
inequality when a policy has the effect of discriminating
against the poor.178 Risk assessment may also diminish
racial or socioeconomic disparities by counterbalancing
implicit or explicit biases of judges.179 To the extent that
evidence-based methods run counter to those biases, a
jurisdiction may achieve significantly fairer outcomes.

must carefully consider how to characterize different
risk levels. Risk assessment tools typically define certain
risk levels as “high,” “moderate,” or “low,” but that
characterization is a policy judgment, not a statistical
one. Calling a risk score “high” is likely to significantly
impact how judges, and the public, view particular
outcomes. An initial decision over where to set that
threshold – is a “high risk” defendant one with a 30%
risk of failure, or should that label be reserved for 50% or
75% risk? – should take place transparently and with the
involvement of all criminal justice system stakeholders.
Second, judges and other system actors must undergo
training that allows them to understand precisely what
it is that a risk score conveys: a statistical estimate of a
particular outcome based on the observed outcomes
among a population of individuals who share certain
characteristics. In many instances, an actuarial tool
may be very predictive for the group on average but
not accurate for any given member of that group.184
If a judge relies on a risk score without considering
other factors that may be relevant in making a bail
determination, the risk score could carry undue weight.

Initial experiences in some jurisdictions suggest that risk
assessment tools may improve pretrial outcomes on many
dimensions. After Kentucky began to use a risk assessment
tool, the state was able to increase the percentage of
defendants who were released before trial while reducing
re-arrests among defendants on pretrial release.180 Virginia
has also kept pretrial failure rates low by using a risk
assessment tool. In fiscal year 2012, Virginia defendants
who were released pretrial had a 96.3% appearance rate
in court and less than 4% of released defendants were
re-arrested.181 Mecklenburg County, North Carolina has
been able to reduce the number of people held in jail
pretrial since using a risk assessment tool.182 Just a month
after Allegheny County, Pennsylvania instituted changes
to its pretrial services program, including the use of risk
assessment tools to inform bail determinations, the number
of defendants held in jail after their first appearance was
reduced by 30 percent.183

It is also important to ensure a consistent structure
for balancing a risk assessment score with other
considerations. If the point of risk assessment is to
displace arbitrary or biased decision-making, that
purpose would be defeated if the ultimate pretrial
decisions are not structured to ensure consistent riskbased decision-making. Jurisdictions should issue
guidance for judges to structure the relationship
between a defendant’s risk score and other
considerations. This might include a list of factors
that can justify departing from what the instrument
indicates. Such criteria should embody the principle that
a pretrial decision should impose the least restrictive
conditions necessary.185 It could do this, for example,
by requiring that any outcome more restrictive than a
risk score indicates must be justified in writing based

b.	 Implementation Considerations
Capturing the potential benefits of risk assessment
requires close attention to several important
implementation considerations. First, policymakers

Risk assessment tools may transform some of the worst
pathologies of the pretrial process by replacing arbitrary or
discriminatory decision making with a more systematic method
grounded in evidence.



on certain enumerated criteria. Judges, prosecutors,
defense attorneys, and other practitioners will need
to be trained in how to interpret and utilize risk
assessment scores before a jurisdiction implements an
actuarial risk instrument in the pretrial setting. Defense
counsel should also have a role in the application of
a risk assessment instrument – this may include being
present with a defendant during an initial interview and
promptly receiving a copy of the data inputted into an
APRAI and the ultimate report. Finally, implementation
of any APRAI should be accompanied by a robust datacollection requirement that allows a jurisdiction and
outside observers to measure the instrument’s effects
in terms of overall detention rates, pretrial failure rates,
and racial disparities.

reflects risk of arrest for a serious violent crime, whether
the arrest will be occurring during the pendency of the
defendant’s case, or which interventions are likely to be
effective in mitigating that risk.188
The potentially negative effects of risk assessment,
moreover, may disproportionately impact Black and
Latino defendants or other minority groups. In particular,
many critics argue that by relying on underlying factors
that are shaped by race discrimination, statistical tools
may reinforce and deepen inequalities in the criminal
justice system.189 To the extent that risk scores reflect
prior interaction with the criminal justice systems, the
disproportionate exposure of African Americans and
Latinos to law enforcement will skew those assessments
– even where those underlying disparities reflect
discrimination or other unjust patterns.190 Similarly,
risk assessment scores that incorporate educational
history, housing instability, or other socioeconomic
factors that correlate with race may also import serious
racial disparities.191

Potential Harms of Risk Assessment
Despite the potentially promising aspects of risk
assessment, policymakers should also consider the
very serious possible drawbacks. For one thing, all
of the potential benefits of risk assessment hinge on
generating consistently accurate predictions. That
requires a reliable method of gathering data for the
underlying algorithm and properly inputting information
about individuals who the risk assessment instrument
evaluates. But “criminal justice data is notoriously
poor,”186 and reliance on an ostensibly scientific process
fueled by faulty data may skew outcomes.187 Before
utilizing risk assessment, many jurisdictions will need
to improve the collection of criminal justice data that
they will rely on. This is an ongoing process. It means
having sufficiently reliable means for collecting data
relevant to individual defendants to input into their
risk calculation; depending on the instrument, it may
also mean continually collecting reliable information
about the overall population of pretrial defendants and
other related aggregate-level data to ensure that the
instrument reflects current populations and pretrial
outcomes. In many jurisdictions, the costs related to
data collection and maintenance may significantly strain
limited budgets.

Former Attorney General Eric Holder has expressed
the concern that, in the sentencing context, actuarial
risk assessment “may inadvertently undermine our
efforts to ensure individualized and equal justice.”192
In Holder’s view, “[b]y basing sentencing decisions on
static factors and immutable characteristics – like the
defendant’s education level, socioeconomic background,
or neighborhood – [risk assessment instruments] may
exacerbate unwarranted and unjust disparities that are
already far too common in our criminal justice system
and in our society.”193 This can lead to a vicious cycle:
because pretrial detention has been shown to lead to
worse criminal justice outcomes, the characteristics of
the individuals detained pursuant to risk assessment will
gain an even stronger association with pretrial failure
over time, thus strengthening the seeming predictive
power of those features.194 Indeed, because APRAIs are
based on empirically-derived factors, it is possible that risk
assessment tools will not only entrench but exacerbate
existing racial and socioeconomic disparities by appearing
to give a scientific imprimatur to unequal outcomes.

In addition to the possibility of inaccuracies flowing from
erroneous inputs, risk assessment tools may distort pretrial
outcomes to the extent that the “risk” they forecast is
ambiguous or otherwise subject to broad interpretation.
In many instances, prediction tools do not distinguish
between risk of pretrial flight and risk of arrest. Even when
tools make that basic distinction, a simple designation of
“high risk” may not tell a decision-maker whether that


Some critics of risk assessment have also argued that
the very premise of an actuarial model – drawing on
aggregate data to make decisions about individuals
– violates fundamental norms of fairness. While an
individual’s conduct is within his control, that individual
cannot control the aggregate conduct of others who
share some characteristic deemed relevant for the risk


assessment instrument.195 Therefore, because actuarial
models derive outcomes from aggregated data, the
individual’s treatment is based, at least partially, on the
independent decisions of others.196

answers to jurisdictions navigating that constitutional
terrain. Rather, it outlines the principal constitutional
considerations likely to be relevant to any jurisdiction
considering the use of quantitative risk assessment tools
as part of their pretrial system.
Any risk assessment tool that determines or influences
pretrial outcomes must conform to the Equal Protection
Clause of the Fourteenth Amendment. Equal protection
principles generally prohibit the government from taking
adverse action against a person on the basis of certain
protected characteristics, particularly race, national
origin, and sex.198 Typically, this prevents the government
from acting on the basis of racial classification except
in exceedingly narrow circumstances; classifications
based on sex will receive slightly more deference but
must also satisfy exacting judicial scrutiny.199 In the risk
assessment context, those “classifications” will consist
of the inputs that drive an assessment tool’s statistical
analysis. As a starting point, then, equal protection
considerations counsel strongly against using a system in
which race or sex are incorporated into risk scores.200 It
is important to note that equal protection principles will
generally prohibit express classification based on race
or sex or intentional discrimination on those bases, but
the Constitution does not proscribe policies that have
an unintentional disparate impact on particular groups,
even if those disparities are foreseeable.201 While such
disparities will not violate constitutional guarantees,
they may violate core policy imperatives to avoid racially
unjust outcomes. Jurisdictions should carefully consider
these policy issues before implementing a risk assessment
tool. Those considerations are discussed further in
Section III.B.1.

Jurisdictions considering the use of actuarial risk
assessment tools should consider the policy considerations
outlined above in deciding what framework will deliver
a fair and effective pretrial system. Significantly, the
determination of whether to use actuarial risk assessment
is inherently a relative decision. In other words, the
potential costs and benefits – including the effects on
detention rates, efficacy in improving pretrial outcomes,
fairness to individual defendants, and racial disparities –
must be considered relative to a preexisting status quo or
a likely alternative pretrial framework. In making those
judgments, the details will matter. The potential benefits
and drawbacks of risk assessment will vary depending
on how an instrument is implemented – whether it is
accompanied by a reliable system for collecting and
maintaining data, whether judges and other system actors
receive proper training, whether appropriate procedural
safeguards are in place, and how risk assessment is
integrated into an overall decision-making framework.
Policymakers considering a risk-based system should
begin from the premise that the efficacy and fairness
of risk assessment instruments are not matters that can
be resolved in a vacuum. Rather, the policy value of risk
assessment should be measured against the kinds of
potential advantages and hazards outlined in this section.
2.	 Constitutional Considerations
There is very little judicial guidance on the constitutional
implications of risk assessment tools, and the cases that
have examined issues related to risk assessment have not
arisen in the pretrial context.197 Depending on how such
tools are used, substantial constitutional considerations
may come into play. Much will depend on the specific
context in which risk assessment tools are used. For
example, one set of constitutional implications may
attach to risk assessment instruments used to determine
what conditions are necessary to ensure appearance at
trial; different constitutional considerations may apply
where risk scores are incorporated into a decision of
whether to preventively detain an individual deemed
dangerous to the community. The discussion here
does not attempt to provide definitive or exhaustive

Incorporating risk assessment tools into pretrial
decision-making may also implicate constitutional due
process guarantees. Again, the dimensions of any due
process analysis will depend on what purpose the risk
assessment instrument serves. Decisions about whether
or not to detain someone pretrial will demand more
stringent due process protections than decisions about
what array of non-detention conditions – such as
check-in requirements or electronic monitoring – may
be necessary to ensure appearance at trial. But all of
these decisions involve potential infringements on a
defendant’s pretrial liberty, which means that any risk
assessment tool must be consistent with a defendant’s
due process rights.



The Constitution’s due process protections require
that, before a person is deprived of liberty by the
government, she must enjoy sufficient procedural
safeguards to “minimize substantively unfair or mistaken
outcomes.”202 The hallmarks of such procedures are
reasonable notice and an opportunity to be heard.203 In
the pretrial context, the Supreme Court has emphasized
that, at least for a preventive detention decision, the
procedural due process inquiry turns on whether a
defendant enjoys “procedures by which a judicial officer
evaluates the likelihood of future dangerousness [that]
are specifically designed to further the accuracy of that
determination.”204 These principles should be reflected
in any procedures that rely on actuarial risk assessment.
Generally, that means that a defendant must have
some opportunity to contest potentially inaccurate or
substantively unfair risk assessment procedures.

are able to identify groups of high-risk offenders, not a
particular high risk offender; that some studies of the tool
being used have “raised questions about whether they
disproportionately classify minority offenders as higher
risk of recidivism”; and that the tool is based on a national
sample that has not been validated for Wisconsin and that
risk assessment tools must be constantly monitored and
re-calibrated for accuracy as the population changes.206
In light of these due process principles, numerous
safeguards should be in place when risk assessment
instruments are used in the pretrial context. Those
safeguards should reflect the weighty liberty interests
involved in the pretrial setting, where presumptively
innocent defendants face a deprivation of liberty.207
While the specific framework will depend on the
instrument being used and its role in pretrial decisionmaking, a defendant should be provided with a
substantive understanding of how the instrument works
and a meaningful opportunity to contest its application
in his case. This means disclosing the defendant’s risk
assessment score, the factors considered in determining
the score, the relative weights given to different factors,
and information about when and how the instrument
was validated and re-normed, including information
about the population samples used in validating it.208 A
procedural framework should also ensure disclosure of
relevant information about the instrument’s accuracy
– including studies demonstrating unwarranted race
disparities or other inaccuracies – and set out clear
parameters about precisely what role the instrument
may play in shaping pretrial decisions.

There is no case law at this point elaborating what that
should mean in the pretrial context, but case law in other
areas suggests some ways jurisdictions might ensure
adequate procedures. In one recent case, the Wisconsin
Supreme Court upheld the use of a risk assessment
instrument in the sentencing context, but outlined
several requirements for applying it consistently with
due process. The court held that the instrument could
be used to determine whether portions of a sentence
could be served in the community instead of prison. But
the court went on to hold that the instrument could
not be used “to determine the severity of the sentence
or whether an offender is incarcerated” and the court
imposed “the corollary limitation that risk scores may
not be considered the determinative factor in deciding
whether the offender can be supervised safely and
effectively in the community.”205 The court further held
that sentencing judges considering risk reports must
receive an accompanying advisory alerting them to
four points: that the company that created the tool has
invoked its proprietary interest to prevent disclosure of
how factors are weighted or risk scores are determined;
that risk assessment scores are based on group data and

One of the most significant pathologies of money bail is
its use as a subterranean mode of preventive detention;
judges address perceived risk to the community by
setting bond at a level that will be presumptively out
of reach to a defendant.209 Using cash as a proxy to
preventively detain defendants viewed as dangerous

“In our society liberty is the norm, and detention prior to trial or
without trial is the carefully limited exception.”



is indefensible as a matter of principle, but it reflects a
real concern by many judges about the risk that certain
people will commit serious crimes while on pretrial
release. For this reason, the discussion of moving to a
risk-based bail system is often accompanied by a call for
risk-based preventive detention.

use of preventive detention in the pretrial context.
In upholding the constitutionality of the federal Bail
Reform Act, the Salerno Court first emphasized the
importance of the statutory purpose of preventive
detention: detention that is “regulatory, not penal”
does not constitute “impermissible punishment before
trial.”212 The test for determining whether a preventive
detention policy is regulatory or punitive depends,
first, on whether there was an express legislative intent
to punish; if not, the inquiry turns to whether there is
a rational connection between the policy and a nonpunitive justification and, finally, whether the policy is
proportional to that justification.213 In Salerno, the Court
found that the federal bail statute fell on the regulatory
side of this distinction. Significantly, in examining
whether the preventive detention scheme embedded in
the Bail Reform Act was proportionate to the regulatory
interest in preventing danger to the community, the
Salerno Court emphasized the statute’s limited reach and
detailed safeguards:

At least twenty-two states, the District of Columbia,
and the federal system already authorize the use of
pretrial preventive detention in some circumstances.210
Many more states are likely to consider using or
expanding preventive detention in conjunction with a
risk assessment system. But a jurisdiction that chooses
this path should do so with extreme care. Insofar as
states choose to utilize preventive detention as an
aspect of pretrial reform, this section outlines the
baseline legal and policy considerations that should
guide policymaking.
This primer does not take a position on whether, as
a policy matter, preventive detention is appropriate.
Indeed, many observers raise grave concerns about
the use of preventive detention. Among other things,
critics point out that there is no guarantee that
authorizing judges to use preventive detention will
reduce the number of individuals detained pretrial – if
the standards are open-ended enough, or define pretrial
risk broadly enough, a tool intended to reform excessive
jail populations could have the opposite effect. More
fundamentally, some question whether preventive
detention is legitimate as a matter of principle.211 Pretrial
defendants are presumed innocent and using a mere
arrest as a trigger for depriving a person of his liberty
strikes some as contrary to the basic underpinnings
of a free society. On the other hand, many reformers
have championed risk-based pretrial detention as a
way to cure the arbitrary and discriminatory practices
inherent in money bail while providing judges with a
more transparent and rational tool for addressing serious
risk to the community. Proponents of limited authority
for pretrial detention note that the Supreme Court
has ruled that such mechanisms can be consistent with
constitutional guarantees, and they maintain that the
Court’s rulings will ensure robust procedural safeguards
as a prerequisite to any pretrial detention authority.

The Bail Reform Act carefully limits the
circumstances under which detention may
be sought to the most serious of crimes. The
arrestee is entitled to a prompt detention
hearing, and the maximum length of pretrial
detention is limited by the Speedy Trial Act.
Moreover...the conditions of confinement
envisioned by the Act appear to reflect the
regulatory purposes relied upon by the
Government...[T]he statute at issue here
requires that detainees be housed in a facility
separate, to the extent practicable, from
persons awaiting or serving sentences or being
held in custody pending appeal.214
Having determined that the statutory authority to detain
pretrial defendants was regulatory rather than punitive,
the Court went on to decide that the restrictions
the statute imposed on pretrial liberty could be
adequately justified by the compelling governmental
interest. In doing so, the Court emphasized the
“narrow circumstances” in which preventive detention
was authorized.215 Once again, the Court’s detailed
description of the Bail Reform Act’s procedural
framework reveals the considerations it deemed vital to
the constitutional analysis:

1.	 Constitutional Requirements

The Bail Reform Act...narrowly focuses on
a particularly acute problem in which the
Government interests are overwhelming. The

The Supreme Court’s decision in United States v. Salerno
articulates the constitutional principles governing the



“testify in their own behalf, present information by
proffer or otherwise, and cross-examine witnesses,”219
which necessarily must be part of an adversarial hearing.
Similarly, the Court emphasized that in detention
hearings under the Bail Reform Act the government bears
the burden of demonstrating by clear and convincing
evidence that no less restrictive conditions suffice; this
kind of stringent burden of proof implies the use of an
adversarial hearing to test the government’s showing.
While the exact protections within a hearing may vary,
the Court’s reasoning assumes an adversarial hearing to
be an essential component of a constitutional preventive
detention framework.

Act operates only on individuals who have been
arrested for a specific category of extremely
serious offenses. Congress specifically found
that these individuals are far more likely to
be responsible for dangerous acts in the
community after arrest. Nor is the Act by any
means a scattershot attempt to incapacitate
those who are merely suspected of these
serious crimes. The Government must first
of all demonstrate probable cause to believe
that the charged crime has been committed
by the arrestee, but that is not enough. In a
full-blown adversary hearing, the Government
must convince a neutral decision maker by clear
and convincing evidence that no conditions of
release can reasonably assure the safety of the
community or any person.216

b.	 Right to Counsel
Just as an adversarial hearing provides the structure in
which the procedural protections outlined in Salerno
can operate, the right to counsel ensures that a pretrial
defendant can enjoy those protections in a meaningful
way. Like the right to an adversarial hearing, the right to
counsel is an indispensable safeguard. Indeed, Salerno
stressed the importance of a combination of procedures
and rights “specifically designed to further the accuracy”
of a determination of dangerousness.220 Many of the
same safeguards that imply the structure of an adversarial
hearing – the ability to testify, present evidence, and
cross-examine adverse witnesses – will typically require
the presence of counsel to ensure they are meaningful.
The Court also noted that the ultimate detention or
release decision must be rooted in statutorily enumerated
factors.221 Those protections will lack any functional
significance unless defendants have competent lawyers
to take advantage of such procedural opportunities.222
Because failing to provide a right to counsel would, in
practical terms, vitiate the other procedural safeguards
emphasized in upholding the Bail Reform Act, it should
be regarded as a bedrock requirement in any system
allowing preventive detention.

Given this detailed and robust procedural framework,
the Court ruled that, “[w]hen the Government proves by
clear and convincing evidence that an arrestee presents
an identified and articulable threat to an individual or
the community...a court may disable the arrestee from
executing that threat.”217
Jurisdictions contemplating the use of preventive
detention should adopt the safeguards emphasized
by the Salerno Court to the greatest degree possible.
While the Salerno Court never stated explicitly which
individual safeguards may be constitutionally mandatory,
two appear to be particularly important components of
ensuring the constitutionality of preventive detention
schemes: an adversarial hearing and the right to the
presence of counsel at bail hearings. As described in
more detail below, those two features are elemental to
the broader array of procedural protections at the heart
of the court’s analysis. Beyond these two overarching,
structural protections, the Court’s analysis gives useful
guidance for states seeking ways to structure preventive
detention authority. As a matter of law and policy, such
systems should treat as a first principle one of the Court’s
concluding remarks: “In our society liberty is the norm,
and detention prior to trial or without trial is the carefully
limited exception.”218

2.	 Vital Procedural Protections
Salerno did not dictate a universal statutory
architecture for preventive detention. While the rights
to an adversarial hearing and an attorney emerge as
indispensable elements, the Court’s analysis suggests
that standing alone, those safeguards would be
insufficient. The following procedural protections would
fortify a preventive detention framework’s compliance
with due process.

a.	 Adversarial Hearings
All preventive detention frameworks should provide
defendants with an adversarial hearing. The statutory
provisions identified by the Salerno Court as sufficient
to satisfy due process included defendants’ ability to



Court placed significant weight on the limited entry
points to the scheme of preventive detention – the
carefully circumscribed threshold circumstances under
which any defendant might face a preventive detention
determination. Policymakers and advocates seeking
to implement preventive detention schemes should
carefully limit the entry points to preventive detention
hearings. There are three different types of entry points
that may be utilized to preventively detain a defendant –
risk assessment score, offense charged, and motion by a
prosecutor – each of which is discussed in turn.

a.	 Speedy Trial
Where a defendant’s liberty is substantially impaired prior
to trial, the pretrial period should be limited. The specific
language used to guarantee a speedy trial for pretrial
detainees may vary from state to state, but it should be
defined for preventively detained individuals in particular.
Some jurisdictions have implemented statutory language
designed to give effect to this principle:
b.	D.C.
The case of the person [preventively] detained
pursuant to ... this section shall be placed on
an expedited calendar and, consistent with
the sound administration of justice, the person
shall be indicted before the expiration of 90
days, and shall have trial of the case commence
before the expiration of 100 days.223

One way to limit the entry points to preventive detention
determinations is to use actuarial risk assessment scores
as a necessary, but not sufficient, basis to trigger a
hearing. Significantly, this will require that states rely on
risk assessment instruments geared specifically to the
risk of re-arrest for violent or serious crime, as opposed
to instruments that lump together re-arrest for serious
and non-serious crime or do not distinguish between
re-arrest and non-appearance. Kentucky’s pilot program
is one example. That system allows the state to conduct
initial assessments that channel individuals with high
risk assessment scores into hearings that afford greater
rights and safeguards in order to make more accurate
individualized determinations.229 A jurisdiction might
further assure limited entry points by only utilizing risk
assessment tools for individuals charged with particular
offenses, as is the case in New Jersey.230

c.	 Vermont
(a) Except in the case of an offense punishable
by death or life imprisonment, if a person is held
without bail prior to trial, the trial of the person
shall be commenced not more than 60 days
after bail is denied.
( b) If the trial is not commenced within 60
days and the delay is not attributable to the
defense, the court shall immediately schedule a
bail hearing and shall set bail for the person.224

Some jurisdictions have automatically triggered preventive
detention hearings based on the offense charged, even
though the offense charged may not correspond to risk
of reoffending. For example, under both the D.C. and
federal system, particular types of offenses create a
rebuttable presumption that no condition or combination
of conditions will reasonably assure appearance of public
safety.231 These rebuttable presumptions trigger detention
hearings and lead to the detention of many charged
individuals. Offense-based triggers are problematic
because they are not tied to individual circumstances of
a defendant and reflect the relatively low threshold for
issuing a charge. If used, it is crucial that such enumerated
offenses remain narrow and that, even when they trigger
hearings, they do not dictate outcomes or prevent an
individualized determination based on the defendant’s
circumstances. This is especially important because
prosecutors exercise wide discretion in making charging
decisions, and inappropriate charging decisions could lead
to unnecessary preventive detention.232

Additionally, states should examine their speedy trial
statutes to ensure that carve-outs do not render the law
ineffective. For example, under New York’s “ready rule,”
as long as the prosecutor has declared that he or he is
ready for trial, the delays from “court congestion” or
even an adjournment because the prosecutor failed to
turn over evidence are not counted as part of the
trial clock.2255
d.	Limited Entry Points
In Salerno, the Court repeatedly emphasized the narrow
scope of the preventive detention authority in the
federal pretrial system – it noted that the challenged
statute “carefully limits the circumstances under
which detention may be sought to the most serious
crimes;”226 that the statute “narrowly focuses on a
particularly acute problem in which the Government
interests are overwhelming;”227 and that it applied
only in “narrow circumstances.”228 In other words, the



resources, length of residence in the community,
community ties, past conduct, history relating to
drug or alcohol abuse, criminal history, and record
concerning appearance at court proceedings; and

Another potential pathway to preventive detention
hearings is authorizing prosecutors to move for such
hearings. Both the D.C. and federal systems also allow
the prosecutor to move for pretrial detention based on a
number of grounds.233 This discretion may be appropriate
in some circumstances, but it should be structured so
that prosecutors may only move for pretrial detention
based on clearly defined, limited circumstances. To the
extent a defendant is detained prior to the hearing, the
prosecutor should be required to make a substantial
initial showing to justify that detention.

e.	 Statutorily Enumerated Factors Guiding
Bail Determinations

4.	 The nature and seriousness of the danger to any
person or the community that would be posed by the
person's release.235

The Salerno Court noted that, in the federal scheme,
judicial officers must follow statutory guidelines and
make a finding by clear and convincing evidence that
there is a statutorily permissible reason for detention.234
Imposing clear and stringent standards that must be
satisfied to preventively detain a defendant helps ensure
adherence to constitutional standards.

Many other states mirror D.C.’s statute.236
Given the forward-looking, regulatory nature of
preventive detention, states should not place exclusive
or predominant weight on the nature of the charged
offense, or the weight of the evidence, in prescribing
standards for pretrial release or detention. The
charged offense may deserve some weight in those
determinations, insofar as the most serious charges
carry elevated penalties that may increase a defendant’s
incentive to abscond. But it is important that these
considerations not subsume the individual determination
focused on a defendant’s particular circumstance, nor
should a focus on the charged offenses give rise to a
mini-trial on the defendant’s guilt or innocence. To the
extent that the gravity of the charged offense informs
the pretrial release decision, it should be just one
consideration that may be reinforced or counterbalanced
by other factors. Policymakers should avoid statutory
language that requires or implies that the charged
offense is the sole or predominant consideration.

In addition to imposing a stringent standard, jurisdictions
should supply courts with clear criteria to apply in
weighing a preventive detention decision. The D.C.
statute offers an example of the types of factors that
states should address:
1.	 The nature and circumstances of the offense
charged, including whether the offense is a crime of
violence or dangerous crime...or involves obstruction
of justice...;
2.	 The weight of the evidence against the person;
The history and characteristics of the person,

B.	Whether, at the time of the current offense or
arrest, the person was on probation, on parole,
on supervised release, or on other release
pending trial, sentencing, appeal, or completion
of sentence for an offense under local, state, or
federal law; and

A.	The person's character, physical and mental
condition, family ties, employment, financial

Additionally, while actuarial risk assessment tools may be
utilized in an initial screening, they should not displace

Policymakers and advocates seeking to implement preventive
detention schemes should carefully limit the entry points to
preventive detention hearings.



the other listed factors. Where actuarial risk assessment
tools suggest a high risk of committing some future
crime, a judicial officer should still consider the nature
and seriousness of the danger and allow the defendant
to rebut the risk assessment by providing additional
evidence through an adversarial hearing with the
assistance of counsel.



I V.


The country's approach to the pretrial process is
undergoing intensive reexamination and may be on
the verge of fundamental change. Money bail, nearly
ubiquitous and deeply entrenched for decades, is now
subject to scrutiny and criticism from a broad array of
observers and advocates. Litigation challenging practices
that result in wealth-based detention have gained
traction, creating an opening for remaking pretrial
systems in jurisdictions around the country. An energized
movement for reform has embraced a risk-based model
that a number of jurisdictions have now implemented,
with many others watching closely. These trends are
encouraging and should spur further action. At the same
time, all stakeholders need to ensure that this wave
of reform yields workable new models that solve the
problems plaguing the current system without producing
new forms of injustice. Striking that balance will require
careful attention by all stakeholders to the legal and
policy questions outlined in this primer. With those
considerations in mind, and guided by local needs and
opportunities, advocates and policymakers should forge
a new path for pretrial justice that furthers the highest
ideals of our legal system and ensures fair, consistent,
and efficient administration of justice.





10.	Prison Pol’y Inst., Detaining the Poor 2-3 (2016), available at

See infra, note 36, listing cases.


See, e.g., Nick Pinto, The Bail Trap, N.Y Times Magazine, Aug.
13, 2015; Leon Neyfakh, Is Bail Unconstitutional?, Slate, June 30,
2015, available at
system_keeps_the_poor_in_jail_and_lets.html; Last Week
Tonight with John Oliver, Bail (June 7, 2015), available at https://

11.	 See Richard Williams, Bail or Jail: State Legislatures (May 2012),
available at
12.	Todd Minton & Zhen Zeng, Bureau of Justice Statistics, Jail
Inmates at Midyear 2014 1-3, 8 (2015), available at http://www.

See, e.g., Ram Subramanian et al., Vera Inst. of Justice,
Incarceration’s Front Door: The Misuse of Jails in America,
July 29, 2015, available at
incarcerations-front-door-report_02.pdf; Timothy R. Schnacke,
Nat’l Inst. of Corrs., Money as a Criminal Justice Stakeholder:
The Judge’s Decision to Detain or Release a Person Pretrial (Sept.
2014), available at
Christopher T. Lowenkamp et al., Laura & John Arnold Found.,
The Hidden Cost of Pretrial Detention (Nov. 2013), available at
pdf; Michael R. Jones, Pretrial Justice Inst., Unsecured Bonds: The
Most Effective and Most Efficient Pretrial Release Option (Oct.
2013), available at


See, e.g., Carl Hulse, Unlikely Cause Unites the Left and the Right:
Justice Reform, N.Y. Times, Feb. 18, 2015.


See Amer. Bar Ass’n, ABA Standards for Criminal Justice:
Pretrial Release, Standard 10-1.4(a) (3d ed. 2007) (“additional
conditions should be imposed on release only when the need
is demonstrated by the facts of the individual case reasonably
to ensure appearance at court proceedings, to protect the
community, victims, witnesses or any other person and to
maintain the integrity of the judicial process.”).

13.	David C. May, et al., Going to Jail Sucks (And It Really Doesn’t
Matter Who You Ask), 39 Am. J. Crim. Just. 250, 251 (2014),
available at
14.	Nicholas Freudenberg, Jails, Prisons, and the Health of Urban
Populations: A Review of the Impact of the Correctional System
on Community Health, 78 J. of Urban Health: Bulletin of the New
York Academy of Medicine 214 (June 2001).
15.	Justice Pol’y Inst., Bailing on Baltimore: Voices from the Front
Lines of the Justice System 15 (Sept. 2012), available at http://
16.	 See Norimitsu Onishi, In California, County Jails Face Bigger Load,
N.Y. Times, August 6, 2012, available at http://www.nytimes.
17.	Subramanian, supra note 3, at 12.
18.	Noonan et al., U.S. Dept. of Justice, Bureau of Justice Statistics,
Mortality in Local Jails and State Prisons, 2000- 2013, Statistical
Tables 1 (August 2015), available at
19.	 See Laura Sullivan, Inmates Who Can’t Make Bail Face Stark
Options, Nat’l Public Radio (Jan 22, 2010), available at http:// See
also Barker v. Wingo, 407 U.S. 514, 532-33 (1972) (“The time spent
in jail awaiting trial has a detrimental impact on the individual.
It often means loss of a job; it disrupts family life; and it enforces
idleness. Most jails offer little or no recreational or rehabilitative
programs. The time spent in jail is simply dead time. Moreover,
if a defendant is locked up, he is hindered in his ability to gather
evidence, contact witnesses, or otherwise prepare his defense.
Imposing those consequences on anyone who has not yet been
convicted is serious. It is especially unfortunate to impose them
on those persons who are ultimately found to be innocent.”).

6.	Timothy R. Schnacke, U.S. Dept. of Justice, Nat’l Inst. of Corrs.,
Fundamentals of Bail: A Resource Guide for Pretrial Practitioners
and a Framework for American Pretrial Reform 2 (2014), available
fundamentals_september_8,_2014.pdf (recognizing that
“depending on the source, one will see bail defined variously as
money, as a person, as a particular type of bail bond, and as a
process of release”). When this primer refers generally to “money
bail” or “cash bail,” it is referring to secured money bonds.
7.	Subramanian et al., supra note 3, at 32.

20.	Laura & John Arnold Found., Pretrial Criminal Justice Research
2 (2013), available at



21.	 See id.



22.	Megan Stevenson, Distortion of Justice: How the Inability to



on Bail Reform, KQRE News 13, Feb. 12, 2016, available at http://

Pay Bail Affects Case Outcomes 22 (Working Paper, 2016),
available at
workingpapers/Distortion-of-Justice-April-2016.pdf; Heaton
et al., The Downstream Consequences of Misdemeanor
Pretrial Detention 3 (July 2016), available at

See State v. Brown, 338 P.3d 1276, 1278 (N.M. 2014) (concluding
that “the district court erred by requiring a $250,000 bond when
the evidence demonstrated that less restrictive conditions of
pretrial release would be sufficient”).

23.	Heaton et al., supra note 22, at 3.

See Walker v. City of Calhoun, 2016 WL 361580 (N.D. Ga. 2016)
(court order certifying as a class for declaratory and injunctive
relief “all arrestees unable to pay for their release who are or will
be in the custody of the City of Calhoun as a result of an arrest
involving a misdemeanor, traffic offense, or ordinance violation”);
Jones v. City of Clanton, 2015 WL 5387219, at *4-5 (M.D. Ala.
2015) (granting declaratory judgment); Rodriguez v. Providence
Cmty. Corr., Inc., 155 F. Supp. 3d 758, 767 (M.D. Tenn. 2015)
(enjoining the practice of arresting and detaining probationers for
nonpayment of court costs); Pierce v. City of Velda City, 2015 WL
10013006, at *1 (E.D. Mo. 2015) (issuing declaratory judgment that
“no person may, consistent with the Equal Protection Clause of the
Fourteenth Amendment to the United States Constitution, be held
in custody after an arrest because the person is too poor to post
a monetary bond”); Thompson v. Moss Point, 2015 WL 10322003,
at *1 (S.D. Miss. 2015) (issuing declaratory judgment); Mitchell v.
City of Montgomery, 2014 WL 11099432 (M.D. Ala. 2014) (granting
declaratory and injunctive relief).

24.	 Id. at 3-4, 19, 21.
25.	 See, e.g., Rakesh Kochhar & Richard Fry, Pew Research Ctr.,
Wealth Inequality has Widened Along Racial, Ethnic Lines Since
End of Great Recession, Dec. 12, 2014, available at http://www. (discussing the wealth gap between white and black
households). Studies have also shown that white defendants
receive more favorable bail determinations than similarly situated
African American defendants, either because of racial animus or
implicit bias on the part of decision makers. Cynthia E. Jones, Give
Us Free, 16 Legis. & Pub. Pol’y 919, 943 (2013).
26.	 S
 ee, e.g., Traci Schlesinger, Racial and Ethnic Disparities in Pretrial
Criminal Justice Processing, 22 Justice Quarterly 170, 187-88 (2005);
Stephen DeMuth, Racial and Ethnic Differences in Pretrial Release
Decisions and Outcomes: A Comparison of Hispanic, Black and White
Felony Arrestees, 41 Criminology 873, 895, 897 (2003); Ian Ayres &
Joel Waldfogel, A Market Test for Race Discrimination in Bail Setting,
46 Stanford L. Rev. 987 (1994).

37.	 S
 ee City of Riverside v. McLaughlin, 500 U.S. 44 (1991) (requiring
that probable cause determination for arrestees occur within 48
hours of arrest); Gerstein v. Pugh, 420 U.S. 103 (1975) (holding
that the Fourth Amendment requires a judicial determination of
probable cause as a prerequisite to extended restraint of liberty
following arrest).

See, e.g., Arpit Gupta, et al., The Heavy Costs of High
Bail: Evidence from Judge Randomization 3, 19 (May 2,
2016), available at
GuptaHansmanFrenchman.pdf (studying the assessment of
money bail in Philadelphia and Pittsburgh courts and finding that
the imposition of money bail led to a 6-9 percent yearly increase
in recidivism;) Laura & John Arnold Found., supra note 20, at 5
(“Compared to individuals released within 24 hours of arrest,
low-risk defendants held 2-3 days were 17 percent more likely
to commit another crime within two years. Detention periods
of 4-7 days yielded a 35 percent increase in re-offense rates.
And defendants held for 8-14 days were 51 percent more likely
to recidivate than defendants who were detained less than 24

38.	U.S. Const. amend. VIII
39.	 S
 ee U.S. v. Salerno, 481 U.S. 739, 752 (1987); Carlson v.
Landon, 342 U.S. 524, 546 (1952) (“[T]he very language of the
Amendment fails to say all arrests must be bailable.”).
40.	 Stack v. Boyle, 342 U.S. 1, 3 (1951).
41.	 Id. at 5.
42.	 Id.

28.	Justice Pol’y Inst., Bail Fail: Why the U.S. Should End the Practice
of Using Money for Bail 25 (2012), available at http://www.

See Salerno, 481 U.S. at 754 (holding that under Excessive Bail
clause “proposed conditions of release or detention” may not be

29.	 Id. at 26.

44.	 Id. at 751.

30.	Laura & John Arnold Found., supra note 27, at 5 (finding that “[l]
ow risk defendants held for 2-3 days were 22 percent more likely
to fail to appear than similar defendants (in terms of criminal
history, charge, background, and demographics) held for less
than 24 hours”).

45.	 R
 eno v. Flores, 507 U.S. 292, 302 (1993); see also Washington v.
Glucksberg, 521 U.S. 702, 721 (1997); Salerno, 481 U.S. at 749-51.
See Mathews v. Eldridge, 424 U.S. 319, 335 (1976); see also
Salerno, 481 U.S. at 751-52.
47.	 G
 riffin v. Illinois, 351 U.S. 12, 19 (1956); see also, Williams v. Illinois,
399 U.S. 235, 241 (1970).

31.	Colo. H.B. 13-1236 (2013). See also, Timothy R. Schnacke, Best
Practices in Bond Setting: Colorado’s New Pretrial Bail Law 20-24
(July 3, 2013), available at

Bearden v. Georgia, 461 U.S. 660, 672–73 (1983).
49.	 See, e.g., Pugh v. Rainwater, 572 F.2d 1053, 1058 (5th Cir. 1978)
(en banc) (“We have no doubt that in the case of an indigent,
whose appearance at trial could reasonably be assured by one of
the alternate forms of release, pretrial confinement for inability
to post money bail would constitute imposition of an excessive
restraint.”); Jones v. City of Clanton, 2015 WL 5387219, at *3
(M.D. Ala. 2015) (“Criminal defendants, presumed innocent,
must not be confined in jail merely because they are poor. Justice
that is blind to poverty and indiscriminately forces defendants
to pay for their physical liberty is no justice at all.”); Pierce v. City
of Velda City, 2015 WL 10013006, at *1 (E.D. Mo. 2015) (“No
person may, consistent with the Equal Protection Clause of the
Fourteenth Amendment to the United States Constitution, be
held in custody after an arrest because the person is too poor

32.	N.J. P.L. 2014, Ch. 31, §1-20. See also, Amer. Civil Liberties Union,
ACLU-NJ Hails Passage of NJ Bail Reform as Historic Day for Civil
Rights, (Aug. 4, 2014), available at
33.	Christine Stuart, Malloy Pitches Bail Reform as Part of
Connecticut’s Second Chance Society 2.0, New Haven Register,
Jan. 28. 2016, available at
34.	N.M. State Legis., Final Senate Joint Resolution 1, available at
pdf. See also, New Mexico House GOP, Senate Dems Reach Deal



12.30 (“Bail”).

to post a monetary bond.”); Walker v. City of Calhoun, 2016 WL
361580 at *49 (N.D. Ga. 2016) (“Any bail or bond scheme that
mandates payment of pre-fixed amounts for different offenses
to obtain pretrial release, without any consideration of indigence
or other factors, violates the Equal Protection Clause.”); Williams
v. Farrior, 626 F.Supp. 983, 985 (S.D. Miss. 1986) (“For the
purposes of the Fourteenth Amendment’s Equal Protection
Clause, it is clear that a bail system which allows only monetary
bail and does not provide for any meaningful consideration of
other possible alternatives for indigent pretrial detainee infringes
on both equal protection and due process requirements.”)

See, e.g., Ariz. Rev. Stat. Ann. § 20, Ch. 2, art. 3.5 (“Bail Bond
Agents and Bail Recovery Agents” appears in the section of the
code devoted to insurance regulation).
See, e.g., Alaska Stat. Ann. § 12.30.016(d) (imposing a minimum
cash bond of $250,000 for those persons charged with
misconduct regarding methamphetamines who have been
previously convicted of possession, manufacture, or delivery of
See, e.g. Ga. Code Ann. § 17-6-1(f)(1) (“the judge of any court
of inquiry may by written order establish a schedule of bails and
unless otherwise ordered by the judge of any court, a person
charged with committing any offense shall be released from
custody upon posting bail as fixed in the schedule.”); Alaska
Code § 12.30.016(d) (requiring $250,000 cash bond for offense
involving manufacture of methamphetamine after a prior
conviction for similar offense).

50.	Ariana Lindermayer, Note, What The Right Hand Gives:
Prohibitive Interpretation of the State Constitutional Right to Bail,
78 Fordham L. Rev. 267, 275 (2009).
51.	 See, e.g., Henley v. Taylor, 918 S.W.2d 713, 714 (Ark. 1996) ( “As
can be seen from the constitutional provision and the criminal
procedure rule [setting out conditions of release that may be
required for defendants who are likely to commit another crime],
a non-capital defendant’s absolute right to bail may only be
curbed by the setting of certain conditions upon his release, and
not its complete denial.”); Sprinkle v. State, 368 So.2d 554, 559
(Ala. Crim. App. 1978) (“In Alabama, an accused upon arrest and
before conviction, is entitled to bail as an absolute right provided
he has sufficient sureties. Bail may only be denied in capital
offenses, when the proof is evident or the presumption great.”);
see also Lindermayer, supra note 50, at 276.

63.	One example is the Illinois Pretrial Services Act of 1990, which
created a legal framework for pretrial services in the state.
Unfortunately, Administrative Office of the Illinois Courts found
that the Act “has become largely aspirational” in Cook County,
at least in part because of concerns about the credibility of risk
assessment determinations made by pretrial services officers. See
Illinois Supreme Court, Administrative Office of the Illinois Courts,
Circuit Court of Cook County Pretrial Operational Review 5 (2014)
available at

52.	 See, e.g., Rendel v. Mummert, 474 P.2d 824, 828 (Ariz. 1970)
(finding that the constitution “does not guarantee bail as a
matter of absolute right but is conditioned upon the giving
of ‘sufficient sureties.’ We are of the opinion that the words
‘sufficient sureties’ mean, at a minimum, that there is reasonable
assurance to the court that if the accused is admitted to bail,
he will return as ordered until the charge is fully determined”);
People ex rel. Hemingway v. Elrod, 322 N.E.2d 837, 840-41 (Ill.
1975) (“In our opinion the constitutional right to bail must be
qualified by the authority of the courts, as an incident of their
power to manage the conduct of proceedings before them, to
deny or revoke bail when such action is appropriate to preserve
the orderly process of criminal procedure.”); see also Lindermayer,
supra note 50, at 276.

See, e.g., Cal. Penal Code § 853.6(a)(1) (“In any case in which a
person is arrested for an offense declared to be a misdemeanor,
including a violation of any city or county ordinance, and does
not demand to be taken before a magistrate, that person shall,
instead of being taken before a magistrate, be released...”).
65.	Amer. Bar Ass’n, supra note 5.
66.	Schnacke, supra note 6.
67.	 Bearden, 461 U.S. at 672-73.
68.	 See Prison Pol’y Inst., supra note 10, at 1 (noting that 34% of
defendants are kept in jail pretrial solely because they are unable
to pay a cash bond, and most of these people are among the
poorest third of Americans); Marie VanNostrand, New Jersey
Jail Population Analysis at 13 (March 2013) (finding that 38.5%
of New Jersey pretrial inmates are held in custody “solely due to
their inability to meet the terms of bail” and that approximately
800 inmates held in custody “could have secured their release
for $500 or less”), available at
pdf; Human Rights Watch, The Price of Freedom: Bail and Pretrial
Detention of Low Income Nonfelony Defendants in New York
City at 20-21 (Dec. 2010) available at
default/files/reports/us1210webwcover_0.pdf (reporting that “at
any given moment, 39 percent of New York City’s jail population
consists of inmates who are in jail pretrial solely because they
have not posted bail”).

See, e.g., Rendel, 474 P.2d at 828, (defining sufficient sureties as a
reasonable assurance to the court that the defendant will appear
for trial if admitted to bail).
See, e.g., Ariz. Const. art II, § 22 (public safety exception for
felony offenses); Cal. Const. art. I § 12 (public safety exception
for sexual assault and violent felonies); Mo. Const. Art. 1 § 32.2
(“Notwithstanding section 20 of article I of this Constitution
[which states that '[a]ll persons shall be bailable upon sufficient
sureties'"], upon a showing that the defendant poses a danger
to a crime victim, the community, or any other person, the
court may deny bail or may impose special conditions which the
defendant and surety must guarantee.”).
55.	 S
 tate v. Cardinal, 147 Vt. 461, 465 (1986) ("[If a] defendant
violates conditions of release other than an appearance
condition, a court can impose increasingly more restrictive
conditions, as well as revoke the right to bail altogether, if the
court determines that no conditions of release will assure the
defendant’s appearance at trial.”) (citations omitted).

69.	 Turner v. Rogers, 131 S. Ct. 2507, 2520 (2011).
70.	 Id. at 2511.
71.	 Id. at 2519.

See, e.g., Cal. R. Super. Ct., County of Colusa Local R. Ct. 6.02 (“Bail”).

72.	 See Agreement to Settle Injunctive and Declaratory Relief Claims,
Mitchell v. City of Montgomery, No. 2:14-cv-186-MHT-CSC (M.D.
Ala. Nov. 17, 2014), available at
pdf (setting out a standard “Affidavit of Substantial Hardship
Form” to prevent indigent people from being jailed based on
their inability to pay fines and fees).

See, e.g., Ala. Code § 15, chapter 13 (“Bail”); Alaska Stat. Ann. §

See id. (creating a presumption that those with income at or

See Ga. Const. art. 1, § 1, ¶ XVII; Haw. Const. art. 1, § 12; Md.
Const. [Declaration of rights] art. 25; Ma. Const. Pt. 1, art. 26;
N.H. Const. Pt. 1, art. 33; N.Y. Const. art. 1, § 5; N.C. Const. art. 1,
§ 27; Va. Const. art. 1, § 9; W.V. Const. art. 3 § 5.
57.	 See, e.g., Ala. R. Cr. Proc. 7.2.



85.	Alaska Stat. Ann. § 12.30.016.

below 125% of the Federal Poverty Level, who do not have
substantial assets, will be considered indigent).

86.	 Stack, 342 U.S. at 4-5.

74.	 See Douglas L. Colbert, Thirty-Five Years After Gideon: The
Illusory Right to Counsel at Bail Proceedings, 1998 U. Ill. L.
Rev. 1, 5 (1998) (noting that “most state courts decline to
provide poor people with a lawyer during the first stage of a
criminal case when her presence truly matters: at the initial bail

87.	Amer. Bar Ass’n, supra note 5, at 17.
88.	 Id.
89.	 Id.
See, e.g., Statement of Interest of the United States, Jones v. City
of Clanton, 2:15-cv-34 (M.D. Ala. Feb. 13, 2015), at 14, available
at (“The use
of a more dynamic bail scheme...not only ensures adherence to
constitutional principles of due process and equal protection, but
constitutes better public policy. Individualized determinations,
rather than fixed-sum schemes that unfairly target the poor,
are vital to...providing equal justice for all.”); Walker v. City
of Calhoun, 2016 WL 361612 at *10 (N.D. Ga. 2016) (granting
preliminary injunction) (holding “any bail or bond scheme that
mandates payment of pre-fixed amounts for different offenses
to obtain pretrial release, without any consideration of indigence
or other factors, violates the Equal Protection Clause”).

75.	Douglas L. Colbert, Prosecution Without Representation, 59 Buff.
L. Rev. 333, 345 (2011).
76.	 Cf. Turner, 131 S. Ct. at 2519.
77.	For example, in November 2016, New Mexico voters may
decide to enact a constitutional amendment with the following
protections: “A person who is not detainable on grounds of
dangerousness nor a flight risk in the absence of bond and is
otherwise eligible for bail shall not be detained solely because
of financial inability to post a money or property bond. A
defendant who is neither a danger nor a flight risk and who has
a financial inability to post a money or property bond may file
a motion with the court requesting relief from the requirement
to post bond. The court shall rule on the motion in an expedited
manner.” Supra note 34, at 1.

91.	Alysia Santo, When Freedom Isn’t Free, The Marshall Project,
(Feb. 23, 2015),
92.	 See International Ass’n. of Chiefs of Police, Law Enforcement’s
Leadership Role in the Pretrial Release and Detention Process 7
(Feb. 2011), available at

78.	The use of bail schedules has been challenged through litigation
in California, Georgia, and Mississippi. See Compl., Buffin, v. City
and County of San Francisco, 2015 WL 6530384 (N.D. Ca. Oct. 28,
2015) (order certifying class); Walker v. City of Calhoun, Georgia,
2016 WL 361580 (N.D. Ga. 2016); Thompson v. Moss Point, 2015 WL
10322003 (S.D. Miss. 2015) (declaratory judgment).

93.	 See, e.g., The Nat’l Ass’n. of Pretrial Services Agencies, The
Truth About Commercial Bail Bonding in America (Aug. 2009)
available at
Facts%20and%20Positions%201.pdf; Shane Bauer, Inside the
Wild, Shadowy, and Highly Lucrative Bail Industry, Mother Jones,
May/June 2014, available at

79.	 See, e.g., Ga. Code Ann. § 17-6-1 (f)(1) (“Except as provided in
subsection (a) of this Code section or as otherwise provided in
this subsection, the judge of any court of inquiry may by written
order establish a schedule of bails and unless otherwise ordered
by the judge of any court, a person charged with committing any
offense shall be released from custody upon posting bail as fixed
in the schedule.”); Utah Code of Judicial Administration, Rule
4-302 (“The Uniform Fine/Bail Schedule Committee shall establish
a uniform fine/bail schedule setting forth recommended fine and
bail amounts for all criminal and traffic offenses, pursuant to the
Utah Code...[w]hen imposing fines and setting bail, courts should
conform to the uniform fine/bail schedule except in cases where
aggravating or mitigating circumstances warrant a deviation from
the schedule.”) Cal. Penal Code § 1269b(b) (“If a defendant has
appeared before a judge of the court on the charge contained
in the complaint, indictment, or information, the bail shall be in
the amount fixed by the judge at the time of the appearance.
If that appearance has not been made, the bail shall be in the
amount fixed in the warrant of arrest or, if no warrant of arrest has
been issued, the amount of bail shall be pursuant to the uniform
countywide schedule of bail for the county in which the defendant
is required to appear, previously fixed and approved as provided in
subdivisions (c) and (d).”).

94.	 See Justice Policy Inst., For Better or For Profit: How the Bail
Bonding Industry Stands in the Way of Fair and Effective Pretrial
Justice 40-42 (Sept. 2012), available at http://www.justicepolicy.
95.	 See Brian R. Johnson & Ruth S. Stevens, The Regulation and
Control of Bail Recovery Agents: An Exploratory Study, 38 Crim.
Justice Rev. 190, 193 (2013); Brian Liptak, Illegal Globally, Bail
for Profit Remains in U.S., N.Y. Times, Jan. 29, 2008. See also
Schnacke, supra note 6, at 36-37 (examining the historical
origins of the modern surety system, comprised of “what we
might now call unsecured bonds using co-signors, with nobody
required to pay any money up-front, and with the security on any
particular bond coming from the sureties...who were willing to...
acknowledge the amount potentially owed upon default.”).
96.	Ky. Rev. Stat. Ann. § 431.510 (West 1976); 725 Ill. Comp. Stat.
5/103-9 (1986); 725 Ill. Comp. Stat. 5/110-7 & -8.

80.	Ala. Code § 15-13-103.

97.	Fred Contrada, Bail Bondsmen are a thing of the Past in Mass.,
Mass Live (2014),

81.	 Id.
82.	Ala. Rules of Crim. Proc., Rule 7.2(b).

98.	The American Bar Association standards for pretrial release
reflect these principles, noting that the “presumption that
defendants should be released under the least restrictive
conditions necessary to provide reasonable assurance they will
not flee or present a danger is tied closely to the presumption
favoring release generally. It has been codified in the Federal
Bail Reform Act and the District of Columbia release and pretrial
detention statute, as well as in the laws and court rules of a
number of states.” Amer. Bar Ass’n, supra note 5, at 39-40.

Jones v. City of Clanton, No. 215CV34-MHT, 2015 WL 5387219, at
*3 (M.D. Ala. 2015).
84.	Alaska Stat. Ann. § 12.30.016. Alaska is in the process of
reviewing its bail practices as part of the Justice Reinvestment
Initiative, and this provision may change. See Alaska Crim. Justice
Comm’n., Justice Reinvestment Initiative Report 8 (Dec. 2015),
available at
intiative_report_to_acjc_12-9.pdf. (identifying “one of the likely
contributors to pretrial length of stay in Alaska is the use of
secured money bail”).


99.	Spurgeon Kennedy, Freedom and Money – Bail in America,
available at


116.	Matt O’Keefe, Court Appearance Notification System: 2007
Analysis Highlights, LPSCC (2007), available at http://www.

100.	 Id.
101.	Pretrial Services Agency for the District of Columbia, Research
and Data, Performance Measures, available at https://www.psa.

117.	Wendy F. White, Court Hearing Call Notification Project, Coconino
County, AZ: Criminal Coordinating Council and Flagstaff Justice
Court (2006)
Hearing%20Notification%20Project%20(2006).pdf. (finding that
advance phone calls to defendants reduced the percentage of
failure to appear from over 25% to less than 13%).

102.	 Id.
103.	 Id.
104.	 See, e.g., Ariz. St. Code of Jud. Admin. §5-201; Colorado Rev. Stat.
Ann. §16-4-105; New Jersey P.L. 2014, Ch. 31, §1-20.
105.	National Center for State Courts, Pretrial Services and

118.	Schnacke, supra note 115, at 86 (2012) (finding that phone calls
from volunteers increased court appearance rates from 79%
to 88%).

106.	Colo. Rev. Stat. Ann. §16-4-105.
107.	Kathy Rowings, Nat’l Ass’n of Counties, County Jails at a Crossroads
– Mesa County, CO, (July 6, 2015), available at http://www.naco.

119.	Pretrial Justice Inst., Using Technology to Enhance Pretrial
Services: Current Applications and Future Possibilities 14-16,
available at
PRETRIAL%20SERVICES.pdf (discussing the potential for
implementation of new technologies like e-mail and text
message reminders).

108.	See Pretrial Justice Inst., Mesa County Pretrial SMART Praxis (2013),
available at
109.	 Id.

120.	See Marie VanNostrand, et al., Pretrial Justice Inst., State of the
Science of Pretrial Release Recommendations and Supervision 29
(2011), available at
PJI percent20State percent20of percent20the percent20Science
percent20Pretrial percent20Recommendations percent20and
percent20Supervision percent20(2011).pdf (finding that “review
of supervision strategies in numerous pretrial services agencies
nationally revealed disparate practices for what constitutes
pretrial supervision. The frequency and types of contacts ranged
from monthly phone contacts with an automated calling system
to daily in-person reporting by defendants”).

110.	Nat’l Ass’n of Pretrial Services Agencies, Standards on Pretrial Release,
Third Edition 14 (Oct. 2004), available at
111.	 Id.
See, e.g., Superior Court of Fulton County, Pretrial Services –
Savings to Taxpayers, available at
pretrial/savings.php (cost-benefit analysis estimating a $51.8
million dollar savings in fiscal year 2011 comparing the pretrial
services budget in Fulton County, Georgia to the cost of keeping
defendants in jail); Alex Piquero, Cost-Benefit Analysis for Jail
Alternatives and Jail 5 (Oct. 2010) available at http://criminology. (comparing daily costs of jail and pretrial
services in Broward County, Florida and concluding that “cost
savings via pretrial (in lieu of jail) are over $100 million in both 2009
and 2010”); Marie VanNostrand, Alternatives to Pretrial Detention:
Southern District of Iowa (2010), available at https://www.pretrial.
VanNostrand%202010.pdf (evaluating pretrial services programs in
Iowa, showing pretrial services resulted in cost savings of $15,393
per defendant and a cost avoidance of $5.33 million in the 2008
and 2009 fiscal years).

121.	 Id.
122.	John S. Goldkamp & Michael D. White, Restoring Accountability
in Pretrial Release: The Philadelphia Pretrial Release Supervision
Experiments, 2 J. of Experimental Criminology 143 (2006).
123.	Laura & John Arnold Found., supra note 20, at 6.
124.	 Id.
125.	 Id.
126.	Sharon Aungst, ed., Partnership for Community Excellence,
Pretrial Detention and Community Supervision: Best Practices and
Resources for California Counties, 13 (2012), available at http://

See United States v. Scott, 450 F.3d 863, 868 (9th Cir. 2006)
(“[O]ne who has been released on pretrial bail does not lose his or
her Fourth Amendment right to be free of unreasonable seizures.”)
(quoting Cruz v. Kauai County, 279 F.3d 1064, 1068 (9th Cir. 2002)).

127.	Eric Markowitz, Electronic Monitoring Has Become the New
Debtors Prison, Newsweek, Nov. 26, 2015, available at http://

114.	 See, e.g., id. (holding that mandatory random drug testing as a
condition of release constituted an unreasonable search under
the Fourth Amendment where there was no showing that
drug testing advanced goals of public safety of guaranteeing
appearance at trial); United States v. Karper, 847 F. Supp. 2d 350
(N.D.N.Y. 2011) (finding that mandatory electronic monitoring
conditions on defendants accused of certain sex offenses violates
due process and constitutes excessive bail); United States v.
Polouizzi, 697 F. Supp. 2d 381 (E.D.N.Y. 2010) (same).

128.	 See Aungst, supra note 126, at 13; Keith Cooprider & Judith
Kerby, Practical Application of Electronic Monitoring at the Pretrial
Stage, 1 Federal Probation 28, 33 (1990) (noting that “[t]he higher
violation rate(s) of electronically monitored defendants is probably
related to the fact that, as a rule, the riskier clients (serious charge
in terms of felony class, recidivist, already on some other form
of community supervision, FTA history, chemical dependency,
etc.) are supervised with electronic monitoring”); Timothy P.
Cadigan, Electronic Monitoring in Federal Pretrial Release, 55
Federal Probation 1, 26 (1991) (finding “the electronic monitoring
defendants were charged more frequently with serious offenses”
than average, which helps explain their higher rates of re-arrest)
29-30; Albert J. Lemke, Institute for Court Management, Evaluation
of the Pretrial Release Pilot Program in the Mesa Municipal Court
50 (2009), available at
Lemke_EvalPretrialReleaseProg.ashx; Nat’l Inst. of Justice, Office

115.	 See Timothy R. Schnacke, et al., Increasing Court-Appearance
Rates and Other Benefits of Live-Caller Telephone Court-Date
Reminders: The Jefferson County, Colorado FTA Pilot Project
and Resulting Court Date Notification, 48 Court Rev. 86 (2012);
Wendy F. White, Court Hearing Call Notification Project, Coconino
County, AZ: Criminal Coordinating Council and Flagstaff Justice
Court (2006).



Chester L. Britt, et al., Drug Testing and Pretrial Misconduct: An
Experiment on the Specific Deterrent Effect of Drug Monitoring
Defendants on Pretrial Release, J. of Crime and Delinquency (1992)
(finding that drug testing only leads to a slight decrease in pretrial
arrest but does not alter failure to appear).

of Justice Programs, Electronic Monitoring Reduces Recidivism 2
(2011), available at
(discovering quantitatively “significant decreases in the failure rate
for all groups of offenders, and the decreases were similar for all
age groups”).
129.	VanNostrand, supra note 120, at 27.

144.	VanNostrand, supra note 120, at 24.

See generally Avlana K. Eisenberg, Mass Monitoring, 90 S. Cal. L. Rev.
__ (forthcoming 2017).

145.	A series of Supreme Court cases has addressed the Fourth
Amendment considerations surrounding drug testing in a variety
of settings. While government-ordered drug tests indisputably
constitute searches for Fourth Amendment purposes, the Court
has generally scrutinized the particular contexts in which such
tests are imposed to determine whether they are constitutionally
reasonable searches. Compare Ferguson v. City of Charleston, 532
U.S. 67 (2001) (holding that mandatory urinalysis of pregnant
mothers violates Fourth Amendment) and Chandler v. Miller, 520
U.S. 305 (1997) (striking down drug testing for candidates for
designated state offices) with Bd. of Educ. Of Independent Sch.
Dist. No. 92 of Pottawatomie Cty. v. Earls, 536 U.S. 822 (2002)
(upholding mandatory drug testing of high school students
participating in extracurricular activities) and Skinner v. Railway
Labor Executives’ Ass’n, 489 U.S. 602 (1989) (upholding drug and
alcohol testing of railway operators).

131.	 See Nat’l Inst. of Justice, supra note 128, at 1-4 (describing
the results of a survey conducted by Florida State University’s
Center for Criminology and Public Policy Research comparing the
experiences of more than 5,000 medium-and high-risk offenders
who were monitored electronically to more than 266,000
offenders not placed on monitoring over six years). See also
M.M., Living With an Ankle Bracelet: Freedom, With Conditions,
The Marshall Project, (July 16, 2015), available at https://www.
132.	Nat’l Inst. of Justice, supra note 128.
133.	 Id. at 2.
134.	 Id.

146.	See Scott, 450 F.3d at 870 (“Drug use during pretrial release may
result in a defendant’s general unreliability or, more nefariously,
an increased likelihood of absconding. Whether this is plausible
depends on whether drug use is a good predictor of these harms
– a case that must be established empirically by the government
when it seeks to impose the drug testing condition.”); Berry
v. District of Columbia, 833 F.2d 1031, (D.C. Cir. 1983) (holding
that to justify mandatory pretrial drug testing “the District must
proffer reliable evidence, statistical or otherwise, from which the
trial court can reasonably conclude that drug testing makes it
significantly more likely that an arrestee will commit crimes or fail
to appear for scheduled court dates”).

135.	 Id.
136.	 See Polouizzi, 697 F. Supp. 2d at 389 (“Required wearing of
an electronic bracelet, every minute of every day, with the
government capable of tracking a person not yet convicted as if
he were a feral animal would be considered a serious limitation
on freedom by most liberty-loving Americans.”).
137.	 See, e.g., United States v. Karper, 847 F. Supp. 2d 350 (N.D.N.Y.
2011) (finding that mandatory electronic monitoring conditions
on defendants accused of certain sex offenses violates due
process and constitutes excessive bail); Polouizzi, 697 F. Supp. 2d
381 (same). But see, e.g., United States v. Gardner, 523 F. Supp.
2d 1025 (N.D. Cal. 2007) (rejecting due process and excessive bail
challenges to mandatory electronic monitoring).

147.	Markowitz, supra note 127.
148.	 See Jason Blalock, Profiting from Probation: America’s “Offenderfunded” Probation Industry, Human Rights Watch (2014),
available at 2014/02/05/profitingprobation/americas-offender-funded-probation-industry.

138.	United States v. Jones, 132 S. Ct. 945, 955 (2012) (Sotomayor,
J., concurring) (“In cases involving even short-term monitoring,
some unique attributes of GPS surveillance...will require particular
attention. GPS monitoring generates a precise, comprehensive
record of a person’s public movements that reflects a wealth of
detail about her familial, political, professional, religious, and
sexual associations.”).

149.	Charles Summers & Tim Willis, Bureau of Justice
Assistance, Pretrial Risk Assessment: Research Summary
2 (2010), available at

139.	All states, except Hawaii and the District of Columbia, charge fees
to those who are under electronic monitoring. Joseph Shapiro,
As Court Fees Rise, The Poor Are Paying the Price, NPR, (May 19,
2014), available at

150.	Cynthia Mamalian, State of the Science of Pretrial Risk
Assessment 7 (2011), available at
Science%20Pretrial%20Risk%20Assessment%20 (2011).pdf; see
also, Pretrial Justice Institute, Pretrial Risk Assessment: Science
Provides Guidance on Assessing Defendants 3 (2015), available

140.	Markowitz, supra note 127.
141.	VanNostrand, supra note 120, at 24.
142.	Pretrial Justice Inst., Survey of Pretrial Programs 47 (2009),
available at

151.	 See Council of State Govts, Risk and Needs Assessment and Race
in the Criminal Justice System, May 31, 2016, available at https:// (discussing the ways
in which a risk assessment tool must be used properly and
accurately in order to ensure that it does not perpetuate racial
biases in the criminal justice system).

143.	Stefan Kapsch & Louis Sweeny, Bureau of Just Assistance,
Multnomah County DMDA Project: Evaluation Final Report
(1990); see also VanNostrand, supra note 120, at 21-22 (discussing
separate studies of Washington D.C. and Milwaukee which found,
respectively, that drug testing pretrial did not reduce failure to
appear and that increasingly severe judicial sanctions changed
nothing); Michael R. Gottfredson, et al., U.S. Dept. of Justice,
Nat’l. Inst. of Justice, Evaluation of Arizona Pretrial Services Drug
Testing Programs: Final Report to the National Institute of Justice
(1990) (finding that “knowledge of drug test results does not
appreciably improve the ability to estimate pretrial misconduct”);


152.	 See Timothy Cadigan, et al., Implementing Risk Assessment in the
Federal Pretrial Services System, 75 Fed. Prob. 30, 31 (Sept. 2011)
(noting that “[r]isk tools, while tremendously useful in improving
agency decision making and ultimately release recommendations,
have limitations...[and] the tool should not be followed blindly”).
153.	 See id. at 32.
154.	Jeff Larson et al., How We Analyzed the COMPAS Recidivism


Defendants to Detain Prior to Trial, June 26, 2015, available at

Algorithm, May 23,2016, available at
155.	Samuel Wiseman, Fixing Bail, 84 Geo. Was. L. Rev. 417, 442 n. 145
(2016) (citing Charles Summers & Tim Willis, Bureau of Justice
Assistance, Pretrial Risk Assessment Research Summary 2 (2010);
Vera Inst. of Justice, Evidence-Based Practices in Pretrial Screening
and Supervision, 2–3 (2010)). The use of risk assessment in the
federal system has been limited: In 2011, there were an average
of 26,000 pretrial defendants each quarter in the federal system,
but only 4,000 reports with Pretrial Risk Assessment scores.
Cadigan, supra note 152, at 30, 33.

173.	 See Shima Baradaran & Frank L. McIntyre, Predicting Violence, 90
Tex. L. Rev. 497, 553 (2012) (finding among pretrial defendants
nationally that “almost exactly half of those reality have
a lower than 20% chance of rearrest, while an equivalent number
of those released have a higher than 20% chance of committing a
crime”) (emphasis added).
174.	Samuel Wiseman, Fixing Bail, 84 Geo. Was. L. Rev. 417, 467-68 (2016).

156.	For example, New Jersey has partnered with the Laura and John
Arnold Foundation to implement a pretrial risk assessment tool
as part of its bail reform efforts. See N.J. State Legis., FY 20152016 Budget, Judiciary Response 3, available at http://www.njleg. New
Mexico is also investigating the use of risk assessment. See N.M.
Senate, Joint Resolution 1, Fiscal Impact Report, Feb. 16, 2016,
available at

175.	Vernon Lewis Quinsey, et al., Violent Offenders: Appraising and
Managing Risk 171 (1998).
176.	Baradaran & McIntyre, supra note 173, at 553.
177.	 See, e.g., Justice Pol’y Inst., supra note 28, at 13-14 (discussing
the potential interruption to jobs, housing, health insurance,
familial and community well-being that can occur as a result of
pretrial detention).
178.	See supra note 25.

157.	 See, e.g., Pretrial Justice Inst., The Colorado Pretrial Assessment
Tool, Revised Report, Oct. 19, 2012, available at http://www.
pdf; Ohio Dep’t of Rehabilitation and Correction, Ohio Risk
Assessment System, Nov. 17, 2014, available at http://www.drc.; Northpointe, Northpointe Software
Suite, available at

179.	Even one of the scholars most critical of actuarial risk assessment
tools has observed that the bail context constitutes “one instance
of actuarial progress that unquestionably has benefited poor and
minority communities.” Bernard E. Harcourt, Against Prediction:
Profiling, Policing, and Punishing in an Actuarial Age 216 (2007).
180.	 See Laura & John Arnold Found., supra note 163.
181.	Kenneth Rose, Virginia Dep’t of Criminal Justice Services, A “New
Norm” for Pretrial Justice in the Commonwealth of Virginia 6
(2013), available at

158.	A 2011 report noted that the cost of validating a risk assessment
instrument could range from $20,000 to $75,000, depending
on the jurisdiction and the type of study that is being done. See
Mamalian, supra note 150, at 35 n. 91.

182.	 See Mecklenburg County Criminal Justice Services Planning, Jail
Population Trend Report, January-March 2016, available at http://
In 2008, the average pretrial jail population was 1,953 people
but by December 2015 the daily population was reduced to
817 people. Nat’l. Ass’n. of Counties, Effectively Framing the
Pretrial Justice Narrative, Webinar, April 14, 2016, available at

159.	Indeed, two of the most recent states to undertake major
reform, New Jersey and New Mexico, have explored the use of
national APRAI models. See N.J. State Legis., supra note 156 at 3;
N.M. Senate, supra note 156 at 1.
160.	The Arnold Foundation has funded the Criminal Justice Policy
Program on work that is unrelated to this primer.
161.	Laura & John Arnold Found., Developing a National Model for
Pretrial Risk Assessment (Nov. 2013), available at http://www.

183.	Pretrial Justice Inst., The Transformation of Pretrial Services in
Allegheny County, Pennsylvania: Development of Best Practices
and Validation of Risk Assessment, vii (Oct. 9, 2007), available at

162.	VanNostrand, supra note 120, at 29.
163.	 See Laura & John Arnold Found., Results from the First Six
Months of the Public Safety Assessment - Court in Kentucky (July
2014), available at

168.	Id.

184.	 See Sonja B. Starr, Evidence-Based Sentencing and the Scientific
Rationalization of Discrimination, 66 Stan. L. Rev. 803, 842 (2014)
(“[In the sentencing context,] the models are designed to predict
the average recidivism rate for all offenders who share with the
defendant whichever characteristics are included as variables
in the model. If the model is well specified and based on an
appropriate and large enough sample, then it might perform this
task well. But because individuals vary much more than groups
do, even a relatively precisely estimated model will often not do
well at predicting individual outcomes in particular cases.”).

169.	 Id.

185.	Amer. Bar Ass’n, supra note 5.

170.	 Id.

186.	 Executive Office of the President, Big Data: A Report on
Algorithmic Systems, Opportunity, and Civil Rights 21 (May 2016),
available at

164.	Laura & John Arnold Found, Developing a National Model for
Pretrial Risk Assessment, (Nov. 2013), available at http://www.
165.	 Id.
166.	 Id.
167.	 Id.

171.	 Id.
172.	Laura & John Arnold Found., More Than 20 Cities and States
Adopt Risk Assessment Tool to Help Judges Decide Which

187.	 See, e.g., Angele Christin et al., Courts and Predictive Algorithms



“most modern sentencing systems” and the Equal Protection
underpinnings of that practice). See also Starr, supra note 184, at
823-24 (arguing that risk assessment instruments in the sentencing
context which rely on “statistical generalizations about groups”
based on gender and socioeconomic status violate the Equal
Protection Clause). Observers have reached competing conclusions
about whether including sex as a variable in an actuarial risk
assessment would survive equal protection scrutiny. See Melissa
Hamilton, Risk-Needs Assessment: Constitutional and Ethical
Challenges, 52 Am. Crim. L. Rev. 231, 250-53 (2015) (outlining
divergent views by scholars and commentators). The Wisconsin
Supreme Court recently upheld the use of a risk assessment
instrument in the sentencing context that included gender as a risk
factor. See Loomis, 881 N.W.2d at 766. Significantly, however, the
court in that case considered a claim based on due process, not
equal protection. Id.

7-8 (Oct. 27, 2015) (outlining ways that inputting and
interpretation of data may result in shifting discretion rather than
rationalizing decisions), available at http://www.datacivilrights.
188.	 See, e.g., Christopher Slobogin, Risk Assessment and Risk
Management in Juvenile Justice, 27 Wtr. Crim. Justice 10,
16-17 (2013).
189.	 See, e.g., Michael Tonry, Legal and Ethical Issues in the Prediction
of Recidivism, 26 Fed. Sent. Rep. 167, 173 (2014) (concluding that
reliance on criminal histories intertwined with socioeconomic
factors, such as age at time of first arrest, custody status at
time of first arrest, and total number of convictions, inherently
disadvantage minority defendants).
190.	 See, e.g., Marc Mauer & Ryan S. King, The Sentencing Project,
Uneven Justice: State Rates of Incarceration by Race and Ethnicity
4 (July 2007) (finding that “[t]he American prison and jail system
is defined by an entrenched racial disparity in the population of
incarcerated people”).

201.	 See, e.g., Personnel Adm’r of Massachusetts v. Feeney, 442 U.S.
256, 279 (1979).
202.	Fuentes v. Shevin, 407 U.S. 67, 81 (1972).

191.	For example, the nine factors that the Virginia Pretrial Risk
Assessment Instrument (VPRAI) considers include whether the
defendant has lived at their current residence for a year or more
and whether he or she has been employed continuously or served
as a primary caretaker for children over the past two years.
Virginia Pretrial Risk Assessment Instrument (VPRAI) Instruction
Manual, 5-7
files/publications/corrections/virginia-pretrial-risk-assessmentinstrument-vprai.pdf. Factors like these may systematically
import racial bias. See, e.g., Matthew Desmond, Eviction and the
Reproduction of Urban Poverty, 118 Am. J. of Sociology 88, 110
(July 2012) (explaining the structural reasons why black women
are overrepresented in evictions in Milwaukee).

203.	See Mathews v. Edridge, 424 U.S. 319 (1976).
204.	 Salerno, 481 U.S. at 751.
205.	 Loomis, 881 N.W.2d at 760.
206.	 Id. at 769.
207.	 See generally Salerno, 481 U.S. at 750-51; Mathews, 424 U.S. at
208.	See Hamilton, supra note 200, at 271 (2015) (considering potential
due process requirements in various contexts).
209.	 See John S. Goldkamp, Danger and Detention: A Second
Generation of Bail Reform, 76 J. Crim. Justice & Criminology 1,
4 (1985) (listing five common objections to cash bail systems
including that judges often set bond at a level without relation
to the dangerousness of the defendant and which may handicap
the defendant at later stages in the criminal procedure).

192.	Atty. Gen. Eric Holder, Remarks of Attorney General Eric Holder
at National Association of Criminal Defense Lawyers 57th Annual
Meeting (Aug. 1, 2014), available at

210.	 See Alaska Stat. Ann. § 12.30.011(d)(2); Arizona Rev. Stat. Ann. §
13-3961; Colorado Rev. Stat. Ann. § 16-4-101; D.C. Code §23-1322;
Florida Const. Art. 1 § 14; Hawaii Rev. Stat. §804-3; Illinois 725
Ill. Comp. Stat. 5/110-6.1; Indiana Code Ann. §35-33-8-2 (only for
murder charges where “the proof is evident or the presumption
strong”); Louisiana Code Crim. Proc. Ann. art. 330.1; Maine Rev.
Stat. Ann. tit. 15 § 1027, § 1029 (for crimes that are or were
formerly capital offenses); Maryland Rules Crim. Proc. § 5-202;
Massachusetts G. L. 276 § 58A; Michigan §765.5; Mississippi
Const. Art. 3, §29; Mo. Const. Art. 1 § 32.2; New Jersey P.L. 2014,
Ch. 31, §1-20; Ohio Rev. Code Ann. §2937.222; Oregon Rev. Stat.
Ann. § 135.240; Pennsylvania Const. Art. 1 § 14; Rhode Island
Gen. Laws Ann. § 12-13-1.1; Texas Const. Art. 1 § 11a; Washington
Rev. Code Ann. § 10.21.040, § 10.21.060 (for capital offenses and
offenses punishable by life in prison); Wisconsin Const. Art. 1 § 8.

193.	 Id.
194.	See Harcourt, supra note 179, at 220 (2007) (discussing the
“ratchet effect” that may occur when using risk assessment in the
pretrial context).
195.	 See Atty. Gen. Eric Holder supra note 192 (“Criminal sentences
must be based on the facts, the law, the actual crimes committed,
the circumstances surrounding each individual case, and the
defendant’s history of criminal conduct. They should not be
based on unchangeable factors that a person cannot control.”).
196.	 See, e.g., Sonja Starr, Risk Assessment Era: An Overdue Debate, 27
Fed. Sent’g Rep. 205 (Apr. 2015)
197.	 See, e.g., State v. Loomis, 881 N.W.2d 749 (Wisc. 2016) (examining
the use of risk assessment at sentencing); State v. Duchay, 647
N.W.2d 467, 2002 WL 862458, at *1-2 (Wis. Ct. App. May 7, 2002)
(holding that a court’s reliance on a risk assessment instrument
in sentencing was not a due process violation because the
defendant did not show that the information was inaccurate);
Malenchik v. State, 928 N.E.2d 564, (Ind. 2010) (upholding the use
of a risk assessment tool in the sentencing context).

See, e.g., Shima Baradaran Baughman, Restoring the Presumption
of Innocence, 72 Ohio State L.J. 723 (2011) (questioning bail
and pretrial detention given the presumption of innocence);
R.A. Duff, Pre-trial Detention and the Presumption of Innocence
(2012), available at (assessing
whether pretrial detention can coexist with the presumption of
innocence); Sandra G. Mayson, Dangerous Defendants, University
of Pennsylvania Law School, Public Law Research Paper No. 16-30
(August 15, 2016), available at

198.	 See, e.g., Fisher v. University of Texas at Austin, 133 S.Ct.
2411, 2419 (2013) (government policies that rely on “suspect
classifications” will survive judicial scrutiny only if they are
narrowly tailored to serve a compelling governmental interest).

212.	 Id. at 746. See also Bell v. Wolfish, 441 U.S. 520, 535-39 (1979)
(“In evaluating the constitutionality of conditions or restrictions
of pretrial detention that implicate only the protections
against deprivation of liberty without due process of law, we
think the proper inquiry is whether those conditions amount
to punishment of the detainee...[I]f a particular condition
or restriction of pretrial detention is reasonably related to a

199.	 See United States v. Virginia, 518 U.S. 515, 532-33 (1996).
200.	See Carissa Byrne Hessick, Race and Gender as Explicit Sentencing
Factors, 14 J. Gender Race & Justice 127 (2010) (discussing the
“explicit commitment to ensuring that a defendant’s sentence
is not affected by the defendant’s race or gender” present in



226.	 Salerno, 481 U.S. at 747.

legitimate government objective, it does not, without more,
amount to ‘punishment.’”).

227.	 Id. at 750.

213.	 Salerno, 481 U.S. at 747.

228.	 Id.

214.	 Id.

229.	Laura & John Arnold Found., supra note 20, at 4-5.

215.	 Id. at 750.

230.	N.J.P.L. 2014, Ch. 31, § 1 (defining “eligible defendants” for the
purpose of administering risk assessment based on the crime for
which a defendant is charged).

216.	 Id. at 750 (internal quotations and citations omitted).
217.	 Id. at 751.

231.	 See D.C.Code §23-1322; 18 U.S.C § 3142(e).

218.	 Id. at 755.

232.	Cf. Kate Stith, The Arc of the Pendulum: Judges, Prosecutors, and
the Exercise of Discretion, 117 Yale L. J. 1420 (2008) (arguing that
the federal sentencing guidelines have increased prosecutorial
discretion, not only in charging decisions but also in sentencing).

219.	 Id. at 751.
220.	 Id.
221.	 Id. at 751–52.

233.	D.C. Code §23-1322(b)(1); 18 U.S.C § 3142(f).

222.	The right to counsel at bail hearings is crucial not only to
reduce unnecessary pretrial detention, but also to ensure that
defendants are able to preserve their right to a fair trial. Having
counsel involved at an early stage allows the attorney to begin a
prompt investigation of the case and build trust with the client.
See Colbert, supra note 74, at 6.

234.	 Salerno, 481 U.S. at 751.
235.	D.C.Code § 23-1322(e).
236.	 See, e.g., Ohio Rev. Code § 2937.222; Washington Rev. Code Ann.
§10.21.050 (mirroring almost exactly the D.C. language); Mass.
Gen. Laws Ann. ch. 276, § 58A (“The nature and seriousness of
the danger posed to any person or the community that would
result by the person’s release, the nature and circumstances of
the offense charged, the potential penalty the person faces, the
person’s family ties, employment record and history of mental
illness, his reputation, the risk that the 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, his record of convictions, if any, any illegal drug
distribution or present drug dependency, whether the person is
on bail pending adjudication of a prior charge...”). New Jersey’s
preventative detention statute contains a discussion of similar
factors in addition to the “release recommendation of the pretrial
services program obtained using a risk assessment instrument
under section 11 of P.L.2014, c.31 (C.2A:162-25).” N.J. P.L.2014,
c.31 C.2A:162-20(6)(f).

223.	D.C. Code § 23-1322(h)(1). Significantly, the statute allows for
20 day extensions when good cause is shown if a judicial officer
approves of the requested extension.
224.	Vt. Stat. tit. 13, § 7553b.
225.	 See Thomas M. O’Brien, The Undoing of Speedy Trial in New
York: the “Ready Rule,” N.Y. Law Journal (Jan. 14, 2014), available
curindex=0&curpage=ALL (noting the case of Kalief Browder who
was jailed for three years in Rikers Island awaiting trial); William
Glaberson, Justice Denied: Inside the Bronx’s Dysfunctional
Court System, N. Y. Times, available at http://www.nytimes.



Austin Hall, 1st floor | 1515 Massachusetts Ave | Cambridge, MA 02138
Phone: 617-495-4848 |