In the Shadows: A Review of the Research on Plea Bargaining

In the Shadows: A Review of the

Research on Plea Bargaining

Ram Subramanian, Léon Digard, Melvin Washington II, and Stephanie Sorage

September 2020

In the Shadows: A Review of the Research on Plea Bargaining iii

From the Director

When most people think of the American justice

system, they likely picture a courtroom with lawyers,

a judge, and a jury waiting to determine the facts of

the case and provide a just outcome. But the majori

of people “found guil” in America never stand trial:

their fate is determined in a courthouse hallway or

prosecutor’s oﬃce via a quick conversation between

aorneys.

Plea bargaining is more a part of the American justice

system than the formal trial and, in fact, makes up the

vast majori of criminal justice transactions today.

Only 2 percent of federal criminal cases—and a similar

number of state cases—are brought to trial. More than

90 percent of convictions, at both federal and state

levels, are the result of guil pleas. Plea bargaining is

so ndamental to the system that even in 1970, Chief

Justice Warren Burger of the U.S. Supreme Court

estimated that a 10 percent reduction in guil pleas

would require doubling the amount of judicial capaci

in the system. Scholars in recent years have suested

that the criminal legal system could be brought to a

halt by a mass resal to plead guil.

And yet lile is known about plea bargaining.

Pleas are oﬀered and retracted at the unfeered

discretion of prosecutors. Bargains themselves are

undocumented and largely unchallenged, save for a

few formal questions meant to establish that the plea

is “voluntary, intelligent, and knowing.” To understand

plea bargaining, then, we must depend on a small but

growing body of research. rough interviews, data

gathered by courts, and other means, scholars are

aempting to understand the factors that inﬂuence

plea bargaining as well as whether a plea bargain is a

“bargain” at all.

ese studies are vital to understanding a process that

is so central to how the criminal legal system currently

operates, but they remain woelly inadequate and

incomplete. e sheer lack of trials, for example,

means that it is diﬃcult to ﬁnd analogous cases to

compare to those that end in pleas. In addition, the

dynamic and recursive nature of bargaining is diﬃcult

to isolate into discrete questions for study. Finally,

the most important voices—the people subject to

these bargains—are largely absent from these studies.

Instead, data is provided by police, courts, prosecutors,

and jails. is one-sided understanding of the bargain

explains a good deal about how the legal characteristics

of a case, or pressures such as increasing caseloads,

drive prosecutorial bargaining, but it oﬀers lile

insight into the reasons that people trade their right to

a trial for a faster and more certain conviction.

But why does it maer?

e history of the American justice system is a

history of mass incarceration, with wildly disparate

consequences for Black and white people. We are

faced with substantial evidence that people are put

in untenable positions aer arrest. ey are kept in

jails away from their families and communities if

they cannot aﬀord cash bail or an aorney to argue

iv Vera Institute of Justice

for them. And although representation is guaranteed,

the tremendous caseload borne by public defenders

means that “representation” before trial is likely to be

pernctory and impersonal. Pretrial incarceration has

been deﬁnitively linked to the likelihood of conviction,

and most convictions are obtained via plea. And

although the Supreme Court has indicated that even

knowing you may be put to death if you face trial is

not “coercive” pressure to plead guil, it is diﬃcult

to say that an incarcerated person faced with the loss

of their income, housing, family, and communi is

truly free to make a choice. Even aer people return

to their communities, a conviction—whether it stems

from a guil plea or a jury verdict—carries collateral

consequences that will follow them for years, if not the

rest of their lives.

Today, the American justice system is in crisis, but

it is also in a moment of unique opportuni. is

review and analysis of the available literature on plea

bargaining—released by Vera and the Safe and Justice

Challenge—represents decades of work by dozens

of scholars. It brings light to the shadowed hallways

where the majori of justice is transacted and new

aention to the ways the criminal legal system has

become an ad hoc administrative process.

Elizabeth Swavola

Acting Project Director

Vera Institute of Justice

About this report

is report is one of a series that the Vera Institute of Justice (Vera) is

releasing with the Safe and Justice Challenge—the John D. and Catherine

T. MacArthur Foundation’s initiative to reduce overincarceration by

changing the way America thinks about and uses jails. e initiative

is supporting a network of competitively selected local jurisdictions

commied to ﬁnding ways to safely reduce jail incarceration. Other

publications in the series to date include:

› Incarceration’s Front Door: e Misuse of Jails in America;

› e Price of Jails: Measuring the Taxpayer Cost of Local Incarceration;

› Overlooked: Women and Jails in an Era of Reform;

› Out of Sight: e Growth of Jails in Rural America;

› Divided Justice: Trends in Black and White Incarceration 1990-2013;

› e New Dynamics of Mass Incarceration;

› Justice Denied: e Harml and Lasting Eﬀects of Pretrial Detention;

› Gatekeepers: e Role of Police in Ending Mass Incarceration;

› Broken Ground: Why America Keeps Building More Jails and What It Can

Do Instead; and

› the multimedia storytelling project, e Human Toll of Jail.

rough the Safe and Justice Challenge, Vera’s oﬃces in Los Angeles and

New Orleans, and direct partnerships with jurisdictions nationwide, Vera

is providing expert information and technical assistance to support local

eﬀorts to stem the ﬂow of people into jail, including using alternatives to

arrest and prosecution for minor oﬀenses, recalibrating the use of bail,

and addressing ﬁnes and fees that trap people in jail. For more information

about Vera’s work to reduce the use of jails, contact Elizabeth Swavola,

acting project director, at eswavola@vera.org. For more information about

the Safe and Justice Challenge, visit .safeandjusticechallenge.org.

Contents

1 Introduction

10 Factors that inﬂuence plea bargaining

10 Coercive factors

19 Legal characteristics

24 Systemic inequities

31 The criminal law

35 Caseloads

39 Plea bargaining outcomes

40 How much of a bargain is a plea bargain?

44 Are innocent people induced to plead guilty?

48 Conclusion: Limitations and future directions

In the Shadows: A Review of the Research on Plea Bargaining 1

Introduction

he common narrative in popular culture is that criminal justice is

meted out in courtrooms around the country. Facts about a particular

case inevitably emerge from adversarial proceedings in which

prosecutors and defense aorneys go to bale in open court over maers

of fact and law, juries decide whether people are guil or not guil, and

judges determine appropriate punishments.

People have their days in

court, and the public—including victims, if there are any—witness whether

justice is done.

In fact, criminal trials are rare.

Instead, most criminal cases that result

in conviction—97 percent in large urban state courts in 2009, and 90

percent in federal court in 2014—are adjudicated through guil pleas.

these, researchers estimate that more than 90 percent are a result of plea

bargaining—an informal and unregulated process by which prosecutors

and defense counsel negotiate charging and sentencing concessions in

exchange for guil pleas and waivers of constitutionally guaranteed trial

rights.

Indeed, by one estimate, a criminal case is disposed of by plea

bargaining every two seconds during a pical work day in America.

Negotiated deals to resolve criminal cases are so ubiquitous that Justice

By one estimate, a criminal case is

disposed of by plea bargaining every

two seconds during a typical work day

in America.

2 Vera Institute of Justice

Anthony Kennedy of the U.S. Supreme Court stated in 2012 that “criminal

justice today is for the most part a system of pleas, not a system of trials.”

Plea deals—which are entirely within the discretion of a prosecutor to

oﬀer (or accept)—pically include one or more of the following:

› the dismissal of one or more charges, and/or agreement to a

conviction to a lesser oﬀense (known as “charge bargaining”);

› an agreement to a more lenient sentence, which can cover both pe

of sanction—custodial or communi-based—and length (known as

“sentence bargaining”); and

› an agreement to stipulate to a version of events that omits certain

facts that would statutorily expose a person to harsher penalties

(known as “fact bargaining”).

Most criminal cases are resolved by plea bargaining.

Researchers

estimate that

more than

90%

criminal cases

that end in

conviction are

the result of plea

bargaining...

...a low-

visibility,

o-the-record,

and informal

process that

usually occurs

far from open

court.

In the Shadows: A Review of the Research on Plea Bargaining 3

Plea negotiations can be quick and straightforward or long and

complicated—or anywhere in between.

When there are no legal issues in

dispute, or in cases in which it is likely that the prosecution will prevail,

plea negotiations can be superﬁcial—with people taking given deals

without any negotiation or counteroﬀer.

is is especially so in cases of

misdemeanors, infractions, and other lesser oﬀenses for which there are

oen standard dispositions routinely oﬀered and accepted with lile, if

any, actual deliberation. (See “Misdemeanor justice and plea bargains” on

page 16.) Oen, in more complex cases where available evidence is subject

to conﬂicting interpretations, and/or where sentencing stakes are higher,

negotiations can be lengthier—potentially occurring over many weeks and

months.

In a smaller proportion of cases, no actual oﬀer or negotiation

occurs, and people plead guil without any speciﬁc promises or assurances

from the prosecution—called variously taking an “open plea,” taking a

“blind plea,” or “pleading to the sheet.”

In whatever form it takes, plea bargaining remains a low-visibili, oﬀ-

the-record, and informal process that usually occurs in conference rooms

and courtroom hallways—or through private telephone calls or e-mails—

far away from the prying eyes and ears of open court.

Bargains are

usually struck with no witnesses present and made without investigation,

A system of plea bargaining.

4 Vera Institute of Justice

testimony, impartial fact-ﬁnding, or adherence to the required burden

of proof.

Moreover, lile to no documentation exists of the bargaining

process that takes place between initial charge and a person’s formal

admission of guilt in open court, and ﬁnal plea deals that close out cases

are themselves rarely wrien down or otherwise recorded.

As such, plea

deals, and the process that produces them, are largely unreviewable and

subject to lile public scrutiny.

us, despite the high frequency with

which plea deals are used, most people—aside from the usual courtroom

actors—understand neither the mechanics of plea bargaining nor the

reasons so many people decide to plead guil.

Plea bargaining has, however, become the central focus of a growing,

but still small, body of empirical research. In recent years, mounting

concerns about plea bargaining’s role in encouraging the widespread

forfeiture of constitutionally guaranteed trial rights and associated

procedural protections—and its critical role in eling mass incarceration—

has stimulated rther urgency in understanding how the process works.

Indeed, an array of questions regarding its fairness have emerged. Over the

In recent years, mounting concerns about

plea bargaining’s role in encouraging the

widespread forfeiture of constitutionally

guaranteed trial rights and associated

procedural protections—and its critical

role in fueling mass incarceration—

has stimulated further urgency in

understanding how the process works.

In the Shadows: A Review of the Research on Plea Bargaining 5

last few decades, prosecutorial leverage in plea negotiations has increased

exponentially as changes in substantive law have bolstered criminal

penalties and given prosecutors a wider range of choices to use when

ﬁling charges (such as mandatory penalties, sentencing enhancements, and

more serious yet duplicative crimes already well covered by existing law).

But increased exposure to harsher penalties has not been matched with

increased procedural protections for defendants. Prosecutors’ wide powers

in plea bargaining still go largely unchecked, and there are no meaningl

oversight mechanisms or procedural safeguards to protect against unfair

or coercive practices, raising fears about arbitrariness and inequali. Given

this lack of regulation, concern has also grown over the extent to which

innocent people are regularly being induced to plead guil, as well as plea

bargaining’s role in perpetuating racial and ethnic disparities in criminal

case outcomes—for example, plea bargaining practices that send more

Black people to prison or jail than similarly situated white people.

Plea bargaining’s ll impact on the legal system and justice-involved

people remains unknown, but empirical research on this lile understood

yet immensely inﬂuential practice has begun to emerge. In order to provide

an accessible summary of existing research to policymakers and the public,

the Vera Institute of Justice (Vera) examined a body of empirical studies

that has developed around plea bargaining. Although this review is not

exhaustive, it provides a picture not only of the current state of scholarship

on plea bargaining, but also of the gaps in knowledge that must be ﬁlled.

As this report will discuss, studies appear to fall into seven main focus

areas, primarily examining either of two broad questions: (1) which factors

most inﬂuence the plea bargaining decision-making process?; and (2) what

is the impact of plea bargaining on case outcomes? e seven focus areas

covered by this report include the following.

› Coercive factors. A number of studies examine whether the

punishing circumstances of pretrial detention or the threat of

onerous sentences—speciﬁcally, the death penal—play an outsized

role in inducing and expediting guil pleas.

› Legal case characteristics. Another extensive body of work looks

at how legal case characteristics—severi of charge, prior record,

evidentiary factors—inﬂuence the pe of plea oﬀers that are made.

6 Vera Institute of Justice

› Systemic inequities. ere is a growing body of research

that considers the inﬂuence of demographic characteristics in

determining plea outcomes. ese studies primarily look at the

potential existence of conscious or unconscious biases that may

create disadvantage and inequali across race, ethnici, gender,

or age.

› e criminal law. Another body of work looks at the organization,

structure, and content of the criminal code to understand how

the criminal law—including the existence and pe of structured

sentencing schemes—can aﬀect the likelihood and substance of a

plea deal.

› Caseloads. Some studies have tested the common assumption that

the frequent use of plea deals is, at least partly, a nction of heavy

caseloads among prosecutors and a resultant pressure to hasten the

disposal of as many cases as possible. (No research considered the

impact of defenders’ caseloads on plea bargaining.)

› Trial penal. A substantial body of the literature explores the

so-called “trial penal” (or “plea discount”)—that is, the diﬀerence

between a criminal sentence produced by guil plea versus by trial.

Much of this work is narrowly based on the assumption that plea

decisions rely on anticipated trial outcomes, such that the higher

the penal (or larger the discount) the higher the likelihood of a

guil plea.

› Innocence. Finally, given the potential beneﬁts to people of

accepting a plea deal, a small body of research has considered the

extent to which innocent people may be coerced into pleading guil

to avoid receiving a “trial penal” if they fail to prove their cases

at trial.

Whether people are charged with serious crimes or low-level

misdemeanors, whether they are before busy ci courts or slower-paced

rural ones, they will likely resolve their cases by negotiated pleas. Plea

bargaining is the de facto system of justice in America. But despite decades

of scholarship, lile is ultimately known about how it works or why people

plead. is review of contemporary scholarship only oﬀers mixed clues

about the criminal legal system’s primary dispositional process. Although

In the Shadows: A Review of the Research on Plea Bargaining 7

there is evidence that most people receive more favorable sentencing

outcomes through plea bargaining than they would if they had taken their

cases to trial, the exact contours of how such bargains are reached, and

the factors—whether individual, legal, institutional, or demographic—

that ultimately play a key role in inﬂuencing plea outcomes remain both

ambiguous and opaque. What exists instead is a mix of complicated,

nuanced, and sometimes contradictory research ﬁndings.

In order to bring plea bargaining out of the shadows and ensure its

equitable use, more transparency is needed about the process by which

most cases are concluded so that safeguards can be put in place to protect

people from its misuse.

Whether people are charged with serious

crimes or low-level misdemeanors, whether

they are before busy city courts or slower-

paced rural ones, they will likely resolve

their cases by negotiated pleas.

8 Vera Institute of Justice

Law of plea bargaining: An overview

Plea bargaining emerged in the early years of the Republic

as a localized legal innovation that would eventually grow to

become—by the early 20th century—this country’s primary

method of criminal conviction.

But despite plea bargaining’s

long history at the center of American criminal justice, a set of

coherent rules governing its use has never emerged. Indeed,

for much of its history, plea bargaining remained a largely

unregulated and informal form of pretrial negotiation.

Although courts acknowledge plea bargaining as an “essential

component of the administration of justice”—or, more recently,

recognized that “it is the criminal justice system”—courts

have generally taken a hands-o approach in regulating its

use, leaving much discretion with prosecutors and defense

attorneys in shaping plea bargaining processes and outcomes.

This reticence is in part due to the approach that courts have

taken to analogize plea bargains to contracts in private law—

describing them as the same as “any other bargained-for

exchange” between autonomous actors who proceed from

a “mutuality of advantage”—even though prosecutor and

defendant have inherently unequal levels of power, particularly

when the accused is being held in jail pretrial.

Placing

substantive limits on bargaining tactics, according to this line

of thought, could potentially cast a chilling eect that might

foreclose plea negotiations altogether.

In only a few cases

have courts attempted to deﬁne plea bargaining’s contours or

set its outer bounds. Moreover, the few statutes and procedural

rules governing plea bargaining processes exert minimal

control over the direction and process of how plea bargains are

made and are largely silent regarding the actual substance

of plea bargains themselves.

However, in recent years, courts

have begun to ﬁll the gaps in procedural rules with case law

clarifying the limits of prosecutorial discretion.

The constitutional law of plea bargaining

At the heart of plea bargaining law is the 1970 seminal case

of Brady v. United States. Long after plea bargaining had

become the norm in resolving the vast majority of criminal

cases, the Brady court formally recognized plea bargaining

as a constitutional method of criminal adjudication and set

out its minimum requirements.

To be constitutionally valid,

guilty pleas must be both “voluntary” and “knowing, intelligent

acts.”

However, courts have since deﬁned both “voluntary” and

“knowing” in manners that are signiﬁcantly dierent from what

the average person might suppose those terms to mean.

Voluntary. Although the voluntariness requirement supposedly

ensures that people are not misled, tricked, or otherwise

coerced into forgoing their constitutional right to trial by

pleading guilty, courts have still regularly avoided reckoning

with the inherent pressures of “the give-and-take of [the]

plea bargaining” process on people, including the inherently

coercive nature of pretrial detention.

Instead, they have

concluded that “the imposition of dicult choices [is] an

inevitable—and permissible—attribute of any legitimate system

which tolerates and encourages the negotiation of pleas.”

As a result, courts have decided that few actions will likely

amount to “improper pressure that would . . . overbear the

will of some innocent persons” such as to render a guilty plea

involuntary.

Prosecutors have been aorded wide latitude in

using every bargaining chip that is permissible by law when

trying to extract guilty pleas from people—including the threat

of indeﬁnite pretrial detention—so long as the tactics do not

involve illegal fraud (such as threatening to use false testimony)

or outright physical harm, and so long as the bargain is

upheld.

These include promises of leniency if a plea is accepted

and threats of worse punishment, additional charges, or even

the prosecution and punishment of family members if a plea

deal is declined.

Knowing. Although courts have been clear that people must

possess several critical pieces of information in order to validly

plead guilty—the nature of the charges, the rights waived by

pleading guilty, and the sentence that will, or is likely to be,

imposed—the level of actual understanding that is required

is very minimal in practice.

Usually, a superﬁcial judicial

inquiry—the “plea colloquy”—probes whether a person’s

plea bargaining choices are suciently informed; these are

typically highly scripted proceedings that are outlined in state

and federal procedural rules governing the formal entrance of

guilty pleas on the record in open court.

For the guilty plea

to pass constitutional muster, people typically need only to

provide short, often perfunctory, armative responses and a

basic explanation of the oense to which they are admitting.

Courts rarely conduct a deeper inquiry into whether people

fully comprehend the consequences of pleading guilty beyond

what is minimally required by the governing standard set out in

Brady and codiﬁed in the rules of criminal procedure.

There are also other larger structural curbs on what people can

know during the plea bargaining process. For example, people

do not have a right to know the full extent of the evidence, or

lack thereof, against them, because there is no constitutional

right to discovery in criminal cases.

The state is not required

to turn over evidence that may establish a person’s guilt and,

under United States v. Ruiz, it need not share material evidence

favorable to the defendant—a due process right established

under Brady v. Maryland and successor cases that only

attaches if a case goes to trial.

In the Shadows: A Review of the Research on Plea Bargaining 9

Recent developments in plea bargaining law

In recent years, the Supreme Court has become incrementally

more active in regulating the plea bargaining process—

particularly regarding the responsibilities of defense counsel.

This is in part due to an apparent change in approach.

Previously, the court viewed criminal trials as the normative

“touchstone” guiding its decisions around criminal law and

procedure—many of which centered on elaborating procedural

protections regarding the jury trial guarantee and attendant

rights.

But since 2010, in decidedly acknowledging plea

bargaining’s centrality to the criminal legal system, the court

has sought to extend legal protections to better ensure that

people receive competent legal advice and thus eective

bargaining by defense counsel during plea negotiations. For

example, in four cases decided from 2010 to 2017, the court

consistently held that people have a constitutional right to

eective legal counsel during the client counseling phase of the

plea bargaining process—a right that is normally associated

with criminal trials rather than a process that sits wholly apart

from trial proceedings.

In another area of development, courts have begun to

reconsider the scope of appeal rights that are waived when

people plead guilty. Generally, people forfeit a number of

rights simply by choosing to plead instead of going to trial. (For

example, people forfeit rights that would operate during the

trial, such as the right against self-incrimination or to confront

accusers, as well as certain claims that could have been raised

pretrial, such as the racial composition of the grand jury or

the credibility of prosecution witnesses.)

People can also

waive other rights through express waiver provisions that may

accompany a plea oer—although there may be some rights

that are not waivable.

The deﬁnitive scope of which rights are

impliedly waived by a guilty plea, as well as which rights one

can expressly waive, remains an open area of debate. But in

two recent cases, the Supreme Court held that:

› a guilty plea alone—absent an express waiver—does not

bar people from challenging the constitutionality of their

convictions on direct appeal; and

› people who suer ineective assistance of counsel retain

a right to appeal, even if their plea agreements speciﬁcally

waive that right.

In a separate development, echoing civil contract law, the U.S.

Second Circuit Court of Appeals nulliﬁed an appellate waiver in

2018 after ﬁnding that the plea agreement was not supported

by “consideration”—the beneﬁt that each party gets or

expects as the part of an agreement. The defendant had been

sentenced to the statutory maximum and given no beneﬁt of

sentence reductions that were available due to his acceptance

of responsibility and timely guilty plea, eectively giving him

nothing in exchange for waiving his right to appeal.

* Box notes at end of report.

10 Vera Institute of Justice

Factors that inﬂuence plea

bargaining

iven how central plea bargaining is to the nctioning of the

American criminal legal system, understanding how plea deals

are reached is deeply important. As explained in this section,

researchers have investigated the extent to which plea deals are shaped

by coercive factors (such as a person’s pretrial detention status or the

possibili of a death penal sentence), the legal characteristics of a case

(such as the strength of the evidence, or the severi of the charges), and the

diﬀerential treatment of people based on their demographic characteristics

such as race, age, and gender.

Coercive factors

Of great concern to advocates, researchers, and defense counsel is the

inherently uneven playing ﬁeld between accused and prosecutor in

plea bargaining situations—especially given the wide arsenal of tools,

particularly around charging, that prosecutors can use to increase their

Of great concern to advocates,

researchers, and defense counsel is

the inherently uneven playing ﬁeld

between accused and prosecutor in plea

bargaining situations.

In the Shadows: A Review of the Research on Plea Bargaining 11

leverage in negotiations. Researchers have looked at how two coercive

factors in particular inﬂuence plea bargaining outcomes: pretrial detention

and the potential for a death sentence. ese studies have concluded that

these factors likely play a signiﬁcant role in inducing guil pleas so that

people can obtain their liber—or even sustain their life.

Pretrial detention

e vast majori of people in local jails are detained pretrial, meaning that

they have not been convicted of any crime, are legally presumed innocent,

and are awaiting resolution of their criminal cases behind bars—most

oen because they cannot pay the bail set in their cases.

Pretrial detention

status has far-reaching consequences for justice-involved people. Concern

about this population—and the potential negative impacts of detention on

their criminal justice outcomes—has spawned a growing body of research

that has established a strong correlation between pretrial detention and an

increased likelihood of conviction, longer custodial sentences, and ture

system involvement.

While it has long been assumed that the pressure

and isolating circumstances of incarceration induce people to plead guil

more readily during the pretrial phase—potentially explaining why people

in pretrial detention are more likely to be convicted than those who are

released—researchers have aempted only recently to speciﬁcally examine

the inﬂuence of pretrial detention on a person’s plea bargaining behavior.

Using a varie of diﬀerent methods, this small body of scholarship has

established a strong association between pretrial detention and pleading

guil. For example, in a 2012 study examining 634 criminal cases in New

Jersey courts, researchers found that people who were detained pretrial

reached faster case dispositions, usually during the pre-indictment phase,

than people who were released, primarily because, as an interviewee

described it, “defendants plead guil to get out of jail . . . get time served

or to get it over with.”

In a 2018 study looking at nearly 76,000 arrests in

Delaware, researchers similarly uncovered that pretrial detention increased

a person’s likelihood of pleading guil by 46 percent—although there were

diﬀerences depending on the person’s race.

Similar to previous research,

the study found that Black people were 10 percent less likely than white

people to enter into guil pleas.

To explain this ﬁnding, researchers

12 Vera Institute of Justice

have postulated that plea bargaining may be less common among Black

people because they may receive less favorable guil plea agreements

from prosecutors than do white people. Another reason may be that Black

people may be more distrustl of a legal system that disproportionately

and unfairly impacts them and, thus, less likely to strike a bargain.

Both the New Jersey and Delaware studies employed regression

analyses—a statistical method of measuring the relationship between

multiple variables—with only a limited set of controls based on data that

was available (for example, current charge aregated into broader oﬀense

categories, criminal history, and demographic information). is means

that they did not account for a number of unobservable confounding

factors that may have also inﬂuenced people’s decisions to plead guil

(such as strength of evidence, quali of defense, individual cognitive

biases, or wealth). To correct for this potential bias and beer estimate

the causal eﬀect of pretrial detention on a person’s propensi to plead

guil, four recent studies conducted natural experiments—non-controlled

observational studies that exploit random assignment that occurs in

“nature” and which provide social scientists with a stronger inferential tool

to potentially improve the quali of their empirical inferences, particularly

when trying to infer causation.

ese studies exploited the random

or rotating assignment of bail judges and either variations in judicial

punitiveness in bail decisions or variations in people’s access to money bail

by day of the week.

Pretrial detention increases a person's likelihood of pleading

guilty by 46 percent.

plea

In the Shadows: A Review of the Research on Plea Bargaining 13

Consistent with other empirical research, all four studies conﬁrmed

that a person’s odds of conviction via guil plea increased when they

were held in pretrial detention—oen resulting in worse criminal justice

outcomes. For example, one study that looked at 331,971 criminal cases in

Philadelphia from 2006 to 2013 found that pretrial detention resulted in

a 4.7 percentage point increase in the likelihood of pleading guil among

people who probably would otherwise have been acquied, diverted,

or had their charges dropped.

Similarly, a second study that examined

nearly one million criminal cases over ﬁve years in New York Ci found

that detained people charged with felonies were 10 percent—and detained

people charged with misdemeanors were just over 7 percent—more likely

to plead guil compared to similarly-situated people who were released.

Again, not only did pretrial detention increase the odds of pleading guil,

but the study also found evidence that plea oﬀers were less favorable for

people detained on felony charges in particular.

Plea deals for this group

were 10 percent less likely to include a charge reduction. e study also

found that pretrial detention may induce people to plead guil in cases

that would have been dismissed if those people had been released: 34

percent of released cases were dismissed, compared to only 19 percent of

cases among those detained.

Four studies conﬁrmed that a person’s

odds of conviction via guilty plea

increased when they were held in pretrial

detention—often resulting in worse

criminal justice outcomes.

14 Vera Institute of Justice

e two other contemporaneous studies came to similar conclusions.

In a Harris Coun (Houston), Texas, study focusing only on misdemeanor

cases resolved from 2008 to 2013, researchers found that people who were

detained pretrial pleaded guil at a 25 percent higher rate than people

who were released and that data suested that 17 percent of the detained

people would likely not have been convicted at all had they been released

pretrial.

Conversely, the fourth study found that pretrial release reduced

a person’s odds of pleading guil and improved plea outcomes. Aer

examining 420,000 felony and misdemeanor cases from Philadelphia and

Miami-Dade counties covering an eight-year study period, researchers

found that people who were released pretrial were nearly 11 percent

less likely to plead guil—although these estimates varied across race.

Released white people were nearly 20 percentage points less likely to plead

guil, while released Black people were only 12.2 percentage points less

likely to plead guil.

Of those who took a plea oﬀer in the pretrial phase,

outcomes were still more favorable for people who had been released. ey

were more likely to be convicted of fewer oﬀenses or convicted of a lesser

charge, suesting that release can improve a person’s bargaining power in

plea negotiations.

In explaining their results, the researchers speculated that release

may enable people to engage in certain “prophylactic measures”—such as

preparing a more robust defense, engaging in treatment (for mental illness,

substance use disorder, or other behavioral health issues), or providing

restitution—that can help lead to charges being dismissed or encourage

more lenient treatment by prosecutors.

Some of these studies also found that the observed impacts of pretrial

detention—or release—on a person’s plea bargaining propensi increased

in magnitude for two particular pes of people: those with no or limited

criminal justice histories and those charged with lower-level oﬀenses.

For example, researchers in Harris Coun found that pretrial detention

more than doubled the likelihood of conviction for people charged with

ﬁrst-time misdemeanors—the vast majori of whom had pleaded guil.

Similarly, the Philadelphia study found that pretrial detention among

people charged with misdemeanors led to statistically signiﬁcant increases

in all observed negative outcomes, including conviction and guil pleas,

larger than those observed with felonies.

e Philadelphia/Miami study

also found that released people charged with misdemeanors were nearly

In the Shadows: A Review of the Research on Plea Bargaining 15

19 percent less likely to plead guil, while released people charged with

felonies were only 1.2 percent less likely to plead guil.

e study also

found that released people who did not have a recent prior oﬀense were

more than 14 percent less likely to plead guil than similarly situated

people who were detained pretrial.

Put together, these ﬁndings are consistent with the theory that detained

people, who may have limited knowledge of the criminal legal system and/

or who are being detained for the ﬁrst time, have strong incentives to cut

a quick deal in order to resolve their cases as soon as possible—especially

when it involves lower-stakes pey oﬀenses. As was noted earlier, this

is because a plea deal is oen less “painl” in the short term because it

can get people out of jail if they are detained, purchase certain in what

can be a long and unpredictable criminal legal process, or sidestep the

possibili of harsher punishment that could result from trial. In line with

these observations about the long and painl process of justice system

involvement and detention—even for low-level oﬀenses—the New York

Ci study found that people detained on misdemeanor charges pleaded

guil not only more frequently but also faster than people detained on

felony charges.

For incarcerated people charged with misdemeanors

whose cases continued beyond their ﬁrst court appearance, the median

time between arraignment and case disposition was less than three weeks

at every predicted sentence length, but almost 50 days for people charged

with felonies in the shortest predicted sentence length category, growing

larger for groups with longer predicted sentence lengths.

Detained people have strong incentives to

cut a quick deal in order to resolve their

cases as soon as possible.

16 Vera Institute of Justice

Misdemeanor justice and plea bargains

More serious crimes—rape, murder, aggravated assault, etc.—

may be more likely to capture media and public attention, but

the American criminal legal system is, in fact, dominated by the

enforcement of a vast array of minor oenses located both in a

state’s criminal code and in local ordinances.

This enforcement

is both wide and deep, and it is the point of contact through

which most Americans experience the criminal legal system.

It touches everything from low-level crimes (such as disorderly

conduct, simple assault, petty theft, low-level fraud, illegal

trespass, prostitution, driving under the inﬂuence or on a

suspended license, and drug possession) to commonplace

behaviors deemed undesirable or “anti-social” that may either

be designated “civil” or “criminal” infractions (such as public

urination, public intoxication, unlicensed vending, littering,

grati, panhandling, jaywalking, and loitering).

Misdemeanors

can involve small harms, or no harms at all, and many are often

symptoms of larger social problems with which communities

grapple—including substance use disorders, mental illness,

domestic violence, and poverty.

Given the breadth of issues covered by minor oenses, it is

unsurprising that the national misdemeanor caseload volume

is large. According to a recent estimate, there were 13.2 million

misdemeanor cases ﬁled in the United States in 2016—or 4,261

misdemeanor cases ﬁled per 100,000 people; accounting for

76 percent of statewide criminal caseloads in 31 states.

It also

appears that the proportion of such cases among all criminal

prosecutions has remained relatively stable over the last

several decades.

Thus, the majority of plea bargains in America

are products of the country’s diverse and sprawling network

of lower criminal courts where minor charges are typically

adjudicated.

Despite the preponderance of minor charges in the criminal

legal system, a substantial proportion of empirical research

on plea bargaining either focuses on people charged (and

convicted) of felony oenses or comingles felonies and

misdemeanors, assuming that plea bargains operate with

similar goals—or under similar conditions and pressures—

across all types of oenses. But a small body of research—

mainly descriptive in nature—that has explored the real-life

operation of the lower courts suggests that guilty pleas

in misdemeanor and other petty cases are products of a

process that may be qualitatively dierent from that of most

serious oenses in ways that prevent aggregation in analysis,

particularly when trying to extend a generalizable explanation

of how plea bargains work.

This body of work conﬁrms, for example, that in the world of

misdemeanors, infractions, and other lesser oenses, there are

often standard “deals”—baseline oers set to speciﬁc oense

types—that are routinely proposed and accepted that involve

little actual bargaining between parties.

For the accused,

these oers are hard-to-refuse dispositional deals that can

eectively end cases quickly, often as early as their ﬁrst court

appearance.

These include:

› case dismissal;

› unconditional or conditional discharge;

› a conviction of a civil (rather than criminal) infraction;

› a sentence to time served (accounting for any time spent in

custody leading up to the court appearance); or

› a community-based sentence of minimal length.

Thus, contrary to the notion that prosecutors and defendants/

defense counsel “bargain toward settlement in the shadow

of expected trial outcomes,” in these cases, initial charging

decisions—which are in the prosecutor’s discretion—determine

the type of plea deal that settles a particular matter under

established court practices, prosecutor oce policies, or

defense bar customs around similar cases in a particular

jurisdiction.

Researchers have proposed two main explanations of why

cases related to petty oenses are resolved quickly:

› Process costs. Prosecutors, defense counsel, defendants,

and judges all put a premium on speed because of the

enormous scale of misdemeanor dockets across the country

and the high process costs that misdemeanor trials present

to all parties when compared to the low-stakes nature of the

oenses.

With a criminal legal process that is both slow and

unpredictable, costs for the accused include uncertainty of

outcome and a protracted pretrial and trial process that can

span many days, weeks, or even months—one that may be

experienced in part, or wholly, behind bars and away from

their families, jobs, and community responsibilities.

Thus,

paying the price of a guilty plea may be, counterintuitively,

the most “rational” option whether a person is guilty or not.

› Managerial justice. Another body of scholarly work

suggests that prosecutors are not actually trying to

maximize punishment or even secure a conviction when

oering standard plea deals to quickly resolve misdemeanor

and other petty cases.

Instead, prosecutors use various

adjudicatory tools that avoid formal punishment but enable

them to document a person’s criminal justice encounters

and track behavior over time (such as conditional discharge

or an adjournment in contemplation of a dismissal) so that

law enforcement agencies (such as police, prosecutors, and

courts) have a record to use in calibrating future responses.

In the Shadows: A Review of the Research on Plea Bargaining 17

The impact of death penalty sentences

Fears about the coercive nature of plea bargaining are perhaps most

pronounced in cases in which prosecutors have the option to pursue a

death penal conviction. Critics note that when taking cases to trial that

may result in a sentence to death, people may be more likely to accept a

plea deal that they would otherwise have rejected—due either to the harsh

terms of the sentence and/or their factual innocence.

e threat of the

death penal is used as leverage by prosecutors in bargaining, and both

prosecutors and defense lawyers agree that the specter of a death penal

puts prosecutors in a uniquely strong position.

Although analysis of

actual cases is hampered by a relatively small sample to draw from, the

quantitative studies suest that the option to pursue a death penal has a

signiﬁcant eﬀect on plea bargaining.

One such study conducted a natural experiment capitalizing on the

1995 reinstatement of the death penal in New York State to analyze

changes in case outcomes before and aer the law change.

e study

found that people charged in murder cases were 25 percent more likely

to plead guil to their charges following the law change, regardless

of whether the prosecutor had explicitly ﬁled a notice to pursue the

death penal in the case.

e reintroduction of the death penal, the

researchers concluded, made people less likely to be oﬀered a charge

reduction (pically the more advantageous pe of plea oﬀer) and more

likely to take a sentence deal—suesting that prosecutors did indeed

have—and use—greater power in plea bargaining.

A study from Georgia

Prosecutors and defense lawyers agree

that the specter of a death penalty puts

prosecutors in a uniquely strong position.

18 Vera Institute of Justice

produced similar results.

e study looked at eight years of murder

cases from across the state that met the criteria necessary to be tried as

capital cases. e researcher compared the outcomes of cases in which

prosecutors pursued the death penal with those in which they did not,

using sophisticated statistical techniques to control for a varie of case

characteristics. ey found that, all else being equal, people charged in

murder cases were approximately 20 to 25 percentage points more likely

to plead guil when faced with the death penal.

Put another way, the

study suests that when prosecutors actively pursue a death penal,

people in an additional two out of every 10 cases are deterred from going

to trial.

ese ﬁndings have been supported nationally. A comparison of 33

counties—some with the death penal, some without—found that, when

the death penal was available, 19 percent of ﬁrst degree murder cases

were resolved with a guil plea leading to a prison sentence of more than

20 years; in counties without the death penal, this was true for only 5

percent of cases.

is large diﬀerence in outcomes remained statistically

signiﬁcant even when controlling for other case characteristics. In pleading

guil to a life sentence, people are relinquishing their right to appeal, the

chance of an acquial, and the possibili of a shorter sentence; avoiding

the death penal is used as a “substantial incentive” to encourage people to

make this otherwise unappealing decision.

People are 25 percentage points more likely to plead guilty when

faced with the death penalty.

death

sentence

plea

In the Shadows: A Review of the Research on Plea Bargaining 19

Legal characteristics

Researchers have looked at how the application of plea bargaining, and the

potential associated discount in punishment, might vary in relation to the

legal characteristics of a case. Of these, the strength of the evidence, the

severi of the charges, and the accused’s criminal history have garnered

most aention.

Strength of evidence

e strength of the evidence against a person charged with a crime is

perhaps the most salient factor prosecutors say they consider when

deciding whether a plea deal should be extended. Interviews and

surveys conducted with prosecutors have aempted to shed light on the

relationship between the strength of the evidence in a case and the plea

bargaining process, oen testing the general assumption that prosecutors

will oﬀer more lenient deals when the evidence is weaker or harsher deals

when the evidence is stronger. In a survey of 166 prosecutors in three

southern states, 82 percent of respondents indicated that they would

reduce the harshness of the plea if the evidence was weak, and 38 percent

indicated that they would increase the harshness of the plea if the evidence

was strong.

However, a diﬀerent survey study asked nearly 400 prosecutors to

respond to hypothetical legal cases and found that the presence or absence

of evidence had very lile eﬀect on prosecutors’ likelihood to recommend

a plea deal, with plea oﬀers being extended in nearly all cases.

Again, the

presence of evidence decreased the leniency of oﬀers recommended by

prosecutors, but only by a small amount.

e authors noted, however,

that ﬁndings from hypothetical cases such as these—with the absence of

real-life court and caseload pressures—may not easily generalize to actual

practice.

As such, a number of studies have aempted to examine the

relationship between plea bargaining and evidence strength in real cases.

An analysis of cases relating to person and proper oﬀenses

(misdemeanor and felony) ﬁled in one jurisdiction across multiple years

sought to determine the relationship between evidence strength and

plea bargaining using quantitative analysis.

As a proxy for evidence

20 Vera Institute of Justice

strength, the research considered the total number of pieces of evidence

listed in a case and the number of pieces of evidence per intake charge.

While controlling for a number of other case and defendant variables, the

researchers found a signiﬁcant relationship between strength of evidence

and plea oﬀers—but not on all metrics. e strength of the evidence was

unrelated to both the seriousness of the top plea oﬀer charge made by

prosecutors and the length of sentence aached to it. Stronger evidence

was, however, associated with increases in the number of charges included

in the plea oﬀer, the total sentence possible if convicted for all charges,

and the likelihood that the prosecutor would recommend incarceration.

A 2015 study of felony drug cases in New York Ci similarly assessed

the relationship between evidence and plea bargain outcomes while

controlling for other legal and extralegal factors.

e researchers found

that when factors indicating strong evidence were present, the accused

were less likely to receive reduced plea charge oﬀers and more likely to

receive custodial, rather than noncustodial, sentence recommendations.

However, while signiﬁcant, the size of these relationships was smaller than

the researchers had expected.

Confession evidence is considered extremely persuasive at trial, which

could make it an important factor during plea bargaining.

An analysis

of more than 500 felony cases from two New York counties showed that

people who confessed (either lly or partially) overwhelmingly pled guil

(more than 97 percent), but received lower plea discounts than others

(people not interviewed by police and people who denied guilt).

As a

rther analysis, researchers in this study coded the apparent strength

of the evidence in each case. Contrary to expectations—and to ﬁndings

from earlier studies—cases with stronger evidence received greater plea

discounts than cases with weaker evidence.

e researchers proposed

that, because in the coun they studied the prosecutor’s oﬃce had a

known policy stating that no oﬀer would be beer than the initial oﬀer,

people with strong cases against them who knew about the policy were

more likely to accept the ﬁrst plea deal oﬀered.

Researching the relationship between plea bargaining and the strength

of the evidence in real cases presents speciﬁc and signiﬁcant challenges:

the strength of evidence is unlikely to be recorded in ways that are

amenable to large scale analysis and is diﬃcult to quanti in a meaningl

way.

ere are nuances to evidence that are diﬃcult to assess and

In the Shadows: A Review of the Research on Plea Bargaining 21

combine in the aregate. For example, eyewitness testimony may vary

in its persuasiveness if it comes from an undercover police oﬃcer or a

witness to the incident, and witnesses may be aﬀorded diﬀerent levels of

credibili based on their demographics and relationship to the accused.

Furthermore, the strength of a speciﬁc piece of evidence is not necessarily

static and may decay over time as witnesses fall away or forget important

details or an investigation throws doubt on what ﬁrst appeared to be

important evidence.

Charge severity

e seriousness of the current oﬀense is a key consideration when

contemplating plea agreements. Research suests that the likelihood of

a person pleading guil and the leniency of the plea deal oﬀered to them

by prosecutors may both be inﬂuenced by the severi of the charges

the person faces.

A study of 200 cases of drunk driving in a southern

California coun in 1993, for example, analyzed the relationship between

multiple legal and extralegal case characteristics and case outcomes and

found that charge severi was the largest predictor of pleas; people with

more severe charges were less likely to plead guil.

e researchers

suested that when the stakes of conviction are higher, there is greater

incentive to take one’s chances in court.

Another study of 464 felony

The relationship between charge severity and plea likelihood.

increase in

charge severity

decreased likelihood

of pleading guilty

22 Vera Institute of Justice

cases from Virginia in the late 1970s, which also controlled for many

demographic and case characteristics, similarly found that when charged

with oﬀenses that could result in longer sentences, people became less

likely to plead guil.

Studies of whether a deal is reached, however, say nothing of the

leniency (or lack thereof) in the plea oﬀers made by prosecutors. A large-

scale study of federal cases that resulted in conviction in 2001 (excluding

cases resolved through trial) found that more serious charges were

associated with a greater likelihood of charge reductions.

e researchers

theorized that this may be, at least in part, because more serious crimes—

and those with more ﬁling charges—allow more opportunities for

reduction than do lower-level crimes.

Criminal history

People’s conviction histories are oen used to guide sentencing decisions

and are sometimes explicitly included in prosecutors’ plea deal guidelines.

As such, it would make sense that criminal histories might exert a

signiﬁcant inﬂuence on the plea bargaining process. Although researchers

frequently control for criminal history when examining the relationship of

other legal and extralegal factors with plea bargaining outcomes, it is less

frequently the subject of direct investigation itself.

Research conducted in coun criminal courts in Pennsylvania (using

data from 1997 through 2000) and, in a separate study, federal criminal

courts (using data from 2000 through 2002) found that people with more

substantial criminal histories are penalized less for going to trial than those

without.

at is to say, the diﬀerence between sentences resulting from

a jury trial compared to those imposed through plea bargaining becomes

smaller as people’s criminal histories increase. Researchers oﬀer several

plausible explanations for this relationship. First, a jury trial may provide

an opportuni for defense counsel to argue that a person’s criminal history

is less serious or meaningl than it may seem on paper and to present

evidence of the accused’s good character.

Second, it is possible that, for

those with longer criminal histories, prosecutors oﬀer less lenient plea

deals or encourage people to plead to the charges set at arraignment.

e sentence they receive as a result is unlikely to diﬀer markedly from

In the Shadows: A Review of the Research on Plea Bargaining 23

a sentence they might receive at trial. Although the ﬁrst explanation is

diﬃcult to quanti and study, research has lent some support to the

second—although, as elsewhere, the ﬁndings have been mixed.

A large study of people who were sentenced in federal cases in 2001

found that criminal history had no inﬂuence on the reduction of charges

during plea bargaining.

However, other studies—which similarly

analyze the relationship between criminal history and plea bargaining

(while controlling for other important variables)—have found that more

substantial criminal histories are associated with harsher plea deals,

at least for some charge pes. For example, a study of nearly 160,000

misdemeanor cases prosecuted by the District Aorney of New York

Coun found that, holding other case and defendant characteristics

constant, people with longer criminal histories were more likely than

people with shorter histories to receive a plea oﬀer that included

incarceration.

Similarly, statistical analysis conducted in one unnamed

jurisdiction found that—for all charge pes studied (proper, personal,

and drug)—more serious criminal histories were associated with an

When evaluating the association between

people’s criminal histories and the

likelihood of receiving a plea deal—or the

harshness or leniency of that deal—it is

important to keep in mind that criminal

histories themselves are subject to deep

rooted racial biases within the criminal

legal system.

24 Vera Institute of Justice

increased likelihood that the plea deal would include incarceration and

with smaller reductions in the total possible sentence a person might

serve.

However, more serious criminal histories were only found to be

associated with smaller reductions in the seriousness of the top charges

and smaller reductions in the total number of charges for drug oﬀenses.

Criminal histories are, of course, not a neutral factor, but rather one

rooted in the nation’s historic and continuing systemic racism. When

evaluating the association between people’s criminal histories and the

likelihood of receiving a plea deal—or the harshness or leniency of that

deal—it is important to keep in mind that criminal histories themselves

are not created with objective neutrali and are subject to deep rooted

racial biases within the criminal legal system.

Indeed, research conducted

in New York Ci found that people with longer criminal histories were

more likely to have their cases dismissed, the result of a dynamic in

which police will more readily arrest people—especially Black and Latinx

people—whether or not they have suﬃcient evidence for a prosecutor to

ﬁle charges.

Systemic inequities

As described below, the complete lack of transparency and scrutiny

inherent in America’s reliance on plea bargaining—combined with the

U.S. justice system’s long history of prejudicial operations—leaves the

administration of plea deals open to bias. Relatedly, people may have

widely diverging levels of trust in the system, which inﬂuence their

willingness to accept a plea deal or face trial. As such, researchers have

investigated the degree to which plea oﬀers are aﬀected by demographic

factors, notably the race, gender, and age of the accused.

Racism and race inequity

Consistent with the experiences of Black and Latinx people throughout

the criminal legal system, several studies have found that people of color

are oen treated less favorably than white people during the plea bargain

process. For example, two studies of data from the New York Coun

District Aorney’s oﬃce found that Black people were signiﬁcantly less

Challenges in researching plea bargaining

Guilty pleas constitute the main mechanism for case resolution

in American criminal courts. Researching the role and

administration of plea bargaining in this process has, however,

proved remarkably dicult.

The process of plea bargaining is rarely recorded in any formal

documentation and, when it is, these records are often not

made available to researchers.

Simply asking prosecutors

about the plea bargaining process undoubtedly has some

value, but researchers have found that there can be substantial

dierences between lawyers’ perceptions of their work, their

values, and their actual behavior in plea negotiations.

importantly, asking prosecutors about plea bargaining ignores

the voice of the person most aected by the bargain, and few

studies focus on the perspective of the person pleading guilty.

With scant records and limited trust in the validity of self-

reported behavior, researchers have attempted to infer

details of the plea bargaining process by looking at case

outcomes. Although there are many variations in methodology,

researchers commonly seek evidence of charge bargaining

by comparing the initial charges a person faced with those

to which they ultimately pled guilty. Evidence of sentence

bargaining is often inferred by comparing the punishments

given following a plea with those given following a trial

for the same charges.

There are limitations to both approaches.

Inferring that charge bargaining has occurred on the basis

of a reduction in charges (either number or seriousness) is

problematic because charges may be reduced prior to case

disposition for a number of reasons not necessarily related

to plea bargaining.

For example, charges could be reduced

to compensate for mistakes or biases that led to initial

overcharging or in light of new evidence collected during an

investigation and not because of any plea deal.

Quantifying

the presence and impact of sentence bargaining is also

problematic. Cases that go to trial dier in signiﬁcant, hard-to-

measure ways from cases that result in a plea, and researchers

are inconsistent in how they actually measure sentences—

disagreeing on whether to include noncustodial punishments

as “zero-length” sentences, sometimes counting acquittals in a

similar way, and diering in whether to account for time served

in pretrial detention.

As prosecutors may employ dierent bargaining strategies in

dierent circumstances, it is preferable that research studies

attempt to measure both charge bargaining and sentence

bargaining, but this does not always happen.

Furthermore,

researchers have also noted that, at least in the federal

court system, there is an additional common form of plea

bargaining in which the prosecutor and defense may reach a

deal over the key “sentencing facts” of a case—agreeing to

omit certain facts about a case that would statutorily expose

a person to harsher sentences. This process is perhaps even

harder to record, identify, or measure than charge or sentence

bargaining, and it has received even less research attention as

a result.

Perhaps the most confounding aspect of plea bargaining, next

to the lack of available records, is its granularity: the type of

plea bargain favored (charge, sentence, or fact), the frequency

with which it occurs, and the degree of impact it has on

sentencing outcomes can all vary by the speciﬁc jurisdiction,

prosecutor, and charges associated with a case.

Studies

that aggregate data across charge types, courts, or type of

plea bargaining may, therefore, dilute any evidence of plea

bargaining and its impact, hiding any important associations;

conversely, studies that focus on speciﬁc charges, courts, or

metrics are likely to be limited in their generalizability.

In the Shadows: A Review of the Research on Plea Bargaining 25