Content uploaded by Allison D Redlich
Author content
All content in this area was uploaded by Allison D Redlich on Sep 07, 2019
Content may be subject to copyright.
DOI: 10.1111/1745-9133.12463
RESEARCH ARTICLE
PROTECTING THE INNOCENT
The criminal costs of wrongful convictions
Can we reduce crime by protecting the innocent?
Robert J. Norris1Jennifer N. Weintraub2James R. Acker2
Allison D. Redlich1Catherine L. Bonventre3
1George Mason University
2University at Albany
3Guilford College
Correspondence
Robert J. Norris, Department of Criminology,
Law,and Society, George Mason University,
4400 University Drive, 4F4, Enterprise Hall,
Room 354, Fairfax, VA 22030.
Email: rnorris4@gmu.edu
Research Summary: In this article, we examine crimi-
nal offending by true perpetrators after innocent people are
arrested and convicted for their crimes. After investigating
a set of cases in which DNA was used to exonerate the inno-
cent and to identify the guilty party, we identified 109 true
perpetrators, 102 of whom committed additional crimes.
We found a total of 337 additional offenses committed by
the true perpetrators, including 43 homicide-related and 94
sex offenses. By extrapolating from our findings, we esti-
mate that the wrong-person wrongful convictions that occur
annually may lead to more than 41,000 additional crimes.
Policy Implications: Our findings indicate that one
consequence of wrongful convictions, allowing the true
perpetrators of crimes to remain at liberty and commit
new crimes that imperil prospective victims, represents an
important threat to public safety and thereby dramatically
compounds the harms caused to innocents. We stress
the importance of framing wrongful conviction issues
to capture these important crime control concerns and,
thus, to help galvanize public opinion and promote policy
reforms that will mutually benefit adherents of both crime
control and due process perspectives.
KEYWORDS
crime control, due process, exoneration, framing, innocence, miscarriage
of justice, true perpetrator, wrongful conviction, wrongful liberty
Criminology & Public Policy. 2019;1–22. wileyonlinelibrary.com/journal/capp © 2019 American Society of Criminology 1
2NORRIS ET AL.
Wrongful convictions represent failures of the justice system. Since 1989, 2,364 people are known to
have been exonerated throughout the United States (National Registry of Exonerations, n.d.).1Most
exonerations (1,502 or 63.6%) concern “wrong-person” cases, in which innocent individuals were con-
victed of crimes committed by someone else. Such miscarriages of justice stand in contrast to “no-
crime” cases, in which an innocent person was convicted although a crime was never committed.2
Regardless of case specifics, wrongful convictions generate a widespread “circle of harm” (Thomp-
son & Baumgartner, 2018). The harms suffered by the innocents and their loved ones are profound
(e.g., Westervelt & Cook, 2012), yet they extend as well to the communities to which they return,
to legal actors involved in the wrongful convictions, to the original crime victims and their social
networks,3and potentially to diminished public confidence in the administration of justice. And
even more of consequence is at stake because “wrong-person” wrongful convictions leave the true
offenders free to commit additional crimes and claim new victims. Preventing and correcting wrongful
convictions thus entails “a double imperative – a justice imperative and a public safety imperative”
(Findley, 2002, p. 337).
That wrongful convictions threaten public safety by allowing actual offenders to remain at liberty
is self-evident and has been recognized for decades (e.g., Givelber, 1997; Huff, Rattner, & Sagarin,
1986; Penzell, 2007; West & Meterko, 2016). A scant amount of research, however, has been devoted
to uncovering the breadth and nature of crimes committed by the true perpetrators who remain at large
while innocent persons have erroneously been convicted in their stead. Such information is essential to
gain a fuller understanding of the dangers associated with wrongful convictions and the social harms
they inflict. Also, a more complete and nuanced discussion among scholars, practitioners, and policy
makers will be possible about the urgency of guarding against wrongful convictions by enacting justice
system reforms (Acker, 2013; Baumgartner, Grigg, Ramirez, & Lucy, 2018).
In this study, we examine the number and types of crimes committed by the true perpetrators after
the arrest and wrongful conviction of another individual. With a set of cases from the Innocence Project
in which DNA testing was used not only to exonerate the innocent but also to identify the guilty party,
we seek to understand the extent to which wrongful convictions contribute to criminal activity and
threaten public safety. We additionally identify the factors that contributed to the erroneous convic-
tions in the cases, which ultimately enabled the true perpetrators to evade justice and commit addi-
tional offenses. We conclude with a discussion of implications, including why wrongful convictions
represent a threat to public safety and trigger more than due process concerns. We thus consider the
importance of framing in helping shape public attitudes, how public opinion influences policymaking,
and why evidence-based practices designed to mitigate wrongful convictions should also be construed
as bolstering efforts to reduce crime.
1CRIMES OF “WRONGFUL LIBERTY”
Although their public safety implications are clear, “wrongful convictions have rarely been conceptual-
ized as a factor contributing to violent crimes in America” (Forst & Huff, 2018, p. 438). Concrete data
about the crimes committed by true perpetrators in wrongful conviction cases are lacking, but Forst
and Huff (2018) in their thought experiment suggested that the criminal consequences of justice system
errors are significant. Based on their review of 2014 Federal Bureau of Investigation (FBI) and Bureau
of Justice Statistics (BJS) data concerning arrests for violent crimes and felony conviction rates, they
estimated that more than 339,000 convictions for violent crimes occurred that year. After assuming a
wrongful conviction rate of 1% to 3%, which is in line with many empirical estimates (e.g., Loeffler,
Hyatt, & Ridgeway, 2018; Ramsey & Frank, 2007; Zalman, Smith, & Kiger, 2008), and that each true
NORRIS ET AL.3
perpetrator in these cases committed only one additional crime, Forst and Huff (2018, p. 442) esti-
mated that their collective offending would contribute approximately 3,000–10,000 additional crimes
per year. (The estimated number of crimes would presumably be reduced to approximately 1,900 to
6,400 if 36% of wrongful convictions are “no-crime” cases, as discussed earlier, and thus lack a true
perpetrator.)
Although their results are provocative, this exercise by Forst and Huff (2018) is primarily hypothet-
ical, and existing research findings indicate that the estimate that true offenders in cases of wrongful
conviction commit only one additional crime is likely too conservative. Acker (2013) provided a series
of narratives of this “flipside injustice” (p. 1629) of wrongful convictions. Even though it was not a
representative sample, powerful accounts of crimes committed by 20 true perpetrators across multiple
jurisdictions were offered. Some offenders committed as many as 20 additional offenses, including
sexual assaults, murder, and other crimes of violence, demonstrating that wrongful convictions have
profound public safety implications. For example, in September 1990, Justin Albert Johnson raped
and murdered a 3-year-old girl in Brooksville, Mississippi. Although Johnson was a suspect—he was
known to frequent the area and had a history of sexual assault—evidence implicating him in the child’s
rape and murder was overlooked. After an erroneous eyewitness identification and use of flawed foren-
sic techniques, Levon Brooks was wrongly convicted of the rape-murder in 1992 and sentenced to life
imprisonment without parole. Around the time of Brooks’s trial, another 3-year-old girl was raped
and murdered in the same general area. Again, Johnson was a suspect, but investigators erroneously
pursued Kennedy Brewer, who was wrongly convicted and sentenced to death in 1995. In 2007, DNA
evidence from the earlier crimes was matched to Johnson, who subsequently pled guilty to both rape-
murders, and Brooks and Brewer were both exonerated in 2008. The facts of this case highlight the
extensive harms that can ensue when justice goes awry. During his period of liberty after a rape and
murder, Johnson not only committed another rape-murder of a child that led to a second wrongful
conviction, but he also committed an additional sexual assault and four drug crimes (see Acker, 2013,
pp. 1665–1669; Balko & Carrington, 2018).
The cases described by Acker (2013) represent only a small sample of the harms caused by known
true perpetrators after wrongful convictions. Researchers from the Innocence Project reported that the
actual offender was identified in nearly half (159, or 48.9%) of the first 325 DNA-based exoneration
cases. Based on publicly available information, West and Meterko (2016) explained that the known
true perpetrators in these cases were responsible for at least 142 additional violent crimes, a total that
in their estimation represented “just a fraction of all subsequent criminal activity” (p. 731, emphasis
added).
To our knowledge, only two other studies to date have attempted to track the crimes of true perpe-
trators systematically (see Table 1 for a summary). The first, the product of a joint investigation by
the Better Government Association and Northwestern University Law School’s Center on Wrongful
Convictions, examined 85 Illinois exonerations from 1989 to 2010. Although investigators could not
identify the true perpetrators in all 85 cases—the actual offenders in 35 murders, 11 rapes, and 2 rape-
murders remained unknown—those they did find committed at least 97 additional felonies, including
14 murders and 11 sexual assaults (Conroy & Warden, 2011).
In the second study, Baumgartner and colleagues (2018) examined known wrongful conviction cases
in North Carolina, attempting to systematically track crimes committed by the true perpetrators dur-
ing the period of “wrongful liberty” between the original crime (for which someone else was wrongly
convicted) and their later arrest. They located 36 exoneration cases, 9 of which involved the identifica-
tion of the true perpetrator. Eight of the nine offenders experienced “wrongful liberty” and could have
committed additional crimes. Six of them were found to have been arrested and convicted of a total of
99 additional crimes, including 35 felonies and 13 violent offenses.
4NORRIS ET AL.
TABLE 1Summary findings of previous true perpetrator studies
Study Wrongful Conviction Cases Examined
Additional Crimes of True
Perpetrators Identified
Conroy and Warden (2011) 85 exoneration cases from Illinois,
1989–2010a
97 additional felonies
West and Meterko (2016) 325 DNA exonerations across U.S.,
1989–2014b
142 additional violent crimes
Baumgartner et al. (2018) 36 exoneration cases from North
Carolina, 1943–2013c
99 additional crimes
aConroy and Warden (2011) noted that they could not identify the true perpetrators in all 85 cases, although it is somewhat unclear how
many cases were included in their total calculations.
bTrue perpetrators were identified in 159 of these exonerations, although it is important to note that multiple exonerations may correspond
to one true perpetrator, or multiple true perpetrators may correspond to one exoneree.
cTrue perpetrators were identified in nine of these cases, although only eight remained at liberty to commit additional crimes. The ninth
was arrested at the same time as the exoneree.
The reported crimes committed by the actual offenders while they remain free after the wrongful
conviction of others are striking, in particular, because these accounts almost certainly underestimate
the true number of additional offenses for which these offenders are responsible. The available data
indicate that many of the true perpetrators in wrongful conviction cases are repeat offenders. For exam-
ple, Laudan (2018) estimated that serial offenders who eluded arrest and conviction for crimes would
commit, on average, 1.2 additional serious crimes while at liberty during the average 3-year period
when they would have been incarcerated if they had been apprehended. Moreover, crimes attributed
to offenders during periods of “wrongful liberty” necessarily will be limited to cases in which wrong-
ful convictions are discovered, a subset likely representing only the “tip of the iceberg” (e.g., Garrett,
2014, p. 980) of all wrongful convictions. Many true perpetrators will never be identified (see Conroy
& Warden, 2011), and even when they are, they may have committed an untold number of offenses
in addition to those for which they are ultimately held accountable. As Baumgartner and colleagues
(2018) urged, systematic study of the crimes committed by the true offenders after they escape jus-
tice in wrongful conviction cases should be conducted across multiple jurisdictions, rather than being
restricted to a single state such as Illinois or North Carolina. We attempt to address this gap in the
literature by examining the crimes committed by true perpetrators in a collection of cases from across
the United States in which DNA was used to exonerate the innocent and to identify the guilty offender.
2CURRENT STUDY
Our goal is to examine the crimes of true perpetrators in cases of wrongful conviction across multiple
jurisdictions. We restrict our analysis to cases compiled by the Innocence Project, which involve only
DNA-based exonerations. Although this restriction limits the types of cases in our study, it ensures that
we know, with a great degree of certainty, that the exonerees were factually innocent and that the true
perpetrators were guilty. As Risinger (2007, p. 769) noted, “the most bulletproof DNA exonerations are
those where not only is the previously convicted person exonerated, but also where the true perpetrator
is identified using the same DNA.”4
The Innocence Project provided us with the initial data on exoneration cases in which DNA testing
also was used to identify a true perpetrator. The files included 143 exoneration cases originating in 28
different states,5the names of true perpetrators identified in those cases (when available), as well as
additional notes for some cases. In several instances, the same case involved multiple exonerees and/or
NORRIS ET AL.5
Exonerations
•Initial list of 143
exonerations
provided by
Innocence Project
Criminal Events
•Because of group
exonerations, 143
exonerations
correspond to 108
different cases or
criminal incidents
True Perpetrators
•Across those 108
cases or criminal
incidents, 121 true
perpetrators were
identified
FIGURE 1 Breakdown of individual exonerations, total cases, and true perpetrators
multiple true perpetrators. For example, in the well-known Central Park Jogger case, five teenagers
were convicted of a rape and an assault in New York that was actually committed by one individual.
Conversely, for example, Michael Anthony Green was convicted of a rape in Texas that was committed
by four men. Because of cases such as these, the ratio of exonerees to true perpetrators is not one to
one. Ultimately, the list of 143 exonerations provided by the Innocence Project corresponded to 121
true perpetrators who were identified through DNA testing. We focused on these true perpetrators in
our search for criminal activity, which is depicted in Figure 1. The complete listing of exonerees and
corresponding true perpetrators is provided in the Appendix, arranged by wrongful conviction case.
We attempted to secure official identification numbers for the 121 true perpetrators through state
departments of corrections (DOCs) to facilitate our search for criminal histories. When this informa-
tion was unavailable, we sought other identifying information (e.g., dates of birth and aliases) through
online searches. We then examined official criminal histories, narratives of exoneration cases provided
by the Innocence Project and the National Registry of Exonerations, and accounts from reputable media
sources. We also used Lexis Nexis and Google searches to obtain information about the true perpetra-
tors’ criminal histories.
We focused on crimes committed by true perpetrators after the date of the crime for which an
innocent person was wrongly convicted. We initially sought to include only subsequent crimes for
which the true perpetrators ultimately were convicted. Because we could not obtain official conviction
records for many people in our data set and because convictions do not capture the full range of
additional crimes for a variety of reasons, we eventually expanded our criteria for inclusion. The crimes
included in our results are those based on an official record of conviction or arrest, or for which reliable
sources—established news outlets or nonprofit organizations—provided information that included
details about the event (type of crime, date of the incident, etc.) and a clear link between the crime and
the individual, such as an admission of guilt by the true perpetrator. A discussion of the limitations of
our approach is provided in a following section.
The difficulties associated with locating and compiling information about crimes that might be
decades old, in combination with our inclusion of crimes in addition to those documented by offi-
cial conviction records, necessarily required a measure of subjective judgment about the incidence of
true perpetrators’ offending. To this end, we followed the advice of Baumgartner and colleagues (2018,
p. 1279), who recommended “using caution and common sense when working with the data and erring
on the side of underreporting.” We consequently made conservative assessments. For example, we
found indications from various sources that some true perpetrators may have engaged in subsequent
criminal activity, but if we could not find additional details, such as specifics of the crime and the
date of offense, we did not include them in our analyses. In addition, several of the true perpetrators’
records indicated multiple charges, but we could not verify whether those charges represented distinct
crimes or resulted from a single criminal event (e.g., the person committed a robbery but was charged
with multiple associated offenses). We recorded such situations as involving only one additional crime
rather than multiple additional crimes.
6NORRIS ET AL.
TABLE 2Factors contributing to wrongful convictionsa
Fact or s
Exonerations in Our Data
Set (n=141)
First 325 DNA Exonerations
(West & Meterko, 2016)
Eyewitness misidentification 57% 72%
Misapplication of forensic science 47% 47%
False confessions 43% 27%
Unreliable informants 23% 15%
aPercentages do not add up to 100% because many cases involve multiple contributing factors.
In employing these criteria, we thus included 109 true perpetrators in our analysis. We could not
locate sufficiently detailed information about the remaining 12 true perpetrators or their activities to
include them.6
3FINDINGS
We present our findings in three parts. We begin by comparing the factors contributing to wrongful
convictions in all of the DNA-based exoneration cases identified by the Innocence Project to the subset
of cases in which true perpetrators were identified. We do so to determine whether the cases involving
known true perpetrators seem to differ systematically from the other Innocence Project exoneration
cases. We next present our findings about the additional crimes committed by the 109 true perpetrators
for whom we could find additional information. By using the previously described heuristic framework
developed by Forst and Huff (2018), we extrapolate from our findings to estimate the potential harm
generated by wrongful convictions that is directly relevant to crime control interests.
3.1 Factors contributing to wrongful convictions
All of the cases in our data set depend on DNA analysis, and they consequently are heavily skewed
toward rape and sexual assaults. Although not representative of all crimes or even of all known wrongful
convictions, the cases we examined can be compared with the larger body of DNA-based exonerations
compiled by the Innocence Project.
The factors that contributed to the wrongful conviction cases in our collection are presented in
Table 2; because we use a set of cases provided by the Innocence Project, we rely on their coding of
the contributing factors, which allows for a comparison with its larger set of cases. The initial list pro-
vided by the Innocence Project contained 143 exonerees. For two of them—Jeramie Davis and Robert
Dewey—however, no information is provided about the contributing factors from either the Innocence
Project or the National Registry of Exonerations. Thus, the calculated percentages for each contributing
factor correspond to the 141 exoneration cases for which this information is available. The most preva-
lent contributing factor to the wrongful convictions in our data set was eyewitness misidentification,
present in 80 (56.7%) of the cases, followed by misapplication of forensic science7(n=66, or 46.8%),
false confessions (n=61, or 43.3%), and the testimony of unreliable informants (n=32, or 22.7%).
As reported in Table 2, the ordering of the contributing factors in terms of prevalence remains the
same in our subset of cases and in the larger set of the first 325 DNA-based exonerations from 1989
to 2014. Eyewitness misidentification, however, appears less often in our subset of cases than in the
larger body of DNA exonerations (56.7% vs. 72%), whereas false confessions (43.3% vs. 27%) and
unreliable informants (22.7% vs. 15%) appear as contributing factors more often in our collection of
cases.
NORRIS ET AL.7
7
102
68
48
30 29
19 14 11 6532
0
20
40
60
80
100
120
0 1 2 3 4 5 6 7 8 9 10 11 12
Number of True Perpetrators
Minimum Number of Addional Crimes
FIGURE 2 Number of additional crimes ×number of true perpetrators [Color figure can be viewed at
wileyonlinelibrary.com]
Two possible explanations for these discrepancies are worth noting. The first is that they may be
attributable to the six different group exonerations in our data set in which three or more people were
exonerated, and in which false confessions and/or use of informants were contributing factors, but not
eyewitness errors.8The second speculative, but more interesting, possibility is that the types of factors
that contribute to erroneous convictions may affect the likelihood that the true perpetrator will be
identified or, alternatively, that the chances of official exoneration increases when the true perpetrator
is identified in cases involving primary (false confessions) and secondary confessions (informants).
Although these points are beyond the scope of our study, they warrant further exploration.
3.2 Crimes committed by true perpetrators during their period of wrongful
liberty
As discussed earlier, our data collection process produced information on 109 of the true perpetrators
from the initial data file. We found no indication that 7 among the 109 (6.4%), some of whom were
deceased or incarcerated for an earlier offense, committed crimes subsequent to the wrongful convic-
tion offense. On the other hand, 102 (93.6%) of the true perpetrators engaged in subsequent criminal
activity after the wrongful conviction offense. Our findings demonstrate that these individuals engaged
in “a collective crime spree” (Conroy & Warden, 2011, para. 2) while experiencing a period of wrong-
ful liberty, committing a total of 337 additional offenses.
At the individual level, 34 true perpetrators seemed to be one-time subsequent offenders, whereas the
remaining 68 committed multiple additional offenses. Most committed five or fewer additional crimes,
although some committed many more, as shown in Figure 2. Among the 68 true perpetrators who
committed two or more additional offenses, five committed at least 10 more crimes, including two who
committed 12 additional offenses while an innocent person was arrested, convicted, and imprisoned.
The types of additional crimes varied, but many were serious offenses that claimed additional vic-
tims. More than 60% (n=206, or 61.1%) of these additional offenses were felonies and/or crimes of
violence against other persons. More specifically, the crimes included 43 attempted or consummated
criminal homicides (murder or manslaughter), 94 completed or attempted sex offenses (rape, sexual
assault, child molestation, indecent or lewd acts with a child), and 35 drug-related offenses. Other
offenses comprised a wide variety of more prevalent crime types including aggravated and simple
8NORRIS ET AL.
12.7%
27.9%
10.4%
49%
Homicide-related Sex offenses Drug offenses Other
FIGURE 3 Types of additional crimes (n=337) [Color figure can be viewed at wileyonlinelibrary.com]
assault, robbery, burglary, theft, and threatening a public official, among others. The relative frequency
of crime types committed is displayed in Figure 3.
3.3 Potential harms of wrongful liberty: Extrapolating from our data
Our findings are telling in their own right, but it is interesting to extrapolate from these results using
the heuristic framework suggested by Forst and Huff (2018) who, as discussed earlier, used FBI and
BJS data to estimate that approximately 339,000 convictions were obtained for violent crimes in 2014.
Assuming a wrongful conviction rate of 1% to 3% and that each true perpetrator committed one addi-
tional crime, they projected that wrongful convictions generate an additional 3,000–10,000 crimes per
year. We can be somewhat more precise in our estimates, however, and our findings indicate that these
projections may underestimate the extent of crimes committed by true perpetrators.
Calculating an exact wrongful conviction rate is impossible; this “dark figure of innocence” (Bedau
& Radelet, 1987, p. 83) is “not merely unknown but unknowable” (Gross, O’Brien, Hu, & Kennedy,
2014, p. 7230). Numerous scholars, however, have attempted to estimate such a rate using various
techniques and a variety of case compilations. Estimates, some more reliable than others, have ranged
from a fraction of 1% (Kansas v. Marsh, 2006; Thomas, 2018) to nearly 40% for some subsets of crime
types (Loeffler et al., 2018; Poveda, 2001).9Most estimates tend to fall between 1% and 5%. One
empirical estimate was offered by Risinger (2007), who generated a conservative wrongful conviction
rate of 3.3% to 5.0% for capital rape-murder cases. Furthermore, Risinger suggested that the rate of
wrongful convictions for noncapital murders and rapes, which make up the majority of our cases,
could plausibly be as high as or higher than for the capital cases he used to generate his estimate. In
a second study of capital cases that resulted in death sentences, published in the Proceedings of the
National Academy of Sciences, the authors conservatively estimated a 4.1% error rate (Gross et al.,
2014). Although death penalty cases are unique, they are of interest concerning convictions for other
serious crimes. Indeed, Loeffler and colleagues (2018) recently used self-report data from Pennsylvania
inmates, with an adjustment for implausible responses, to estimate an overall wrongful conviction rate
of 6.2%.
Although these figures are only estimates, they are consistent with many other wrongful conviction
estimates and even lower than some. For example, Walsh and colleagues (Walsh, Hussemann, Flynn,
Yahner, & Golian, 2017) reviewed more than 400 sexual assault cases in Virginia and, using DNA
testing not available at the time of the trials, estimated that more than 11% of the convictions were
erroneous (for a discussion, see Baumgartner et al., 2018). Thus, with these empirically derived esti-
mates, which range from 3.3% to 6.2%, as our boundaries, it is useful to think about the potential harms
inflicted by true perpetrators during their wrongful liberty.
NORRIS ET AL.9
Based on the estimate of 339,000 convictions annually for violent crime noted earlier (Forst & Huff,
2018), an error rate of 3.3% to 6.2% would yield between approximately 11,000 and 21,000 wrongful
convictions per year. As mentioned, however, approximately 36% of cases in the National Registry of
Exonerations database are no-crime cases, leaving 64% of wrongful convictions with a true perpetrator.
With this adjustment, true perpetrators may escape apprehension and conviction in an estimated 7,040
to 13,440 violent crime cases per year. Among the true perpetrators included in our analysis (n=109),
we found 337 crimes, or an average of 3.1 additional crimes per person. Thus, if 7,040 to 13,440
perpetrators remain free and each commits an average of 3.1 crimes during his or her period of wrongful
liberty, the wrongful convictions that occur each year would ultimately enable true perpetrators to
commit between 21,824 and 41,664 additional crimes.
It is important to note that the average of 3.1 crimes per true perpetrator is only an estimate and that
this calculation may be artificially inflated to some degree if our data set includes true perpetrators who
are at a higher risk of recidivism than other offenders, particularly those in other wrongful conviction
cases. For example, it is possible that the ability to identify true perpetrators increases with their level
of recidivism, thus, skewing our data toward high rates of reoffending. Our data, however, do allow
for a slightly more conservative estimate based on a different assumption, but the calculations are
still telling. As we noted earlier, we initially received a list of 121 identified true perpetrators, but we
included only 109 in our analysis because we could not locate the remaining 12 or secure detailed
information about them. Even if we assume that these additional 12 true perpetrators did not commit
any additional offenses and, thus, use all 121 true perpetrators in our denominator, the resulting average
of 2.8 crimes per offender (337 crimes / 121 true perpetrators) would generate an estimate of nearly
20,000 (7,040 ×2.8) to 38,000 (13,440 ×2.8) total offenses. These projections remain astounding,
especially considering that they include only the wrongful convictions for violent crimes that occur in
a single year. Over time, the potential crime-related costs of wrongful conviction are tremendous.
3.4 Methodological limitations
Our results should be interpreted with some caution because of several methodological limitations.
First, as with all research on wrongful convictions, questions about guilt, innocence, and exonera-
tion may linger. It is not always possible to determine with absolute certainty that individuals iden-
tified on the lists compiled by the Innocence Project and the National Registry of Exonerations are
innocent. Restricting our cases, however, to those in which the use of DNA not only demonstrated
the innocence of the exonerees(s) but also identified the true perpetrator(s) brings us as close to
certain as possible. Yet, the DNA restriction limits us in other ways, such as resulting in an over-
representation of rape and sexual assault cases, which in turn may affect the generalizability of our
results.
It is true that for some cases in our sample we lack complete information about the crimes committed
by true perpetrators. As noted, official arrest and conviction records were not available for all of the
offenders in our data set. Moreover, official crime records obviously may fail to capture the full extent
of criminal activity. We attempted to compensate for these shortcomings by relying on other reliable
sources of information, introducing a measure of subjective judgment to the reported crime totals that,
we believe, on balance enhances the accuracy of the tallies. In short, we attempted to be inclusive
beyond official records while remaining cautious and somewhat conservative in our judgments about
the true perpetrators’ criminal activity during their period of wrongful liberty.
Finally, it is important to reemphasize that our extrapolation is based on estimates that carry their
own shortcomings. As we have noted, the actual rate of wrongful convictions is unknown, and esti-
mates vary widely depending on method and case selection. Our extrapolations generated from these
10 NORRIS ET AL.
estimates are based on an unrepresentative set of cases and on several assumptions, making generaliz-
ing from them preliminary.
Despite these limitations, we believe that our findings raise important issues for the study of wrong-
ful convictions and for criminology more broadly. In light of the implications for policy and practice,
discussed in the next section, it will be important for scholars to turn their attention to the widespread
harms generated by wrongful convictions, including the crimes committed by true perpetrators during
periods of wrongful liberty.
4DISCUSSION AND POLICY IMPLICATIONS
To the best of our knowledge, this study is the first to rely on a national collection of exoneration cases
in which true perpetrators have been identified to provide an estimate of the new crimes committed by
the true perpetrators while they remain at large during their period of wrongful liberty. Our findings
indicate that wrongful convictions pose a serious threat to public safety and that their prevention
is urgent not only to protect innocent accused citizens from injustice but also to protect innocent
victims and their families from future harm. Shifting the conversation to emphasize that wrongful
convictions generate harms well beyond their direct effects on exonerees is important for criminal
justice discourse, policy, and practice.
4.1 Wrongful convictions and crime control: Framing, public opinion, and
policymaking
Contemporary discourse around criminal justice increasingly involves the conviction of the innocent.
Scholars have documented the rise of the “innocence frame” (Baumgartner, De Boef, & Boydstun,
2008) in media coverage since the 1990s. This development, in combination with the mounting number
of exonerations, the founding and growth of the international Innocence Network, and the currency of
the topic of wrongful convictions in books, films, television shows, and podcasts, is an essential element
of the “innocence movement” that has risen over the last two decades (Norris, 2017). The penetration of
wrongful convictions into the public consciousness is reflected in various measures of public opinion
about justice issues.
For example, concerns about wrongful convictions have been empirically linked to declining sup-
port for capital punishment. Recent Gallup Poll data have showed public support for the death penalty
to be at its lowest levels since the 1970s (Gallup, n.d.), and several scholars have suggested that this
trend is related in part to increasing exonerations and concerns that innocent persons are at risk of exe-
cution (e.g., Baumgartner et al., 2008; Fan, Keltner, & Wyatt, 2002; Sarat, 2005). Less is known about
how increased awareness of wrongful convictions has influenced public attitudes on criminal justice
issues beyond the death penalty context, but some research findings indicate it may be meaningful. For
example, evidence reveals that public perceptions of the courts’ ability to establish the truth (i.e., con-
victing the innocent or freeing the guilty) play a role in shaping perceived judicial legitimacy (Tyler
& Sevier, 2013/2014). Thus, it is possible that wrongful convictions can meaningfully affect public
opinion, especially if framed so the full consequences of wrongful convictions are apparent.
Framing has been described as “the process by which people develop a particular conceptualization
of …or reorient their thinking about an issue” (Chong & Druckman, 2007, p. 104). In a vast body of
literature, scholars have analyzed how the manner in which issues are framed or communicated can
influence people’s attitudes and responsiveness to information. In particular, changes in how issues are
presented can dramatically affect public views about them. For instance, Nelson, Clawson, and Oxley
NORRIS ET AL.11
(1997), in their classic study, found that people expressed more tolerance for the Ku Klux Klan when
presented with a news story framing a rally as a free speech issue than when it was described as a dis-
ruption to public order. Similarly, Sniderman and Theriault (2004), in their experimental study, showed
that more than 80% of respondents favored allowing an extremist group to hold a rally when framed as
a free speech issue, although fewer than half favored allowing the rally when it was framed as carrying
a risk of violence. Such framing effects have been found across a range of political attitudes, includ-
ing support for government spending and campaign finance (for a review, see Chong & Druckman,
2007). In the realm of criminal justice, through framing studies, scholars have reported similar results
regarding views about the death penalty (Peffley & Hurwitz, 2007), the Supreme Court (Nicholson &
Howard, 2003), and how the racialized framing of crime affects corresponding public attitudes (e.g.,
Applegate & Sanborn, 2011; Park, Holody, & Zhang, 2012; Ramirez, 2013).
Such framing effects may be equally or more important in the context of wrongful convictions and
public attitudes about the justice system, justice officials, and policy reforms. In most popular and
scholarly writings about wrongful convictions, the authors have emphasized errors of due process, fac-
tors that lead to erroneous outcomes, and the harms faced by exonerees. Accordingly, the issues are
often construed as being liberal and criminal defense oriented (Norris, 2017), sparking critiques and
rebuttals from law enforcement officials and prosecutors (e.g., Marquis, 2004) who, in turn, often are
construed as opponents of the innocence movement. Yet, our findings about the crimes committed by
true perpetrators indicate that strong incentives exist for law enforcement officials to be allies in the
quest to reduce wrongful convictions. To make the justifications for this alliance clearer, discourse
about wrongful convictions should contemporaneously highlight concerns about public safety in addi-
tion to emphasizing due process considerations. In other words, wrongful conviction issues should be
framed not only as due process concerns but also as integral to crime control.
Framing wrongful convictions as a public safety issue is important for several reasons, related to both
public opinion and policymaking, including enacting change in local criminal justice practices. For
instance, such framing may help influence opinions among those who may not be swayed by due pro-
cess and defense-oriented concerns, especially political conservatives who are likely to be more recep-
tive to concerns about public safety and law and order. For example, while advocating for a wrongful
conviction review panel, one senator from New Jersey, a self-described “law and order conservative,”
employed a crime control frame, pointing out that whenever an error occurs, “that means the person
who committed the crime is still out there” (Sullivan, 2017, para. 6). As one innocence advocate in
North Carolina described it, the “true-perpetrator angle” is the “Republican pitch” for wrongful con-
viction reform (Norris, 2017, p. 158). Our findings provide tangible evidence about the potential scope
of crime generated when errors are made in arresting and convicting the innocent.
Not only is employing a crime control frame apt to help influence lawmakers’ views directly, but it
also has the potential to help shape public attitudes and thereby play a substantial role in promoting
policy changes and achieving progressive justice system reforms. The findings reported in an expansive
amount of literature show that public opinion is important for policymaking, particularly when issues
are politically salient (Burstein, 2003). Indeed, in the realm of criminal justice, study findings indicate
that shifts in public opinion have been influential with respect to death penalty practices and policies
(e.g., Mooney & Lee, 2000; Norrander, 2000). In light of the increasing prominence of the innocence
frame since the late 1990s (Baumgartner et al., 2008), presenting different dimensions of the problems
associated with wrongful convictions could be a powerful tool for generating widespread support for
systemic reforms.
Importantly, alternative frames may be particularly effective in helping change or galvanize views
about wrongful convictions. One factor influencing how persuasive a frame may be is the presence of
competing frames that present opposing views on an issue (Chong & Druckman, 2007; Sniderman &
12 NORRIS ET AL.
Theriault, 2004). In the case of wrongful convictions, the “due process” and “crime control” frames
are not necessarily competing, but they can be viewed in many respects as complementary, in that
they do not lead to opposing conclusions. Most notably, whether due process or crime control values
are prioritized, few would desire to see the innocent convicted or to allow the guilty to remain at
large to commit additional crimes. Rather than competing to sway opinion in one direction or another,
the due process and crime control frames may operate in tandem to broaden the appeal of working
collaboratively to improve the system by reducing errors. In this sense, as Findley (2008, p. 133) has
argued, the innocence movement uniquely and effectively “merges crime control and due process.”
4.2 Local policy adoption and implementation: Informed debate and
collaboration
In addition to its potential for affecting public opinion and policymaking on a broad level, framing
innocence in a way that highlights public safety can contribute to achieving reforms in local law
enforcement agencies and policies. The leading contributing factors to the wrongful convictions
in our data set mirror many of those found in broader collections of exoneration cases: eyewitness
misidentification, forensic errors, false confessions, and the use of unreliable informants. The results
of conducting extensive research on these issues have led to incorporation of a series of evidence-based
practices that may help mitigate such errors. For example, psychological researchers have found
that changes to eyewitness identification procedures—double-blind administration, specific witness
instructions, proper lineup construction, and so forth—may reduce the likelihood that witnesses mis-
takenly pick the wrong person out of a lineup or photo array (Wells et al., 2018; Wells, et al., 1998).
Furthermore, interrogation experts have recommended changes such as limiting the length of interro-
gation sessions (e.g., Bull, 2014; Leo, 2008) and video-recording interrogations as measures to reduce
the likelihood of excessive coercion and help guard against false confessions (Kassin et al., 2010).
These and other reforms—forensic lab accreditation and oversight, enhanced discovery and disclosure
when using informants, and the like—have primarily been championed by the innocence community
and recommended by scholars as safeguards to protect the innocent. Given the extensive criminal costs
of wrongful convictions as documented in our results, however, we argue that such reforms can and
should be conceptualized concurrently as ways to reduce crime and enhance public safety. Framing the
prevention of wrongful convictions as a way to enhance crime control as well as to promote due process
not only allows for a more complete and nuanced understanding of the issues and a more informed
policy discussion but also carries implications for collaboration in adopting and implementing
changes.
The crime control implications of our study are important considerations for a fully informed policy
debate about wrongful convictions. One source of hesitation in adopting some of the recommended
best practices to forestall wrongful convictions is their potential cost; the concern is that even though
such changes may reduce the likelihood of wrongful convictions, they might simultaneously increase
the difficulty of obtaining accurate convictions. This concern has been highlighted in debates involving
psychologists and legal scholars about eyewitness reforms. For example, Clark (2012) and Laudan
(2012) have expressed concerns that various recommended reforms in eyewitness identification
policies enhance the risk that guilty offenders will not be identified. In contrast, Wells and colleagues
have suggested that additional accurate identifications of the guilty “are ill-gotten” if they come at
the risk that innocent individuals will erroneously be identified through unnecessarily suggestive
procedures (Wells, Steblay, & Dysart, 2012, p. 265). Newman and Loftus (2012) have similarly argued
that gains in the form of fewer false negative identifications often come at the expense of coercion
and suggestiveness, and should not be embraced even if fewer true identifications are obtained.
NORRIS ET AL.13
Analogous concerns have been raised with respect to interrogation reforms. In particular, some have
argued that the presence of a recording device will inhibit suspects from speaking, thus, reducing
truthful confessions. Proponents have countered that there is scant empirical support for these concerns,
noting that agencies that record interrogations have reported that suspects do not refrain from talking
and may provide more incriminating statements than during sessions that are not recorded (see Leo &
Richman, 2007; Sullivan, 2004).
These debates about the utility of investigative reforms hinge on the assumption that by protecting
the innocent it is necessary to sacrifice catching the guilty, that a choice must be made between the
competing perspectives of Packer’s (1968) “crime control” and “due process” models. As Findley
(2008) has argued, and as our findings show, however, avoiding wrongful convictions can embrace both
ideals. The results of our analysis indicate that when true perpetrators remain at large after an innocent’s
wrongful conviction, their propensity to commit new offenses represents a significant crime control
problem. Of course, trade-offs sometimes do exist between protecting the innocent and apprehending
the guilty, different weights can be assigned to the respective errors, and policy makers will disagree
about their significance. However these challenging issues are resolved, policy makers should at a
minimum be mindful that when actual offenders remain at large after wrongful convictions, the adverse
consequences transcend the injustice experienced by the innocent. Significant social costs are also
likely to be incurred in the form of the additional crimes committed by the true perpetrators during
their periods of wrongful liberty. These considerations should be a vital aspect of informed policy
discussions.
Finally, we note the importance of framing of issues surrounding wrongful convictions for the adop-
tion and implementation of evidence-based reforms. As discussed, many changes advocated by the
innocence community involve police practices. Collaborations with local law enforcement agencies
are important if the innocence reform agenda is to succeed. Presenting innocence-related concerns as
concurrently important to crime control and public safety may encourage greater collaboration among
advocates who traditionally have worked to prevent and correct wrongful convictions and those who
traditionally have supported law enforcement. Indeed, the innocence and law enforcement communi-
ties can both gain by working in concert to improve the accuracy of the justice system, thereby averting
wrongful convictions and preventing the harms committed by true perpetrators enjoying wrongful lib-
erty. Such collaborative efforts are possible.
For example, in 2012, the International Association of Chiefs of Police and the U.S. Department of
Justice hosted a national policy summit involving police, prosecutors, academics, innocence advocates,
and others that developed fruitful strategies to reduce wrongful arrests and convictions (International
Association of Chiefs of Police, 2013). Collaborations of this nature may increase the likelihood that
evidence-based reforms designed to improve the justice system will be adopted and implemented (see
also Batts, deLone, & Stephens, 2014). Such collaborative efforts may increase if the involved parties
understand the full breadth of problems associated with wrongful convictions, including the dangers
posed by true perpetrators during periods of wrongful liberty.
5CONCLUSION
We have attempted to document the crimes committed by known true perpetrators in DNA-based exon-
eration cases while they remain at large after the wrongful conviction of an innocent person for their ear-
lier crime. We catalogued an additional 337 offenses committed by true perpetrators, including many
crimes of violence, during their periods of wrongful liberty. Even these totals are unlikely to represent
all of the additional crimes the true perpetrators committed while others instead were punished.
14 NORRIS ET AL.
Although the findings from innocence scholarship have highlighted the harms suffered by innocents
because of wrongful conviction, it is likewise important to understand the associated costs to public
safety in the resulting additional crimes that are committed and the new victimizations that occur. Of
course, not all wrongful convictions can be prevented, and even with improved justice system measures
that determine guilt and innocence more reliably, some offenders will elude apprehension and convic-
tion and remain at large to reoffend. Our findings nevertheless make clear that efforts to reduce justice
system errors importantly serve the goals of crime control just as they advance the ideals of due process.
We can and should extend our appreciation of the harms generated by wrongful convictions to
include the profound toll taken on public safety and crime control efforts when the true perpetrators of
crime remain at large to reoffend and claim new victims. This perspective is critical to helping reframe
issues associated with wrongful convictions, fashioning and implementing innocence-related policies,
and enacting broader criminal justice reforms. As others have pointed out (Acker, 2013; Baumgartner
et al., 2018; Thompson and Baumgartner, 2018), an awareness of the expansive circle of harm created
when innocent persons are convicted and the guilty escape punishment is important to help inspire a
more fully informed and meaningful debate about the flaws in our justice system and how to alleviate
them. Indeed, efforts to reduce wrongful convictions can benefit a vast body of innocents, not only
the wrongly accused but also untold numbers of citizens who will be spared victimization by the true
perpetrators of crime who have eluded justice.
The National Institute of Justice affirms on its website that “[t]he strength of our criminal justice
system depends on its accuracy—its ability to convict the guilty and to clear the innocent” (National
Institute of Justice, n.d.). As emphasized throughout this article, we concur that these are not mutually
exclusive goals. We can, indeed, reduce crime by protecting the innocent, and our criminal justice
system is much the stronger when convicting the guilty and clearing the innocent are accomplished
simultaneously.
ENDNOTES
1Figure current as of 1/22/19.
2It is important to note that this collection of cases represents exonerations with a basis in factual innocence, but it is
impossible to know with certainty that all of these individuals are actually innocent. As Zalman (2014) noted, many
cases involve issues of both legal and factual guilt, and there is an element of subjectivity in related judgments. As
suggested by the National Registry’s own definitions, and as co-founder Samuel Gross has said, however, most exonerees
have been proven innocent or are likely innocent.
3We are aware of one current ongoing research project in which scholars are exploring the harms suffered by original
crime victims. Furthermore, at least one nonprofit organization, Healing Justice, was founded by Jennifer Thompson—
the original victim in the wrongful conviction case of Ronald Cotton—and takes a restorative justice approach to healing
for exonerees and original crime victims (Healing Justice, n.d.).
4It is worth noting that DNA testing is not infallible and that such exonerations are not guaranteed to be indicative of
actual innocence with 100% certainty (Murphy, 2015). The DNA exonerations we use—cases in which the use of DNA
testing indicated both the factual evidence of the exoneree and the factual guilt of a true perpetrator—are as close to
certain as is possible, however, given the inherent challenges of defining and identifying legal and factual innocence.
5The states included AZ, CA, CO, CT, FL, GA, IL, IN, KS, LA, MA, MD, MI, MO, MS, NC, NE, NJ, NY, OH, OK,
PA,TN,TX,VA,WA,WI,andWV.
6The initial list we received from the Innocence Project contained two cases with extremely limited information, and we
could not secure additional details. One of them indicated that the true perpetrator had been identified, but no name
was provided, and we could not reliably identify and locate the true perpetrator in the case. In another exoneration
case, the true perpetrator was only identified by last name. Although we found a DOC record for an individual who
may have been the same person, some identifying information did not match other sources, including the Innocence
NORRIS ET AL.15
Project’s file. Thus, we did not include this case. For the remaining 10 people, we occasionally came across their names
in passing while searching news sources—in some, there may even have been an indication that they committed another
crime—but we could not find detailed information about their criminal activity or lack thereof.
7As West and Meterko (2016, p. 743) noted, the Innocence Project defines misapplication of forensic science “as an
instance in which we know that forensic evidence was used to associate, identify or implicate someone who was later
conclusively proven innocent with post-conviction DNA testing, thereby demonstrating that the original forensic evi-
dence was incorrect.” It, however, “do[es] not count instances in which forensic science was applied properly” (emphasis
in original). Cases in this category involve a variety of techniques including “serology, hair microscopy, bite mark anal-
ysis, DNA, dog scent, and fingerprints,” among others (p. 745). The types of issues also varied, ranging from laboratory
errors and overstatements by expert witnesses, to negligence and misconduct.
8These cases included the following group exonerations: (1) Barr, Harden, Sharp, Taylor, and Veal; (2) Bradford, Ollins,
Ollins, and Saunders; (3) Dean, Shelden, White, Winslow, Taylor, and Gonzalez; (4) McCray, Richardson, Salaam,
Santana, and Wise; (5) Richardson, Saunders, Swift, and Thames; and (6) Ruffin, Bivens, and Dixon.
9Like all estimates, those at the extreme high and low ends each have their methodological shortcomings, but they
highlight the wide range of findings that may arise based on technique and case selection. For example, Justice Scalia’s
estimation in Kansas v. Marsh (2006) wasdeveloped by simply dividing the number of exonerations over a period of time
by the total number of felony convictions over that same period. The obvious flaw in this approach is that exonerations
are a highly inaccurate proxy for wrongful convictions; they represent just the select few that are reinvestigated and
exonerated, thus, generating an unreasonably low rate. On the other hand, Poveda (2001) used data from RAND’s
Inmate Survey to examine self-reports of innocence. He found an overall self-reported innocence rate of ∼15.4%, but
the rates are especially high among inmates for rape (37.7%) and other sex offenses (26.9%).
REFERENCES
Acker, J. R. (2013). The flipside injustice of wrongful convictions: When the guilty go free. Albany Law Review,76,
1629–1712.
Applegate, B. K., & Sanborn, J. B. (2011). Public opinion on the harshness of local courts: An experimental test of
question wording effects. Criminal Justice Review,36, 487–497.
Balko, R., & Carrington, T. (2018). The cadaver king and the country dentist: A true story of injustice in the American
South. New York: Public Affairs.
Batts, A. W., deLone, M., & Stephens, D. W. (2014). Policing and wrongful convictions. Retrieved from https://www.
ncjrs.gov/pdffiles1/nij/246328.pdf
Baumgartner, F. R., De Boef, S. L., & Boydstun, A. E. (2008). The decline of the death penalty and the discovery of
innocence. New York: Cambridge University Press.
Baumgartner, F. R., Grigg, A., Ramirez, R., & Lucy, J. S. (2018). The mayhem of wrongful liberty. Documenting the
crimes of true perpetrators in cases of wrongful incarceration. Albany Law Review,81, 1261–1286.
Bedau, H. A., & Radelet, M. L. (1987). Miscarriages of justice in potentially capital cases. Stanford Law Review,40,
21–173.
Bull, R. (Ed.). (2014). Investigative interviewing.NewYork:Springer.
Burstein, P. (2003). The impact of public opinion on public policy: A review and an agenda. Political Research Quarterly,
56, 29–40.
Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science,10, 103–126.
Clark, S. E. (2012). Costs and benefits of eyewitness identification reform: Psychological science and public policy.
Perspectives on Psychological Science,7, 238–259.
Conroy, J., & Warden, R. (2011). Special investigation: The high costs of wrongful convictions. Better Govern-
ment Association. Retrieved from https://www.bettergov.org/news/special-investigation-the-high-costs-of-wrongful-
convictions
Fan, D. P., Keltner, K. A., & Wyatt, R. O. (2002). A matter of guilt or innocence: How news reports affect support for
the death penalty in the United States. International Journal of Public Opinion Research,14, 439–452.
Findley, K. A. (2002). Learning from our mistakes: A criminal justice commission to study wrongful convictions. Cali-
fornia Western Law Review,38, 333–353.
16 NORRIS ET AL.
Findley, K. A. (2008). Toward a new paradigm of criminal justice: How the innocence movement merges crime control
and due process. Texas Tech Law Review,41, 133–173.
Forst, B., & Huff, C. R. (2018). Preventing violent crimes by reducing wrongful convictions. In A. T. Vazsonyi, D. J.
Flannery, & M. DeLisi (Eds.), The Cambridge handbook of violent behavior and aggression (2nd ed.). Cambridge,
England: Cambridge University Press.
Gallup. (n.d.). Death penalty. Gallup News. Retrieved from https://news.gallup.com/poll/1606/death-penalty.aspx
Garrett, B. L. (2014). The banality of wrongful executions. Michigan Law Review,112, 979–992.
Givelber, D. (1997). Meaningless acquittals, meaningful convictions: Do we reliably acquit the innocent? Rutgers Law
Review,49, 1317–1396.
Gross, S. R., O’Brien, B., Hu, C., & Kennedy, E. H. (2014). Rate of false conviction of criminal defendants
who are sentenced to death. Proceedings of the National Academy of Sciences,111, 7230–7235. http://doi.org/
10.1073/pnas.1306417111
Healing Justice. (n.d.). Healing justice: Home. Retrieved from http://www.healingjusticeproject.org/
Huff, C. R., Rattner, A., & Sagarin, E. (1986). Guilty until proved innocent: Wrongful conviction and public policy.
Crime & Delinquency,32, 518–544.
International Association of Chiefs of Police (IACP). (2013). National summit on wrongful convictions: Building
a systemic approach to prevent wrongful convictions. Retrieved from https://www.theiacp.org/resources/national-
summit-wrongful-convictions-building-systemic-approach-prevent-wrongful
Kansas v. Marsh, 548 U.S. 163 (2006).
Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., & Redlich, A. D. (2010). Police-induced confes-
sions: Risk factors and recommendations. Law and Human Behavior,34, 3–38.
Laudan, L. (2012). Eyewitness identifications: One more lesson on the costs of excluding relevant evidence. Perspectives
on Psychological Science,7, 272–274.
Laudan, L. (2018). Different strokes for different folks: Fixing the error pattern in criminal prosecutions by “empiriciz-
ing” the rules of criminal law and taking false acquittals and serial offenders seriously. Seton Hall Law Review,48,
1243–1253.
Leo, R. A. (2008). Police interrogation and American justice. Cambridge: Harvard University Press.
Leo, R. A., & Richman, K. D. (2007). Mandate the electronic recording of police interrogations. Criminology & Public
Policy,6, 791–798.
Loeffler, C. E., Hyatt, J., & Ridgeway, G. (2018). Measuring self-reported wrongful convictions among prisoners. Jour-
nal of Quantitative Criminology. Retrieved from http://doi.org/10.1007/s10940-018-9381-1
Marquis, J. (2004). The myth of innocence. The Journal of Criminal Law & Criminology,95, 501–522.
Mooney, C. Z., & Lee, M-H. (2000). The influence of values on consensus and contentious morality policy: U.S. death
penalty reform, 1956–1982. Journal of Politics,62, 223–239.
Murphy, E. E. (2015). Inside the cell: The dark side of forensic DNA. New York: Nation Books.
National Institute of Justice. (n.d.). Postconviction testing and wrongful convictions. Retrieved from https://www.nij.gov/
topics/justice-system/wrongful-convictions/pages/welcome.aspx
National Registry of Exonerations. (n.d.). Retrieved from http://www.exonerationregistry.org
Nelson, T. E., Clawson, R. A., & Oxley, Z. M. (1997). Media framing of a civil liberties conflict and its effect on tolerance.
American Political Science Review,91, 567–583.
Newman, E. J., & Loftus, E. F. (2012). Clarkian logic on trial. Perspectives on Psychological Science,7, 260–263.
Nicholson, S. P., & Howard, R. M. (2003). Framing support for the Supreme Court in the aftermath of Bush v. Gore.
Journal of Politics,65, 676–695.
Norrander, B. (2000). The multi-layered impact of public opinion on capital punishment implementation in the American
states. Political Research Quarterly,53, 771–793.
Norris, R. J. (2017). Exonerated: A history of the innocence movement. New York: NYU Press.
Packer, H. L. (1968). The limits of the criminal sanction. Stanford: Stanford University Press.
Park, S-Y., Holody, K. J., & Zhang, X. (2012). Race in media coverage of school shootings: A parallel application of
framing theory and attribute agenda setting. Journalism and Mass Communication Quarterly,89, 475–494.
Peffley, M., & Hurwitz, J. (2007). Persuasion and resistance: Race and the death penalty in America. American Journal
of Political Science,51, 996–1012.
Penzell, A. (2007). Apology in the context of wrongful conviction: Why the system should say it’s sorry. Cardozo Journal
of Conflict Resolution,9, 145–161.
NORRIS ET AL.17
Poveda, T. G. (2001). Estimating wrongful convictions. Justice Quarterly,18, 689–708.
Ramirez, M. D. (2013). Punitive sentiment. Criminology,51(2), 329–364. http://doi.org/10.1111/1745-9125.12007
Ramsey, R. J., & Frank, J. (2007). Wrongful conviction: Perceptions of criminal justice professionals regard-
ing the frequency of wrongful conviction and the extent of system errors. Crime & Delinquency,53,
436–470.
Risinger, D. M. (2007). Innocents convicted: An empirical justified factual wrongful conviction rate. Journal of Criminal
Law and Criminology,97, 761–806.
Sarat, A. (2005). Innocence, error, and the “new abolitionism”: A commentary. Criminology & Public Policy,4, 45–54.
Sniderman, P. M., & Theriault, S. M. (2004). The structure of political argument and the logic of issue framing. In
W. E. Saris & P. M. Sniderman (Eds.), Studies in public opinion. Princeton: Princeton University Press.
Sullivan, S. P. (2017, March 11). What happens if you’re wrongfully convicted in N.J.? NJ.com.Retrievedfrom
https://www.nj.com/politics/index.ssf/2017/03/what_happens_if_youre_wrongfully_convicted_in_nj.html
Sullivan, T. P. (2004). Police experiences with recording custodial interrogations. Chicago: Northwestern
University School of Law Center on Wrongful Convictions. Retrieved from http://mcadams.posc.mu.edu/
Recording_Interrogations.pdf
Thomas, G. C. III (2018). Where have all the innocents gone? Arizona Law Review,60, 865–890.
Thompson, J. E., & Baumgartner, F. R. (2018). An American epidemic: Crimes of wrongful liberty. Injus-
tice Watch. Retrieved from http://www.injusticewatch.org/commentary/2018/an-american-epidemic-crimes-of-
wrongful-liberty/
Tyler, T. R., & Sevier, J. (2013/2014). How do the courts create popular legitimacy? The role of establishing the truth,
punishing justly, and/or acting through just procedures. Albany Law Review,77(3), 1095–1137.
Walsh, K., Hussemann, J., Flynn, A., Yahner, J., & Golian, L. (2017). Estimating the prevalence of wrongful con-
victions. Washington: Office of Justice Programs, National Criminal Justice Reference Service. Retrieved from
https://www.ncjrs.gov/pdffiles1/nij/grants/251115.pdf
Wells, G. L., Kovera, M. B., Douglass, A. B., Brewer, N., Meissner, C., & Wixted, J. T. (2018). Policy and proce-
dure recommendations for the collection and preservation of eyewitness identification evidence. APLS.Retrieved
from http://ap-ls.wildapricot.org/resources/Documents/Scientific_Review_Paper/APLS%20Scientific%20Review%
20Paper%20initial%20draft%20July%2030%202018.pdf
Wells, G. L., Small, S., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe, C. A. E. (1998). Eyewitness identifi-
cation procedures: Recommendations for lineups and photospreads. Law and Human Behavior,22, 1–39.
Wells, G. L., Steblay, N. K., & Dysart, J. E. (2012). Eyewitness identification reforms: Are suggestiveness-induced hits
and guesses true hits? Perspectives on Psychological Science,7, 264–271.
West, E., & Meterko, V. (2016). Innocence Project: DNA exonerations, 1989–2014: Review of data and findings from
the first 25 years. Albany Law Review,79, 717–795.
Westervelt, S. D., & Cook, K. J. (2012). Life after death row: Exonerees’ search for community and identity.New
Brunswick: Rutgers University Press.
Zalman, M. (2014). Measuring wrongful convictions. In G. Bruinsma & D. L. Weisburd (Eds.), Encyclopedia of crimi-
nology and criminal justice.NewYork:Springer.
Zalman, M., Smith, B., & Kiger, A. (2008). Officials’ estimates of the incidence of “actual innocence” convictions,
Justice Quarterly,25, 72–100.
AUTHOR BIOGRAPHIES
Robert J. Norris is an assistant professor in the Department of Criminology, Law, and Society at
George Mason University. His interests include social change and policy reform, public opinion,
and legal decision-making, with a specific focus on wrongful convictions. He is the author of Exon-
erated: A History of the Innocence Movement (NYU Press, 2017) and co-author of When Justice
Fails: Causes and Consequences of Wrongful Convictions (Carolina Academic Press, 2018). His
recent articles have appeared in such outlets as Political Research Quarterly; Psychology, Public
Policy, and Law; Criminal Justice Policy Review; and the Journal of Experimental Criminology.
18 NORRIS ET AL.
Jennifer N. Weintraub is a doctoral candidate in the School of Criminal Justice at the University
at Albany, SUNY. Her research interests are broadly characterized within miscarriages of justice,
in particular, regarding wrongful convictions, legal decision-making, and policymaking.
James R. Acker is a Distinguished Teaching Professor in the School of Criminal Justice, University
at Albany, SUNY. His research and scholarship are focused on legal issues in criminal justice,
including wrongful convictions, capital punishment, criminal law, criminal procedure, and juvenile
justice. He is co-author, with Allison D. Redlich, of Wrongful Conviction: Law, Science, and Policy
(Carolina Academic Press, 2nd ed., 2019).
Allison D. Redlich is a professor in the Department of Criminology, Law and Society at George
Mason University. Her research is focused on miscarriages of justice, including wrongful convic-
tions (false confessions and false guilty pleas) and violations of due process (factors influencing
knowing, intelligent, and voluntary legal decision-making). Her recent articles have appeared in
such outlets as American Psychologist, Law and Human Behavior, Psychology, Public Policy and
Law, and the Journal of Experimental Criminology.
Catherine L. Bonventre is an assistant professor of justice and policy studies at Guilford College.
Her research interests include criminal justice policy specifically focusing on wrongful convictions;
crime laboratories, and judicial decision-making and policy implementation. She is a co-editor with
Allison D. Redlich, James R. Acker, and Robert J. Norris of Examining Wrongful Convictions:
Stepping Back, Moving Forward (Carolina Academic Press, 2014), as well as a co-author with
Robert J. Norris and James R. Acker of When Justice Fails: Causes and Consequences of Wrongful
Convictions (Carolina Academic Press, 2018).
How to cite this article: Norris RJ, Weintraub JN, Acker JR, Redlich AD, Bonventre CL.
The criminal costs of wrongful convictions: Can we reduce crime by protecting the innocent?
Criminol Public Policy. 2019;1–22. https://doi.org/10.1111/1745-9133.12463
APPENDIX: NAMES OF EXONEREES AND CORRESPONDING TRUE
PERPETRATORS
Exoneree(s) (By Case) True Perpetrator(s)
1. Adams, Kenneth
2. Gray, Paula 1. Johnson, Ira
3. Jimerson, Verneal 2. Robinson, Arthur
4. Rainge, Willie 3. Rodriguez, Juan
5. Williams, Dennis
6. Alexander, Richard 4. Murphy, Michael
7. Anderson, Marvin 5. Lincoln, John
8. Arledge, Randolph 6. Sims, David
NORRIS ET AL.19
Exoneree(s) (By Case) True Perpetrator(s)
9. Avery, Steven 7. Allen, Gregory
10. Avery, William D. 8. Ellis, Walter
11. Ott, Chaunte
12. Barbour, Bennett 9. Glass, James Moses, Jr.
13. Barr, Jonathan
14. Harden, James
15. Sharp, Shainne 10. Randolph, Willie B.
16. Taylor, Robert
17. Veal, Robert Lee
18. Bibbins, Gene 11. Gordon, Emanuel
19. Bloodsworth, Kirk 12. Ruffner, Kimberley Shay
20. Booker, Donte L. 13. Adams, Pettis, Jr.
21. Bradford, Marcellius
22. Ollins, Cavlin 14. Harris, Eddie "Bo"
23. Ollins, Larry 15. Roach, Duane
24. Saunders, Omar
25. Brewer, Kennedy 16. Johnson, Justin
26. Brooks, Levon
27. Briscoe, Johnny 17. Smith, Larry
28. Brown, Danny 18. Preston, Sherman
29. Brown, Keith 19. Mosley, Samuel
30. Brown, Patrick 20. Rickard, Robert
31. Brown, Roy 21. Bench, Barry Allen
32. Capozzi, Anthony 22. Sanchez, Altemio
33. Charles, Clyde 23. Arnold, Floyd Antonio
34. Clark, Robert 24. Charles, Marlo
35. Cole, Timothy 25. Johnson, Jerry Wayne
36. Cotton, Ronald 26. Poole, Bobby
37. Cruz, Rolando 27. Dugan, Brian
38. Hernandez, Alejandro
39. Dail, Dwayne Allen 28. Neal, William Jackson
40. Davis, Cody 29. Prichard, Jeremy
41. Davis, Jeramie Ray 30. Davila, Julio
42. Daye, Frederick 31. Pringle, David
43. Dean, James
44. Gonzalez, Kathy
45. Shelden, Debra 32. Smith, Bruce Allen
46. Taylor, Ada JoAnn
47. White, Joseph
48. Winslow, Thomas
20 NORRIS ET AL.
Exoneree(s) (By Case) True Perpetrator(s)
49. Deskovic, Jeffrey 33. Cunningham, Steven
50. Dewey, Robert 34. Thames, Douglas
51. Edwards, James 35. Whitfield, Hezekiah D.
52. Elkins, Clarence 36. Mann, Earl
53. Erby, Lonnie 37. Moore, Johnnie
54. Fears, Joseph Jr. 38. Knightona
55. Frey, Joseph 39. Crawford, James E.
56. Fritz, Dennis 40. Gore, Glenn
57. Williamson, Ron
58. Giles, James 41. Brown, Michael
42. Bryant, Stanley
59. Gray, Anthony Jr. 43. Fleming, Anthony Gerald
60. Green, Kevin 44. Parker, Gerald
61. Green, Michael 45. Rhines, Rodney
46. Elder, David
62. Green, Michael Anthony 47. Mosley, Lawrence
48. Smith, Michael A
49. Washington, Timothy
63. Halsey, Byron 50. Hall, Clifton
64. Hayes, Travis 51. Love, Rondell
65. Matthews, Ryan
66. Haynesworth, Thomas 52. Davis, Leon W., Jr.
67. Holland, Dana 53. Bolden, Gordon
68. Hunt, Darryl 54. Brown, Willie Elliot
69. Ireland, Kenneth 55. Benefield, Kevin M.
70. Jackson, Raymond 56. Anderson, Frederick
71. Williams, James Curtis 57. Sayles, Marion Doll
72. Jackson, Willie 58. Jackson, Milton
73. Johnson, Anthony 59. Brown, Matthew
74. Johnson, Rickey 60. McNeal, John C.
75. Jones, David Allen 61. Turner, Chester
76. Jones, Joe 62. Russell, Joel L
77. Karage, Entre Nax 63. Jordan, Keith
78. Kelly, William 64. Miller, Joseph
79. Krone, Ray 65. Phillips, Kenneth
80. Lowery, Eddie 66. Brewer, Daniel Lee
81. Lyons, Marcus 67. Anderson, Carl B.
82. McSherry, Leonard 68. Valdespino, George
NORRIS ET AL.21
Exoneree(s) (By Case) True Perpetrator(s)
83. Jenkins, Jerry Lee 69. Derr, Norman Bruceb
84. McCray, Antron
85. Richardson, Kevin
86. Salaam, Yusef 70. Reyes, Matias
87. Santana, Raymond
88. Wise, Korey
89. McGee, Arvin 71. Alberty, Edward
90. McGowan, Thomas 72. Woodson, Kenneth
91. McMillan, Clark 73. Boyd, David Louis
92. Mercer, Michael 74. Brown, Arthur
93. Miller, Jerry 75. Taylor, Lawrence
94. Miller, Neil 76. Weeks, Robert
95. Miller, Robert 77. Lott, Ronald
96. Moore, Curtis Jasper 78. Pope, Thomas, Jr.
97. Morton, Michael 79. Norwood, Mark Alan
98. Mumphrey, Arthur 80. Mumphrey, Charles
81. Thomas, Steve
99. Nelson, Bruce 82. Moore, Terrence
100. Ochoa, Chris 83. Marino, Achim Josef
101. Danziger, Richard
102. Ochoa, James 84. McCollum, James T.
103. Patterson, Maurice 85. Starkey, James
104. Phillips, Steven 86. Goodyear, Sidney Alvin
105. Pierce, Jeffrey 87. May, Omar D., Jr.
106. Pope, David Shawn 88. Roberts, James Milton
107. Powell, Anthony 89. Dixon, Jerry
108. Rachell, Ricardo 90. Hawthorne, Andrew Wayne
109. Richardson, Harold
110. Saunders, Michael
111. Swift, Terrill 91. Douglas, Johnny
112. Thames, Vincent
113. Roman, Miguel 92. Miranda, Pedro
114. Ruffin, Larry
115. Dixon, Bobby Ray 93. Harris, Andrew
116. Bivens, Phillip
117. Ruffin, Julius 94. Doxie, Aaron
118. Whitfield, Arthur Lee
119. Scott, Calvin Lee 95. Sauls, Steven Wayne
120. Smith, Frank Lee 96. Mosley, Eddie
121. Townsend, Jerry
122. Sterling, Frank 97. Christie, Mark J
123. Stinson, Robert Lee 98. Price, Moses
22 NORRIS ET AL.
Exoneree(s) (By Case) True Perpetrator(s)
124. Sutton, Josiah 99. Young, Donnie Lamont
125. Taylor, Ronald Gene 100. Carroll, Roosevelt
126. Thurman, Phillip Leon 101. N/Ac
127. Tillman, James 102. Foster, Duane
128. Vasquez, David 103. Spencer, Timothy
129. Waller, Patrick 104. Bell, Byron
105. Simmons, Lemondo
130. Warney, Douglas 106. Johnson, Eldred
131. Washington, Calvin 107. Carrol, Bennie
132. Washington, Earl 108. Tinsley, Kenneth Maurice
133. Waters, Leo 109. Caulk, Joel Bill
134. Watkins, Jerry 110. McCormick, Joseph
111. Munson, Kenneth
135. Webster, Bernard 112. Powell, Darren Lyndell
136. White, John Jerome 113. Parham, James Edward
114. Pickens, Lacy
137. Wilcoxson, Robert 115. Rutherford, Robert
138. Kagonyera, Kenneth 116. Summey, Bradford F
139.Williams, Willie 117.Wicker, Kenneth
140. Willis, John 118. McGruder, Dennis
141. Woodall, Glen 119. Good, Donald
142. Wyniemko, Kenneth 120. Gonser, Craig Hamilton
143. Youngblood, Larry 121. Cruise, Walter
aFirst name unavailable.
bDerr was linked to another crime for which a different person, Michael McAlister, was wrongly convicted in Virginia. McAlister,
however, was not exonerated through DNA testing, so he is not listed as an exoneree here.
cNo data available on name of true perpetrator.
