Combating Contamination in Confession Cases

[أحمد أبو زنط]

Combating Contamination in Confession Cases


Laura H. Nirider, Joshua A. Tepfer, & Steven A. Drizin




Convicting the Innocent: Where Criminal
Prosecutions Go Wrong
Brandon L. Garrett. Harvard, 2011. Pp 1, 367.
INTRODUCTION
“Sometimes eyewitnesses make mistakes. Snitches tell lies.
Confessions are coerced or fabricated. Racism trumps the truth. Lab
tests are rigged. Defense lawyers sleep. Prosecutors lie.” So declared
Barry Scheck, Peter Neufeld, and Jim Dwyer in their pathbreaking
2000 treatise on wrongful convictions, Actual Innocence: When
Justice Goes Wrong and How to Make It Right.1 As had never been
done before, Actual Innocence presented story after story of
wrongful convictions (and near executions) of the indisputably
innocent, with each chapter devoted to exposing each of these flaws
in the justice system. Actual Innocence was nothing short of a
revelation, a wake-up call concerning the reality of wrongful
convictions and the truth-telling power of DNA evidence. It was not
merely descriptive; it was also prescriptive, setting out a lengthy
recipe of reforms needed to prevent future wrongful convictions.
University of Virginia law professor Brandon L. Garrett’s
extraordinary new book—Convicting the Innocent: Where Criminal
† Adjunct Professor, Northwestern University School of Law; Project Co-Director,
Center on Wrongful Convictions of Youth, Bluhm Legal Clinic, Northwestern University
School of Law.
†† Clinical Assistant Professor of Law, Northwestern University School of Law; Project
Co-Director, Center on Wrongful Convictions of Youth, Bluhm Legal Clinic, Northwestern
University School of Law.
‡ Clinical Professor of Law, Northwestern University School of Law; Legal Director,
Center on Wrongful Convictions; Cofounder, Center on Wrongful Convictions of Youth;
Associate Director, Bluhm Legal Clinic, Northwestern University School of Law.
The authors coteach a clinical course on wrongful convictions of youth and have worked
together as cocounsel on numerous cases of false and contaminated confessions.
1 Jim Dwyer, Peter Neufeld, and Barry Scheck, Actual Innocence: Five Days to
Execution and Other Dispatches from the Wrongfully Convicted xv (Doubleday 2000). An
expanded version was published in 2001. Barry Scheck, Peter Neufeld, and Jim Dwyer, Actual
Innocence: When Justice Goes Wrong and How to Make It Right xx (Signet 2001).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
838 The University of Chicago Law Review [79:837
Prosecutions Go Wrong—picks up the honorable mantle that has so
long been carried by Actual Innocence (pp 5–13). In the twelve years
since Actual Innocence was first published, the aggregate number of
DNA exonerations has more than quadrupled from 67 to 289.2 In
Convicting the Innocent, Garrett analyzes the first 250 of these DNA
exonerations. This much larger data set, coupled with Garrett’s
academically rigorous method of analysis, has enabled him to make
the strongest case yet both for the reforms identified in Actual
Innocence and for additional reforms to the justice system.
Part I of this Review examines Garrett’s comprehensive new
work with an eye to the contributions that it makes to the field of
criminal defense and, in particular, to the critical subfield of
innocence work. Because one problem that Garrett highlights—the
fascinating problem of false confessions—has been the focus of our
own casework and study for many years,3 Part II follows by drawing
special attention to Garrett’s theories and real-world findings
concerning the contamination of confessions. Part III concludes by
presenting a practitioners’ guide for defense attorneys who are
confronted with contaminated confessions. Using one of our own
clients’ confessions as an example, we offer techniques, tactics, and
legal theories that can be used to attack the credibility of—or even
argue for the wholesale suppression of—confessions that are
contaminated and unreliable.
I. REVIEW OF CONVICTING THE INNOCENT
Garrett has designed his book much like Actual Innocence: by
focusing each individual chapter on one particular breed of error
that can cause wrongful convictions. To seasoned criminal justice
practitioners, his book presents a familiar list of errors: eyewitness
misidentifications (pp 45–83), flawed or misleading forensic science
(pp 84–117), false confessions (pp 14–44), jailhouse snitches (pp 118–
44), overzealous or sometimes unethical prosecutors (pp 167–71),
and inadequate or underfunded defense attorneys (pp 165–67). All
of these failures are well represented within the corpus of DNA
exonerations—if not as direct causes of wrongful convictions, then at
2 Innocence Project, Know the Cases: Innocence Project Case Profiles, online at
http://www.innocenceproject.org/know/ (visited Apr 2, 2012); Jim Dwyer, Peter Neufeld, and
Barry Scheck, Five Days to Execution at xiv (cited in note 1).
3 See, for example, Joshua A. Tepfer, Laura H. Nirider, and Lynda M. Tricarico,
Arresting Development: Convictions of Innocent Youth, 62 Rutgers L Rev 887, 904–08 (2010);
Steven A. Drizin and Richard A. Leo, The Problem of False Confessions in the Post-DNA
World, 82 NC L Rev 891, 907–23 (2004).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 839

least as significant factors that contribute to and perpetuate them. In
Convicting the Innocent, we learn, for example, that eyewitness
identification errors were present in 76 percent of the 250
exonerations that Garrett studied (p 48). Of the 74 percent of
exonerations involving forensic evidence, moreover, the great
majority involved invalid or unreliable evidence (pp 89–90), whereas
the use of snitch testimony and false confessions respectively
contributed to 21 percent and 16 percent of them (pp 18, 124).
Garrett’s statistical breakdown of wrongful conviction causes is
thorough and abundantly useful—and it becomes particularly
interesting when one compares his statistics to those from circa 2001
that are presented in Actual Innocence4—but Scheck, Neufeld, and
Dwyer have mined this terrain before, and similar numbers can be
obtained from the Innocence Project’s website.5 For this reason, the
true legacy of Garrett’s work—and what separates it from all
previous scholarship in this area—is the painstaking research Garrett
did before he ever put pen to paper. With admirable ambition,
Garrett immersed himself in each case of wrongful conviction by
reviewing police reports, confession statements, judicial opinions, lab
reports, and, most importantly, court transcripts whenever possible.
Not content only to scour trial-level records, Garrett also reviewed
pleadings and decisions from direct appeals, state postconviction
proceedings, and federal habeas corpus proceedings, becoming as
familiar as possible with the life cycle of each of these 250 cases
(p 184).
4 In the second Appendix to Actual Innocence, Scheck, Neufeld, and Dwyer quantified
the various sources of error in the first seventy-four DNA exoneration cases. They concluded
that 81 percent of those wrongful convictions were the result of mistaken identifications,
19 percent were due to snitch or informant testimony, and 22 percent were attributed to false
confessions. Scheck, Neufeld, and Dwyer, Actual Innocence at 361 (cited in note 1) (providing
graphical data as to the sources of error in seventy-four wrongful convictions). In subsequent
editions, as the authors gathered more data, the percentages differed only slightly. For
example, in the 2003 edition of Actual Innocence, the authors crunched the numbers for 130
DNA exonerations and came up with the following approximate percentages: 78 percent were
the result of mistaken identifications, 16 percent were due to snitch or informant testimony,
and 27 percent were attributed to false confessions. Barry Scheck, Peter Neufeld, and Jim
Dwyer, Actual Innocence: When Justice Goes Wrong and How to Make It Right 365 (New
American Library 2003).
5 Innocence Project, Understand the Causes: The Causes of Wrongful Conviction, online
at http://www.innocenceproject.org/understand/ (visited Nov 15, 2011) (identifying the
contributing causes of wrongful convictions as 77 percent eyewitness misidentification,
52 percent unvalidated/improper forensics, 23 percent false confessions/admissions, and
16 percent informants/snitches, and noting that the total is greater than 100 percent because
wrongful convictions can have more than one cause).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
840 The University of Chicago Law Review [79:837
Garrett’s unprecedented level of familiarity with these cases
enables him to tell the exonerees’ stories with a richness of detail
missing from Actual Innocence. By presenting these gripping tales of
justice gone terribly wrong, Garrett absorbs the reader by speaking
at once of villains who falsify evidence, intimidate witnesses, and
abuse suspects, and of innocents victimized by the law who win back
their freedom after fighting for decades. Indeed, Garrett’s ability to
tell so many previously untold stories allows him to vividly
demonstrate how the protections afforded criminal defendants in our
adversarial system of justice—the privilege against selfincrimination,
6 the rights to the effective assistance of counsel7 and to
a jury trial,8 the requirement that prosecutors prove defendants guilty
beyond a reasonable doubt, the ability to access appellate review9—
have failed again and again to protect the innocent.
For example, in his chapter entitled “Judging Innocence,”
Garrett asks a simple question that—perhaps surprisingly—has
never been asked before: “Why was it so hard for these innocent
people to challenge their flawed convictions?” (p 182) The demon
here, it turns out, is the doctrine of “harmless error,” a bedrock legal
principle created in 1967 by the United States Supreme Court that
permits appellate courts to overlook errors of constitutional
magnitude when the errors did not contribute to the guilty verdict at
trial.10 When properly applied, the harmless error doctrine requires
the reviewing court to assess only whether an error affected the
outcome of the trial beyond a reasonable doubt.11 Reviewing courts,
however, too often contort the meaning of this test and apply one
that is far different—one that asks whether the evidence of guilt in
the record outweighs the impact of the constitutional error.12
Garrett’s findings reveal that appellate courts’ misapplication of this
doctrine has generated a terrible track record when it comes to
assessing evidence of guilt or innocence. Although we now know that
the defendants in the cases Garrett studied were actually innocent,
reviewing courts—often acting on the basis of a mistaken belief that
6 See Murphy v Waterfront Commission of New York Harbor, 378 US 52, 55 (1964).
7 See Strickland v Washington, 466 US 668, 684–86 (1984).
8 See Duncan v Louisiana, 391 US 145, 149, 155 (1968).
9 See Griffin v Illinois, 351 US 12, 18 (1956).
10 See Chapman v California, 386 US 18, 22 (1967).
11 See id at 23–24.
12 See Harry T. Edwards, To Err Is Human, but Not Always Harmless: When Should
Legal Error Be Tolerated?, 70 NYU L Rev 1167, 1170–72 (1995) (noting that appellate judges’
“natural inclination is to view an error as harmless whenever a defendant’s conviction appears
well justified by the record evidence,” and describing the application of the harmless-error
doctrine as the “guilt-based approach”).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 841

the defendants were guilty—routinely classified the defendants’
meritorious claims of constitutional error as harmless and denied
them relief. For those exonerees whose cases included written
appellate decisions, courts found “harmless error” 30 percent of the
time when affirming their convictions—many of them driven to such
a conclusion by a flawed belief in the defendant’s guilt (p 201).
Similarly flawed assessments of guilt caused appellate courts to
routinely dismiss defendants’ claims of ineffective assistance of
counsel and claims that prosecutors failed to disclose exculpatory
evidence to the defense. In both situations, reviewing courts are
required to determine if the defendant was prejudiced by the errors
of trial counsel or the prosecution—an analysis similar to the
harmless error analysis in that it asks whether the errors of trial
counsel or prosecutors resulted in a different outcome for the
defendant.13 Eighteen percent of those exonerees whose cases
included written appellate opinions had their appeals rejected
because they did not suffer prejudice as a result of trial errors
(p 201).
Perhaps most significantly, appellate judges overtly based
denials of relief on the defendants’ likely guilt in a full 47 percent of
the cases Garrett studied; in 10 percent of the cases, appellate judges
were so certain of the defendants’ guilt that they described the state’s
evidence as “overwhelming” (pp 201–02). This alone is powerful
evidence that the appellate court process does not protect the
innocent who are mistakenly convicted. Indeed, a full 62 percent of
the cases Garrett studied involved appellate judges who were so
moved by the apparent evidence of guilt that they commented on
guilt, deemed an error harmless, or found a lack of prejudice (p 202).
If there is one weakness in Garrett’s otherwise illuminating
analysis of the appellate court system, it is that he is short on
solutions that will make the criminal review process a more effective
safety net for the wrongfully convicted. Garrett could and perhaps
should have joined the chorus of commentators who seek to limit the
scope of the harmless error rule, especially in cases of involuntary
confessions.14 Indeed, before 1991, whenever a reviewing court
13 See Strickland, 466 US at 691–96. See also Brady v Maryland, 373 US 83, 86–87 (1963).
14 See, for example, Alan Hirsch, Confessions and Harmless Error: A New Argument for
the Old Approach, 12 Berkeley J Crim L 1, 26–27 (2007); Charles J. Ogletree Jr, Arizona v.
Fulminante: The Harm of Applying Harmless Error to Coerced Confessions, 105 Harv L
Rev 152, 165–66 (1991) (“[C]oerced confessions generally are not susceptible to harmless error
analysis because of the overwhelming, prejudicial effect such confessions have on jurors’
beliefs. . . . The trial becomes skewed, and appellate courts cannot meaningfully apply harmless
error analysis.”).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
842 The University of Chicago Law Review [79:837
determined that an involuntary confession had been used against a
defendant at trial, it was required to reverse the defendant’s
conviction—period.15 The admission of a coerced confession at trial
was believed to be so prejudicial that it could never be truly
harmless.16 In the 1991 case Arizona v Fulminante,17 however, the
Supreme Court ruled that a trial court’s admission of an involuntary
confession was simply another error to be evaluated under the
harmless error rule.18 Fulminante was harshly criticized by numerous
commentators who feared that the decision would encourage
detectives to coerce confessions and that jurors, in turn, would
convict defendants on the basis of unreliable confessions.19 These
arguments have only grown stronger since the discovery of so many
DNA-proven false confessions post-Fulminante.20 In this book, at
least, Garrett may have missed a golden opportunity to use his
powerful statistics and stories to call for the Supreme Court to revisit
Fulminante.
At any rate, perhaps the most haunting takeaway from Garrett’s
study is that after poring over 250 cases of wrongful conviction, he
concludes that “[t]he unremarkable criminal trials of so many of
these exonerees looked no different than those of countless others”
(p 263). When the truth of this sentence washes over the reader, the
implications are gut-wrenching. Both the common use of
untrustworthy evidence (misidentifications, false confessions, and
flawed forensics alike) and the system-wide failure of actors in the
criminal justice system (defense attorneys, prosecutors, and judges
alike) must necessarily occur in many cases in which evidence with
the power to exonerate, like DNA, is irrelevant, unavailable, or
ignored. Without irony, Garrett forces the reader to conclude that
the DNA exonerees he studied, who spent on average thirteen years
incarcerated for other people’s crimes (p 5), were the lucky ones.
Many other innocent people are living behind bars with no real hope
of ever proving their innocence.
Garrett’s own data compels this inescapable conclusion. The
DNA exonerees in Garrett’s study were overwhelmingly accused
15 See, for example, Blackburn v Alabama, 361 US 199, 205 (1960).
16 See id at 210.
17 499 US 279 (1991).
18 Id at 310.
19 See, for example, Hedieh Nasheri and Victor J. DeMarco, True Confessions?
A Critique of Arizona v. Fulminante, 21 Am J Crim L 273, 283–84 (1994); Ogletree, 105 Harv L
Rev at 161–72 (cited in note 14).
20 See Brandon L. Garrett, The Substance of False Confessions, 62 Stan L Rev 1051, 1052
& n 2 (2010).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 843

and convicted of raping a stranger: 68 percent were convicted of rape
alone, and an additional 21 percent were convicted of rape-murders
(p 278). Why do so many exonerations occur in stranger-rape cases?
The answer is simply that DNA evidence is most often available, and
most probative, in these cases. In the overwhelming majority of
other cases, especially cases involving lesser crimes or property
offenses, DNA is unavailable; and if it is available, it is frequently
deemed less probative of a suspect’s guilt. But there is no reason to
believe that the factors contributing to wrongful convictions are any
less likely to occur in other kinds of cases.
Even if one isolates the small number of serious cases in which
DNA-testable evidence might be available, it is still clear that the
250 cases of wrongful conviction in Garrett’s book are just the tip of
the iceberg. Many times, probative DNA evidence is lost or
misplaced by the state; other times, the evidence becomes degraded
due to poor storage conditions or is collected by evidence technicians
in quantities insufficient for effective testing.21 Further, despite the
fact that sixteen individuals in Garrett’s dataset pleaded guilty
(p 278), some states explicitly bar defendants who have pleaded
guilty from subsequently testing the evidence in their case.22 In the
vast majority of states, it is unsettled whether a defendant who has
pleaded guilty can test the evidence in his case, because the
applicable statutes are ambiguous.23 Given that 97 percent of federal
convictions and 94 percent of state convictions are the result of guilty
pleas,24 a significant portion of the convicted population lacks
procedural access to testing.
Other categories of convicted defendants not mentioned by
Garrett may lack access to DNA testing as well. For instance, there
is not a single example in Garrett’s book of a child convicted in
juvenile court who later was exonerated by DNA testing, despite the
fact that juvenile courts may be troublingly predisposed to find the
innocent guilty.25 Some jurisdictions have not granted their children
21 See Cynthia E. Jones, The Right Remedy for the Wrongly Convicted: Judicial Sanctions
for Destruction of DNA Evidence, 77 Fordham L Rev 2893, 2900 (2009); James Herbie
DiFonzo, In Praise of Statutes of Limitations in Sex Offense Cases, 41 Houston L Rev 1205,
1266 (2004).
22 See, for example, Ohio Rev Code Ann § 2953.72(C)(2); 13 Vt Stat Ann § 5561(e);
People v O’Connell, 879 NE2d 315, 319 (Ill 2007).
23 Joshua A. Tepfer and Laura H. Nirider, Adjudicated Juveniles and Post-conviction
Litigation, 63 Maine L Rev *11–12 (forthcoming 2012), online at http://ssrn.com
/abstract=2013530 (visited Apr 18, 2012).
24 Missouri v Frye, 132 S Ct 1399, 1407 (2012).
25 See Steven A. Drizin and Greg Luloff, Are Juvenile Courts a Breeding Ground for
Wrongful Convictions?, 34 N Ky L Rev 257, 321–22 (2007); Tepfer, Nirider, and Tricarico,
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
844 The University of Chicago Law Review [79:837
the same access to postconviction remedies like DNA testing; and in
many other jurisdictions, the issue remains unsettled.26 Even beyond
the question of procedural access to the remedy, it is also
questionable whether such children have the resources or years-long
dedication required to fight for DNA testing.
Finally, even in less serious cases where DNA-testable evidence
could be available, it is rare that journalists, lawyers, or other
individuals ever put resources toward pursuing such relief. Without
this type of extraordinary help, it remains next to impossible for a
pro se litigant locked behind bars to get any court to listen to his
protestations of innocence.
To ensure a more accurate and cost-effective criminal justice
system, Garrett’s excellent list of proposed reforms must be
implemented. Video recording of police procedures, including
custodial interrogations and identification procedures, is an
important and manageable first step. Requiring eyewitnessidentification
procedures to follow best practices supported by a
mountain of social science research—including double-blind and
sequential lineups—is another logical step. And the creation of
independent crime labs—labs that are regularly audited and that are
disassociated from law enforcement—could solve many of the
forensic-fraud issues that contributed to almost three-quarters of the
wrongful convictions in Garrett’s book. Finally, states would be well
served to follow the lead of North Carolina, which has recently
developed a unique body called the Actual Innocence Commission.
This independent commission, which includes law enforcement
officers, prosecutors, defense attorneys, social scientists, victim’s
advocates, law professors, and judges, proposed and recommended
innocence-driven reforms that have been adopted by the legislature.
The commission also developed an independent Innocence Inquiry
Commission that reviews individual cases of innocence and, if
warranted, recommends them for judicial review (pp 241–43).27
62 Rutgers L Rev at 893 (cited in note 3). See also Welch v United States, 604 F3d 408, 432
(7th Cir 2010) (Posner dissenting).
26 Compare 11 Del Code Ann § 4504(a) with 10 Del Code Ann § 1009(h). See also Ala
Code §§ 15-18-200, 15-19-7; In re William M., 795 NE2d 269, 274 (Ill 2003). Tepfer and Nirider,
63 Maine L Rev at *7–8 (cited in note 23).
27 Notably, the North Carolina Innocence Inquiry Commission allows defendants who
pleaded guilty to apply for relief. In fact, on September 22, 2011, the Commission exonerated
Robert Wilcoxson and Kenneth Kagonyera, two men who had each spent almost eleven years
in prison for the murder of Walter Rodney Bowman. Both Kagonyera and Wilcoxson pleaded
guilty to the Bowman murder; DNA and other evidence later linked other men to the murder.
See Jon Ostendorff, Judges Free 2 Men in Innocence Review, USA Today 3A (Sept 23, 2011).
See also Editorial, Scrutinizing Justice, St Petersburg Times P3 (Dec 13, 2009) (lauding the
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 845

Until these and other reforms are instituted, however,
something else can be done: prosecutors, judges, and even defense
attorneys can adjust their mind-sets to acknowledge and account for
the reality of wrongful convictions. Such an adjustment necessarily
includes not immediately discounting non-DNA evidence of
innocence. Because the vast majority of convicted defendants will
never have DNA to test, exculpatory evidence—such as a
recantation from a crucial witness or new information about an
alternative suspect—should be carefully scrutinized. At bare
minimum, courts should order evidentiary hearings to determine the
reliability of this kind of new information from witnesses under oath,
appointing counsel for indigent defendants.
Sadly, we appear to be very far away from this kind of progress.
One of the most disheartening portions of Garrett’s book is his
eighth chapter, entitled “Exoneration.” Here, Garrett details the
“[l]ong [r]oad to [e]xoneration” faced by the exonerees he studied,
even after DNA evidence proved their innocence (pp 215–17). In too
many of the 250 cases, prosecutors and courts often resisted efforts
even to test available and probative evidence (p 216). Equally
troublingly, these 250 individuals had to wait on average over a full
calendar year to become officially exonerated after postconviction
DNA results conclusively proved their innocence (p 216).
II. THE CONTAMINATION OF CONFESSIONS
The first systemic problem underlying wrongful convictions that
Garrett identifies in Convicting the Innocent is the problem of false
confessions (p 8). Given that the number of wrongful convictions
based on false confessions pales in comparison with those involving
eyewitness identification errors, flawed forensics, or even jailhouse
snitches (p 279), his choice to begin with a chapter on false
confessions is bold. It is a wise choice, however, given that the
concept is absolutely foreign to so many. Many jurors—and many
members of the general public and even of the legal profession—
struggle profoundly with the idea that a person would ever confess to
a crime that he or she did not commit.28 It is for this reason that—
creation of a Florida organization modeled after North Carolina’s Actual Innocence
Commision).
28 See Colorado v Connelly, 479 US 157, 182 (1986) (Brennan dissenting) (observing that
“[t]riers of fact accord confessions such heavy weight in their determinations that ‘the
introduction of a confession makes the other aspects of a trial in court superfluous’”); Miranda
v Arizona, 384 US 436, 466 (1966), citing Mapp v Ohio, 367 US 643, 685 (1961) (Harlan
dissenting) (explaining that a confession is “the most compelling possible evidence of guilt”);
Hopt v Utah, 110 US 574, 584–85 (1884) (recognizing that a “voluntary confession of guilt is
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
846 The University of Chicago Law Review [79:837
although false confessions may be less prevalent than other sources
of error—the road to exoneration for false confessors is often
exceptionally long and exceptionally arduous.
Garrett’s analysis and findings concerning false confessions are
nothing short of groundbreaking. At the beginning of his foray into
the case materials, Garrett expected that the DNA exonerees’
confessions would lack detail and be riddled with errors (pp 18–19).
Stunningly, he found just the opposite: in thirty-eight of the forty
false confessions he studied, the confessions were detailed and often
factually accurate descriptions of the criminal acts (pp 19–20). If
these men are truly innocent, Garrett asks the reader, how is it that
they were able to give such detailed and accurate confessions? His
answer is that their DNA-proven false confessions were
“contaminated” forms of evidence—as tainted and unreliable as
contaminated physical evidence.
What is contamination? Many who watch television crime
dramas like Law & Order or CSI probably associate the notion of
contamination with evidence mishandling at a crime scene or in a
crime lab. In this context, contamination is defined as the “unwanted
transfer of material from another source to a piece of physical
evidence.”29 To prevent contamination, state technicians usually
create a perimeter around a crime scene using yellow tape of the
type made familiar in television and movies. Other standard
precautions include identifying a path of entry into the crime scene
and an exit path out of it, preventing unauthorized individuals from
accessing the crime scene, and keeping a log of all those who enter
the scene.30 Police personnel who enter a crime scene must also wear
protective clothing—gloves, masks, eyewear, and footwear—both to
protect themselves from biological or chemical hazards and to
protect the evidence from contamination. In the crime lab, similarly,
technicians must wear masks, hairnets, and gloves to keep their own
hair, skin cells, or saliva off evidence samples. A simple sneeze or an
among the most effectual proofs in the law”). See also Saul M. Kassin and Katherine
Neumann, On the Power of Confession Evidence: An Experimental Test of the Fundamental
Difference Hypothesis, 21 L & Human Beh 469, 481 (1997); Gerald R. Miller and F. Joseph
Boster, Three Images of the Trial: Their Implications for Psychological Research, in Bruce
Dennis Sales, ed, Psychology in the Legal Process 19, 20–21 (Spectrum 1977).
29 National Institute of Justice, Office of Justice Programs, Crime Scene Investigation:
A Guide for Law Enforcement 42 (Department of Justice 2000), online at https://www.ncjrs.gov
/pdffiles1/nij/178280.pdf (visited Nov 16, 2011).
30 Id at 19–20.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 847

errant hair could compromise the evidence and jeopardize the
chances of conviction.31
In the confession context, contamination is the transfer of inside
information—nonpublic details about the crime that only the true
perpetrator could have known—from one person to another person
during a police investigation.32 The problem of contamination in false
confession cases usually arises during interrogation itself, when the
interrogator pressures a suspect to accept a particular account of the
crime story—one that usually squares with the interrogator’s
preordained theory of how the crime occurred. The interrogator
then uses leading questions, deliberately or inadvertently, to suggest
specific facts about the crime to the suspect, which are then parroted
back in the form of a confession. The presence of these types of
specific facts in the suspect’s confession lends it credibility and
creates an all-important illusion of corroboration.
Just as crime scene technicians are trained to avoid
contaminating crime scenes, police interrogators are trained to avoid
contaminating confessions. As Joseph Buckley, the president of the
leading police-interrogation training firm in the United States and
coauthor of the leading interrogation-training manual in the United
States, said in 2006:
[I]t is imperative that interrogators do not reveal details of the
crime so that they can use the disclosure of such information by
the suspect as verification of the confession’s authenticity. In
each case there should be documented “hold back” information
about the details of how the crime was committed; details from
the crime scene; details about specific activities perpetrated by
31 Christopher B. Mueller, Introduction: O.J. Simpson and the Criminal Justice System on
Trial, 67 U Colo L Rev 727, 734 (1996). When people of a certain generation think of
contamination, they think of the O.J. Simpson case. It was during this highly publicized trial
that defense lawyer Barry Scheck introduced the public to the problem of evidence
contamination when he attacked the collection procedures of the LAPD and the testing
procedures of the LAPD’s crime lab. Scheck carefully demonstrated how human error could
have compromised the reliability of the evidence, educating viewers on the importance of
changing gloves each time a piece of evidence is handled. Scheck also highlighted the “cesspool
of contamination” at the LAPD’s crime lab, demonstrating that reference vials containing
blood standards of both Nicole Brown Simpson and Ron Goldman contained DNA alleles that
came from O.J. Simpson. See William C. Thompson, DNA Evidence in the O.J. Simpson Trial,
67 U Colo L Rev 827, 832–40 (1996).
32 See Richard J. Ofshe and Richard A. Leo, The Decision to Confess Falsely: Rational
Choice and Irrational Action, 74 Denver U L Rev 979, 993 (1997).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
848 The University of Chicago Law Review [79:837
the offender; etc. The goal is to match the suspect’s confession
against these details to establish the veracity of the statement.33
Indeed, contamination is counterproductive, because it prevents
police from testing and corroborating the reliability of the
admissions and confessions they elicit. They are simply unable to
distinguish true from false confessions because both contain “guilty”
knowledge.34 More importantly, contamination can jeopardize public
safety by blinding police to the reality that the true perpetrator is still
on the streets. For these reasons, law enforcement trainers and
leaders almost universally agree that contamination has no
legitimate place in American interrogation.
The problem of police contamination in the interrogation room
triggers a number of events at trial that only increase the likelihood
of a wrongful conviction. In his review of the trial transcripts, Garrett
discovered a striking pattern. In 95 percent of the false confessions
he studied, Garrett found that detectives testified that it was the
defendants who offered key details about the crime (p 20). In many
cases, the detectives went a step further, testifying that they
scrupulously avoided leaking such details (p 20). The defendants’
knowledge of these nonpublic facts then became the crux of the
State’s case at trial. In twenty-two of the thirty-six cases for which
Garrett obtained trial transcripts, prosecutors recited the nonpublic
facts in their closing arguments as evidence of the defendants’ guilt
(p 31).
Garrett is not the first researcher to discuss the phenomenon of
confession contamination. Numerous researchers before him have
described both how contamination during interrogation causes
innocent suspects to incorporate accurate crime scene details into
their confessions, and how jurors can be misled into thinking that
defendants are guilty because their confessions appear to be
corroborated.35 Garrett’s major contribution to this literature is his
33 Joseph P. Buckley, The Reid Technique of Interviewing and Interrogation, in Tom
Williamson, ed, Investigative Interviewing: Rights, Research, Regulation 190, 204–05 (Willan 2005).
34 As Dr. Richard Leo has noted, the failure of the police to ferret out false confessions on
the front lines of the investigative process sets in motion a stunning series of consequences for the
innocent. Suspects who confess are more likely to be charged, detained pending trial, pressured
by their attorneys to plead guilty, convicted at trial, and sentenced to serve more time. See
Richard A. Leo, Inside the Interrogation Room, 86 J Crim L & Criminol 266, 298–99 (1996). See
also Richard A. Leo, Police Interrogation and American Justice 248 (Harvard 2008).
35 See Garrett, 62 Stan L Rev at 1066–92 (cited in note 20); Leo, Police Interrogation
at 165–73 (cited in note 34); Richard A. Leo, et al, Bringing Reliability Back In: False
Confessions and Legal Safeguards in the Twenty-First Century, 2006 Wis L Rev 479, 511; Saul
M. Kassin, A Critical Appraisal of Modern Police Interrogations, in Williamson, ed,
Investigative Interviewing 207, 221–23 (cited in note 33); Gisli H. Gudjonsson, The Psychology
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 849

finding that the problem of contamination is epidemic, not episodic,
in cases of false confessions.
At bottom, confession contamination may be far more
dangerous than contaminated physical evidence at a crime scene or
in the crime lab. While forensic testing can reveal the unwanted
presence of a technician’s DNA on crime scene evidence, there is no
such foolproof test to identify confession contamination. For this
reason, we offer in the next Part some guidance for practitioners
concerning how to identify and combat the contaminated confession.
III. A PRACTITIONER’S GUIDE TO COMBATING CONTAMINATION
IN CONFESSION CASES
To facilitate our discussion of contaminated confessions, we will
use the confession of sixteen-year-old Brendan Dassey as a case
study.36 In 2006, Brendan was accused of helping his uncle sexually
assault and murder a young woman in Wisconsin.37 He was a specialeducation
student who had never before been in trouble with the
law—but that would change after a young woman who made a living
photographing cars for classified ads went missing. Suspicion fell on
Brendan’s uncle after police learned that she had last been seen at
his home photographing a vehicle that was for sale. That suspicion
was amplified when her abandoned SUV was found covered in
branches on Brendan’s uncle’s property with its license plates
missing—and it hit a fevered frenzy when her charred remains were
found in a bonfire pit behind his uncle’s garage.38 The media
swarmed over the gruesome story like ants on a juicy piece of fruit—
made all the more juicy, in this case, by the fact that Brendan’s uncle
was Steven Avery, Wisconsin’s first DNA exoneree. If the story of
Avery’s DNA-driven redemption had been trumpeted in the press,
his apparent subsequent fall from grace was blared at top volume.
of Interrogations and Confessions: A Handbook 178–79 (Wiley 2003); Richard A. Leo and
Richard J. Ofshe, The Consequences of False Confessions: Deprivations of Liberty and
Miscarriages of Justice in the Age of Psychological Interrogation, 88 J Crim L & Criminol 429,
438 & n 24, 439–40 (1998); Gisli H. Gudjonsson, The Psychology of Interrogations, Confessions
and Testimony 259 (Wiley 1992).
36 The authors are postconviction counsel for Brendan Dassey, who was convicted in
2007 of murder and is currently appealing his conviction. The facts relating to Brendan’s
confession are taken from Memorandum of Facts and Law Accompanying Section 809.30 Post-
Conviction Motion, State v Dassey, No 2006 CF 88 (Manitowoc County, Wis, filed Aug 21,
2009) (“Dassey Memorandum of Facts”), online at http://cwcy.org/resources/67_attach
_Memorandum%20of%20Facts%20and%20Law%20with%20Sign atures.pdf (visited Nov 16,
2011).
37 Id at *36–37.
38 Id at *2–3, 5.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
850 The University of Chicago Law Review [79:837
Nearly every fact about the investigation into the case was publicized
in television broadcasts and newspaper headlines across Wisconsin
and the country.39
After Steven was arrested, however, the case went cold very
quickly. Little additional evidence against him was found, despite
one of the largest investigations in Wisconsin history.40 Perhaps
frustrated by the unforgiving media spotlight, investigators
eventually turned to his sixteen-year-old nephew, Brendan Dassey,
who lived next door to Avery.41 Brendan was of interest in part
because he had told investigators that he had attended a bonfire at
his uncle’s house on the night that the victim went missing, although
he’d always said he saw nothing suspicious in the fire. Based on little
else, investigators interviewed him outside the presence of a parent
or attorney twice on February 27, 2006, and followed up with a fullblown
interrogation two days later, on March 1, 2006.42
After undergoing interrogation—which would later be
challenged postconviction as psychologically coercive43—Brendan
eventually began making inculpatory statements indicating not only
that he was at the bonfire, but that he had, in fact, seen a body in the
fire.44 When pushed by investigators, he eventually told a story in
which, while delivering the mail to his uncle’s house, he stumbled
upon Avery sexually assaulting the victim and—at Avery’s urging—
took part in the assault and subsequent murder.45
At this point in the interrogation, the investigators needed
Brendan to deliver the correct narrative regarding the victim’s
murder. Interrogators have long held, after all, that the key to
obtaining a convincing confession is to get the suspect to produce a
postadmission narrative that coincides with the facts of the crime.46
And while in the abstract it was plausible that the victim had been
sexually assaulted, there was no evidence of any such assault either
on her remains or in the uncle’s home.47 The only physical evidence
that investigators had been able to glean from her remains, in fact,
indicated that she had been shot in the head.48 While Brendan’s
39 See id at *6–8.
40 See Dassey Memorandum of Facts at *11–12 (cited in note 36).
41 Id at *1.
42 Id at *12–35.
43 See id at *112.
44 Id at Dassey Memorandum of Facts at *14–15 (cited in note 36).
45 Id at *23–27.
46 Fred E. Inbau, et al, Essentials of the Reid Technique: Criminal Interrogation and
Confessions 213–17 (Jones & Bartlett 2005).
47 Dassey Memorandum of Facts at *11–12 (cited in note 36).
48 See id at *5.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 851

account of sexual assault was good theater, the investigators needed
more to prove their murder case. They needed him to give a
statement that matched the physical evidence; they needed him to
say that she had been shot. The following exchange ensued.
Officer: What happens next? . . .
Brendan: Then he went in, back in there and he stabbed
her. . . .
Officer: We know he did something else to her, what else did
he do to her?
Brendan: He choked her. . . .
Officer: What else did he do to her? We know something else
was done. Tell us, and what else did you do? Come
on. Something with the head. Brendan? . . .
Brendan: Huh? . . .
Officer: What else did you guys do, come on. . . .
Brendan: That he cut off her hair. . . .
Officer: OK, what else? What else was done to her head?
Brendan: That he punched her.
Officer: What else? [pause]What else? . . . What did he make
you do to her? [pause] What did he make you do
Brendan? It’s OK, what did he make you do?
Brendan: Cut her.
Officer: Cut her where?
Brendan: On her throat. . . .
Officer: What else happens to her in her head? It’s
extremely, extremely important you tell us this, for
us to believe you. Come on Brendan, what else?
[pause] We know, we just need you to tell us.
Brendan: That’s all I can remember.
Officer: All right, I’m just gonna come out and ask you. Who
shot her in the head?
Brendan: He did.
Officer: Then why didn’t you tell us that?
Brendan: Cuz I couldn’t think of it.49
After obtaining this admission, the investigators went on to elicit
a number of other details from Brendan, including the fact that he
49 Calumet County Sheriff’s Department, Interview of Brendan Dassey, Complaint No
05-0157-955 *578–79, 584–87 (Mar 1, 2006), online at http://
convolutedbrian.com.s3.amazonaws.com/dassey/01Mar2006/01Mar2006Transcript.pdf (visited
Nov 15, 2011). The interrogation excerpts herein have been shortened. In addition, for
simplicity, the questions posed by two interrogators have been collapsed into one “Officer.”
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
852 The University of Chicago Law Review [79:837
and his uncle had cleaned up using a bottle of bleach from Avery’s
bathroom; the fact that he and his uncle had purportedly hidden the
SUV in a section of Avery’s property known as the “pit” area; and
the fact that the SUV’s license plates had been removed. These facts
were elicited in the following exchanges:50
Officer: [W]hat do you guys do next?
Brendan: Go take the jeep down in the pit.
Officer: Tell us how. . . .
Brendan: He drove it down there and . . . he put it back by the
trees and covered it with branches and a hood. . . .
Officer: OK. After [Avery] put the car there, what do you do
next?
Brendan: We walk out. . . .
Officer: [T]he license plates were taken off the car, who did
that?
Brendan: I don’t know.
Officer: Did you do that?
Brendan: [shakes head “no”] No.
Officer: Did [Avery] do that?
Brendan: Yeah. . . .
Officer: If we took you to that garage, would you be able to
show us where [you cleaned]?
Brendan: Yeah. . . .
Officer: Where’d you get the bleach from?
Brendan: In his house by, in his bathroom.
Brendan’s entire interrogation was captured on camera, per the
Wisconsin Supreme Court’s July 2005 ruling that required any
statements made by juveniles during interrogations to be
electronically recorded.51
How should a defense practitioner proceed when confronted
with a detailed confession like this one? The answer is simple. As
Garrett notes, most false confessions nonetheless include largely
correct statements about how the crime occurred due to
contamination (p 20). Therefore, a defense practitioner must review
the confession with an eye to answering one key question: was the
client ever able to produce any verifiably true facts about the crime,
absent contamination?
The first step in answering this question is to investigate the
confession for potential contamination. Contamination usually arises
50 Id at *600–02, 610.
51 See In re Jerrell C.J., 699 NW2d 110, 123 (Wis 2005).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 853

out of fact-feeding by police, as Garrett identifies (p 28), but it can
also result from a suspect’s exposure to media coverage, knowledge
of community gossip or rumors about the case, and independent
familiarity with the crime scene. Each of these sources should be
investigated in turn.
Examining whether police fed facts to a suspect is a simple task
that can be enormously fruitful—but it is not always possible. As
Garrett emphasizes, the review of an interrogation for fact-feeding is
almost impossible in the absence of an electronic recording of the
interrogation in its entirety. He points out that many jurisdictions
record only the suspect’s final statement, in which the suspect delivers
a story of guilt in response to investigators’ open-ended questions
(p 32)—a story that has often been rehearsed, sometimes for hours,
before the recorder was turned on (p 32). Such recordings are not
useful for the purpose of detecting contamination.52 Instead, it is
essential to have a recording of the entire process of interrogation
(pp 43, 247–48), including the all-important stage during which police
extract a postadmission narrative account of how the crime unfolded.
If the suspect cannot get the facts of the crime right during this stage,
police tend to begin questioning him in a leading style, perhaps
inadvertently and often out of frustration (pp 33–34). Those leading
questions, however, feed the suspect facts about the crime that can be
later parroted back (pp 33–34, 43–44). An inquiry into whether factfeeding
occurred, accordingly, requires a transcript of the
interrogation and a close and careful hunt for leading questions.
While most contamination takes the form of fact-feeding, other
sources of contamination should also be investigated. Whether a
suspect could have learned facts about the crime from media
coverage, of course, will only be a factor in relatively rare highprofile
cases. In those cases, however, a defense practitioner should
first learn the sources of media to which the client had been exposed
in the time period between the crime and his confession; and he
should make every effort to obtain copies of those news stories,
whether in the form of television news transcripts, newspaper
articles, blog postings, or police press releases. By reviewing those
52 The failure to record the interrogation process in its entirety has implications for trial.
If an interrogation is unrecorded, then the interrogator is free to assert that crime facts were
volunteered by the suspect, at which point the trial may well devolve into a swearing contest
between the suspect and the interrogators concerning who was the true source of the details in
the confession. If the entire process is recorded, however, then it may be possible to trace
contamination where it exists.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
854 The University of Chicago Law Review [79:837
news stories, the practitioner will have an understanding of what
information about the crime was public knowledge.
Even where the suspect cannot point to particular media stories
he himself viewed, the practitioner should nevertheless familiarize
himself with all information that was in the public domain. Even if he
cannot prove that the suspect read or viewed the source of media
that publicized a particular fact, other people that the suspect may
have known did. The information could have filtered back to the
suspect through those sources.
Similarly, it is also important for the practitioner to investigate
whether the client could have learned details of the crime from
rumors or gossip. This avenue of inquiry, of course, is particularly
fruitful in cases that generate buzz in the client’s own circle of
acquaintances—in cases, for instance, in which an involved party was
a classmate, family member, coworker, or neighbor of the client.
When that factual premise exists, practitioners should interview
other members of the client’s circle of acquaintances to see what
facts they had heard about the crime through the rumor mill.
Finally, practitioners should also ask common-sense questions
concerning the client’s independent familiarity with the crime scene.
Imagine a case in which a husband confesses to killing his wife with a
knife that he had gotten out of the left-hand kitchen drawer. The fact
that the husband correctly identified the drawer in which his own
knives were stored can hardly be viewed as evidence of guilt, of
course, for the simple reason that his knowledge of that fact need not
have arisen solely from his participation in the crime. When
reviewing a confession for this sort of contamination, accordingly, a
practitioner simply needs to ask his client about his prior experience
with and exposure to the crime scene.
Once all these sources of potential contamination have been
identified, the practitioner will be well on his way to answering the
key question: Was the client ever able to produce any verifiably true
facts about the crime, absent contamination? To be able to answer
that question completely, however, the practitioner must turn his
attention to the matter of “fit.” Fit can be roughly defined as the
degree to which the suspect’s narrative matches the provably true
facts of the crime.53 It is a matter that, as previously referenced,
preoccupies officers during interrogation, but it should concern
defense practitioners even more. Police reports, crime scene
photographs and videos, witness statements, and laboratory reports,
53 See Ofshe and Leo, 74 Denver U L Rev at 990–94 (cited in note 32).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 855

of course, are all familiar sources of information concerning what can
be objectively known about the crime itself.
As may be already clear, the sheer task of collecting and
managing all the information that bears on this confession analysis
can be daunting. It can be tempting to shortcut one’s task by
identifying a few instances in which the confession narrative fits the
crime scene facts and concluding that the confession must be true; it
can be equally tempting to identify a few instances in which the
suspect appeared to volunteer unprompted details during
interrogation and conclude on that basis that the confession must be
true. Such shortcuts, however, are mistakes—and indeed, as
Garrett’s exposé makes clear, such mistakes play out in courtrooms
across the country all too frequently.
Instead, a defense practitioner must cross-reference in a
systematic and complete fashion the results of his investigation of
contamination and fit. The best method for doing this is to create a
chart that breaks down the confession, detail by painstaking detail,
and sets forth (1) the sources of contamination for each detail, if any,
and (2) the degree to which that detail can be corroborated by
objective, physical evidence.
Brendan’s confession can be charted in this manner. His
knowledge of certain details—the fact that the victim was shot in the
head and the fact that her vehicle’s battery had been disconnected—
is quite obviously the product of police fact-feeding, as can be
discerned from a careful review of the transcript of his interrogation.
His knowledge of other details, however, is not the product of factfeeding.
In fact, some of those details at first glance appear to be
volunteered in response to open-ended questions.
On closer review, however, his knowledge of these other details
can be traced to other sources of contamination. The media had
widely publicized the fact that the victim’s SUV and incinerated
remains had been found on Avery’s property, for instance.54
Brendan, who lived next door to his uncle, was also quite familiar
with these facts from overhearing his own family’s discussions of the
case as it unfolded.55 Even the fact that the SUV had been covered
with branches had been well publicized.56 Further, Brendan was
independently familiar with his uncle’s home, which he had visited
many times before, so his ability to describe where his uncle kept the
bleach—in the bathroom, a room likely to be seen by a visitor—
54 See Dassey Memorandum of Facts at *6–7 (cited in note 36).
55 Id at *8–9.
56 Id at *7.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
856 The University of Chicago Law Review [79:837
again did nothing to confirm guilty knowledge.57 And on even closer
examination, those few details that Brendan gave during his
confession that are not traceable to any contamination at all prove to
be incorrect or, at the least, unverifiable. Despite tearing Avery’s
home apart, police never were able to come up with any physical
evidence—no speck of DNA, no loose strand of hair, no trace of
blood—establishing that the victim had been sexually assaulted, that
her hair had been cut, or that she had been stabbed.58 The below
table summarizes these findings.
57 Id at *1, 35.
58 See Dassey Memorandum of Facts at *11–12, 35–36 (cited in note 36).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 857

TABLE 1: CONTAMINATION IN BRENDAN DASSEY’S CONFESSION
Source: Dassey Memorandum of Facts at *1–64 (cited in note 36);
Interview of Brendan Dassey at *578–79, 584, 585–87, 600–02, 610 (cited in
note 49).
When viewed in this comprehensive manner, it is plain that
Brendan was not able to say anything provably correct about the
crime absent the guiding hand of contamination. Such a result is a
red flag of unreliability.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
858 The University of Chicago Law Review [79:837
Determining that a confession bears signs of unreliability is one
significant hurdle in a confession case, but another even more
significant hurdle is convincing the fact finder to embrace that view.
The traditional course that most criminal cases follow provides
multiple opportunities to raise unreliability both before and during
trial—and for innovative practitioners willing to deviate from
tradition, there may be ways to generate even more such
opportunities.
The first such opportunity occurs within the context of a pretrial
written motion to suppress the confession on voluntariness grounds.
Unfortunately, the vast majority of jurisdictions currently draw a
bright-line distinction between voluntariness and reliability.59 Many
courts accordingly refuse to entertain an argument that appears to
relate only to reliability within the context of a voluntariness
motion.60 This does not mean, however, that all of them do. From
time to time, courts have suppressed or admitted confessions using
reliability rationales to bolster their conclusions regarding
voluntariness.61 The recurring nature of such rulings provides reason
enough to pepper a motion to suppress on voluntariness grounds
with evidence of unreliability, in hopes that the judge will at least
tacitly consider that evidence in the voluntariness analysis. If nothing
else, doing so will frame the issue of unreliability for the judge to
consider later in the proceedings.
Another option is to argue that when contamination takes the
form of fact-feeding by police, such contamination is itself a form of
coercion—and thus that evidence of contamination is directly
relevant to the court’s voluntariness analysis. Such an approach was
presaged in the Miranda v Arizona62 decision itself, in which the US
Supreme Court questioned police tactics designed to cause a suspect
to “merely confirm[] the preconceived story the police seek to have
him describe.”63 This approach has received more concrete support,
moreover, from the New York Court of Appeals’s recent decision in
59 See, for example, Colorado v Connelly, 479 US 157, 167 (1986) (affirming the
admissibility of an apparently unreliable confession, because “[a] statement rendered by one in
the condition of respondent might be proved to be quite unreliable, but this is a matter to be
governed by the evidentiary laws of the forum . . . and not by the Due Process Clause of the
Fourteenth Amendment”).
60 Id.
61 See, for example, People v Ventura, 786 NYS2d 469, 470 (NY App 2004) (affirming the
conviction upon finding that the evidence established that the defendant’s confessions were
both voluntary and reliable).
62 384 US 436 (1966).
63 Id at 455.
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 859

Warney v New York.64 There, the state’s highest tribunal reviewed a
lower court’s dismissal of a civil rights lawsuit brought by exoneree
Douglas Warney. Warney had argued that his false confession had
been coerced in part because the police had fed him facts during his
interrogation that he was then pressured into repeating.65 In
reversing the dismissal, the court of appeals held that Warney’s
allegation that, “whether negligently or through intentional
manipulation, police misconduct led to the inclusion of these details
in Warney’s statement” was sufficient to state a claim that he was
“coerced into adopting the false confession.”66 By linking
contamination and coercion in this manner—albeit in a civil context
premised in part on DNA proof of Warney’s actual innocence—the
court may well have launched a new avenue of argument for
practitioners.
Following the filing of the written motion to suppress, any
ensuing pretrial suppression hearing represents a chance to
introduce evidence of contamination and unreliability via live
testimony. At most suppression hearings, of course, the interrogating
officers take the stand to testify. A defense practitioner concerned
about reliability can establish through cross-examination two
important facts that relate to contamination: first, the officer’s
knowledge prior to interrogation of specific information about the
crime and, second, the officer’s use of leading questions during the
interrogation that, in effect, fed that information straight to the
client. The goal here is to document exhaustively every instance in
which the police revealed correct facts about the case to the client
during interrogation.67 Such a presentation, in fact, can be quite
compelling when done in a thorough manner.
While cross-examination of the interrogating officers is an
essential part of any effort to prove unreliability, it is not the only
avenue of attack. Frequently, the testimony of a false confession
expert—in other words, a social psychologist with expertise in police
interrogation tactics—can also be useful. Such an expert should be
able to testify about the phenomenon of contamination and explain
to the judge how contaminated confessions can be unreliable—
perhaps reciting some of the same troubling statistics and stories of
64 947 NE2d 639 (NY 2011).
65 Id at 642.
66 Id at 644.
67 See Deja Vishny, Cross Examining Police in False Confession Cases, 16 Wis
Defender *2–5 (Winter–Spring 2008), online at http://www.wisspd.org/htm/ATPracGuides
/WisDef/WinSpr08/CE.pdf (visited Nov 16, 2011).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
860 The University of Chicago Law Review [79:837
contamination that are included in Garrett’s book. After this
foundation has been laid, the expert should go on to analyze the
client’s confession for contamination and must identify every
possible instance of contamination in the client’s statement. And
finally, the expert should testify that contamination is not something
that concerns only the defense bar; she or he can explain that police
officers are commonly trained to withhold facts during interrogation
because the use of leading questions can suggest information to the
suspect and taint the resulting statement. Although false confession
experts generally are not permitted to opine on the truth or falsity of
the confession, they can give jurors an analytical framework against
which they can assess its reliability.
If the judge will not allow this type of presentation during a
pretrial voluntariness hearing, practitioners should consider filing a
separate pretrial motion in limine requesting the suppression of the
confession squarely on reliability—not voluntariness—grounds.68
Such a motion could fall under the auspices, for example, of a local
rule of evidence modeled on Federal Rule of Evidence 403, which
excludes evidence when the risk of prejudice outweighs the
probative value.69 Confessions, after all, are extraordinarily
prejudicial evidence precisely because they appear to be compelling
evidence of guilt; but when a confession is demonstrably
contaminated and unreliable, it may have little real probative value.
While such an approach is admittedly innovative, there is
substantial and wide-ranging precedent for the notion that judges
may exclude unreliable evidence. Federal judges, for instance, have
long had the power to exclude even voluntary confessions in the
absence of “substantial independent evidence which would tend to
establish the trustworthiness of the statement.”70 Similarly, several
states allow trial courts to hold pretrial reliability hearings—often
called “taint” hearings—to evaluate whether child witnesses’
statements should be excluded as unreliable.71 Judges in several states
also are allowed to exclude unreliable or untrustworthy eyewitness
identifications before trial as a matter of routine.72 More broadly,
almost every state permits its judges to admit or exclude hearsay
68 See Leo, et al, 2006 Wis L Rev at 531–33 (cited in note 35).
69 See FRE 403.
70 Opper v United States, 348 US 84, 93 (1954).
71 See, for example, Fischbach v State, 1996 WL 145968 *2 (Del); State v Michaels,
642 A2d 1372, 1382–83 (NJ 1994); People v Michael M., 618 NYS2d 171, 178 (NY Sup Ct 1994).
72 See, for example, Manson v Brathwaite, 432 US 98, 110–14 (1977); People v Rodriguez,
475 NE2d 443, 444–45 (NY 1984).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
2012] Combating Contamination in Confession Cases 861

testimony based on a generalized evaluation of its reliability.73 Judges
in most states also regularly weigh the reliability of expert opinions
before permitting them to be introduced.74 Plainly, judges are often
asked to engage in pretrial evaluations of the reliability of certain
pieces of evidence; confessions should be no different.
Finally, of course, the issue of contamination must be raised
directly to the fact-finder at trial. Indeed, proving contamination and
coercion must be the heart of any false confession defense. Such a
defense can be accomplished largely by combining effective crossexamination
of the interrogating officers with expert testimony, as
described above in the context of a suppression hearing. It bears
noting, however, that any discussion of contamination can become
repetitive and even boring, because it necessarily entails a detail-bydetail
review of the confession. To keep jurors engaged, practitioners
are urged to use visual aids during this portion of the defense case.
The contamination and fit chart described above could be enlarged
and filled out, line by line, during an expert’s testimony about
contamination—or, even more effectively, during cross-examination
of the officer who fed each fact to the defendant. During closing
argument, the defense attorney could then display the filled-out
chart to the jury in an effort both to remind them of the underlying
testimony and to provide them with a visual illustration of the
contamination argument. If the confession was videotaped,
moreover, the defense should replay the tape during its case in chief,
stopping and starting it as appropriate to draw the jury’s attention to
each instance of fact-feeding and police prompting. A jury that sees
the confession tape for the first time during the prosecution’s case
may be blinded to any fact-feeding by the emotional force of the
confession itself. Accordingly, it is the duty of the defense attorney
to replay the tape and highlight fact-feeding when it occurs.
CONCLUSION
There can be no question that with the publication of Convicting
the Innocent, Brandon Garrett has established himself as one of this
generation’s most significant wrongful convictions scholars. Simply
73 See FRE 807 (allowing for the admissibility of hearsay testimony, even if it does not fit
into one of the established hearsay exceptions, if it bears “circumstantial guarantees of
trustworthiness”); Alex Stein, Constitutional Evidence Law, 61 Vand L Rev 65, 106 (2008).
74 See Daubert v Merrell Dow Pharmaceuticals, Inc, 509 US 579, 591–96 (1993); Kumho
Tire Co, Ltd v Carmichael, 526 US 137, 147 (1999). See also Martin S. Kaufman, The Status of
Daubert in State Courts 4–23 (Atlantic Legal Foundation 2006), online at http://
www.atlanticlegal.org/daubertreport.pdf (visited Nov 16, 2011).
07 NIRIDER, DREPFER, DRIZIN BKR (DO NOT DELETE) 6/5/2012 1:13 PM
862 The University of Chicago Law Review [79:837
put, this is one of the most important books ever written on the
subject of wrongful convictions. It is a deserving heir to the critical
work done by Actual Innocence. We are fortunate, moreover, that
Garrett’s work may well turn out to be the proverbial gift that keeps
on giving; his data and methodology are publicly available on his
University of Virginia website, accessible to other researchers who
themselves will be able to contribute to and make use of them.75
Under Garrett’s leadership and direction, the ever-growing corpus of
DNA exoneration data promises to be a treasure trove for many
years to come.
Before Convicting the Innocent will have its maximal effect,
however, much more must happen than just the reading of a book or
the collection of a dataset. Defense lawyers must take Garrett’s book
off their shelves and use its data in the courtroom, in the defense of
their clients. They must cite Garrett’s findings in their legal briefs
and motions; they must use his statistics to effectively examine
witnesses; and they must use his stories of tragedy and redemption to
educate jurors about the causes and consequences of wrongful
convictions. Perhaps nowhere is this lesson more important than in
confession cases, given the unquestioning faith that many jurors
place in the reliability of confessions. This Review attempts to give
practitioners some ideas concerning how to address the problem of
the contaminated confession that Garrett identifies; but we recognize
and acknowledge that our ideas are far from comprehensive.
Together, practitioners—whether trial attorneys, appellate
advocates, or clinical professors—can develop strategies to combat
this problem. Together, we can teach ourselves and future
generations how to use the lessons of Convicting the Innocent to
prevent future injustices.
75 http://www.law.virginia.edu/html/librarysite/garrett_innocent.htm (visited Nov 16, 2011).