LITTLE M ACRO.DOC

10/6/08 11:23 PM

ADDRESSING THE EVIDENTIARY
SOURCES OF WRONGFUL CONVICTIONS:
CATEGORICAL EXCLUSION OF EVIDENCE
IN CAPITAL STATUTES
Rory K. Little∗
In an effort to provoke realistic remedial thinking,1 this essay presents
and discusses a proposed model statute that would ban any capital
prosecution that is based primarily on the types of evidence we now know

∗ Professor of Law, U.C. Hastings College of the Law. Many thanks to Professor Myrna
Raeder for allowing me to present these thoughts in unvarnished form at the Southwestern Law
School’s Symposium on Wrongful Convictions in February 2008. The Law Faculty at the
University of Leiden also graciously listened to a presentation of these ideas during my visit there
in April 2008; my thanks go particularly to Dean Carel Stoelker, Criminal Law Department Chair
Professor C.P.M. Cleiren; and Els LeMaire, director of the International Education Office at
Leiden. Thanks also to Morgan Weibel and Todd Daloz (both Hastings 2009) for their research
assistance; to Hastings Research Librarian Chuck Marcus and his co-workers; and the
Southwestern staff for patience and editing assistance. Finally, Professors Kate Bloch, Sam
Gross, Rick Marcus, Richard Leo, Chris Slobogin, and Scott Sundby offered helpful thoughts (if
not always agreement) and saved me from a number of errors.
1. Contemplating what to say at a symposium on Wrongful Convictions, I was surprised to
find nowhere in the mass of literature generated since 1989, any proposals to flat-out ban the
sources of evidence that show up in case after case of demonstrated “actual innocence.” I am not
sure I would endorse the statute proposed here, were I legislator or judge. As a former prosecutor,
I believe in the ideal of ethically exercised, and supervisorily reviewed, prosecutorial discretion.
And I am not an abolitionist regarding the death penalty. See Rory K. Little, What Federal
Prosecutors Really Think: The Puzzle of Statistical Race Disparity Versus Specific Guilt, and the
Specter of Timothy McVeigh, 53 DEPAUL L. REV. 1591, 1597 (2004). Nevertheless, the revelation
of so many “actual innocence” cases involving poorly exercised prosecutorial discretion must give
anyone pause.
The idea of excluding known evidentiary sources of wrongful conviction, as opposed to
merely reducing them, at least in capital cases, seems an obvious one that should be considered. If
we really believe that particular types of evidence can, too easily, lead to wrongful convictions,
why do we allow it, when a life hangs in the balance? One response is Justice Scalia’s, in Kansas
v. Marsh, 548 U.S. 163, 199 (2006) (Scalia, J., concurring): “One cannot have a system of
criminal punishment without accepting the possibility that someone will be punished mistakenly.
That is a truism, not a revelation.” This may be true generally; the question posed in this article is,
are we willing to accept it in the capital context?

101

LITTLE M ACRO.DOC

102

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

to be sources of convicting the innocent.2 I focus solely on capital cases, as
they are irremediable once the sentence is carried out, and thus (perhaps3)
the most disturbing category of wrongful convictions.4 Moreover, this is an
essay, not a tenure piece, and its brief discussion surely does not answer, or
2. Unsurprisingly in such an emotionally powerful area, the semantics of “wrongful
convictions” are underlaid with deep disagreements and suspicion. See Adam Liptak, Consensus
on Counting the Innocent: We Can’t, N.Y. TIMES, Mar. 25, 2008, at A14 (“[E]xoneration in the
capital concept is a funny concept. It suggests complete vindication, but its real meaning is
generally narrower.”). Thus Professor Gross and his co-authors use and define the term
“exonerations,” and note that they do not propose “an independent judgment on the factual
innocence of each” of the 340 defendants who they say have been exonerated. Samuel R. Gross et
al., Exonerations in the United States, 1989 Through 2003, 95 J. CRIM. L. & CRIMINOLOGY 523,
524, 526 (2005). Others use the term “wrongful conviction” in a similar fashion, describing any
case in which the conviction has been reversed for reasons of misconduct and/or in situations
where innocence is strongly asserted. Still others include the “actually innocent” but also “those
who are guilty of something, just not the crime of conviction.” Andrew D. Leipold, How the
Pretrial Process Contributes to Wrongful Convictions, 42 AM. CRIM. L. REV. 1123, 1124 (2005).
And yet others use the term “wrongful conviction” without defining it, likely considering the
concept well enough settled to not merit a distracting definitional struggle. E.g., Myrna S. Raeder,
See No Evil: Wrongful Convictions and the Prosecutorial Ethics of Offering Testimony by
Jailhouse Informants and Dishonest Experts, 76 FORDHAM L. REV. 1413, 1413 (2007).
It now seems undisputable that DNA revelations have demonstrated beyond all doubt that
some (whether a “few” or “many” is hotly disputed) convicted capital defendants have been
factually, “actually,” innocent. See Kansas v. Marsh, 548 U.S. 163, 209-10 (2006) (Souter, J.,
dissenting) (“many” death row exonerees have been “cleared by DNA evidence”). Justice
Scalia’s response to this is to claim only that there is “not a single verifiable case” of wrongful
“execution”—not conviction—and that the death row error rate “has been reduced to an
insignificant minimum,” not zero. Id. at 199 (Scalia, J., concurring). See also Gross, infra note 7
(detailing four wrongful conviction cases the author believes showcase not only “actual
innocence” but actual executions).
At the other extreme, however, it is (I hope) equally undisputable that not all convictions
are “wrongful.” Some (indeed, we hope, the great majority) are very “right” indeed. Nor are all
convictions that are reversed on appeal “wrongful.” In our system of constitutionally enshrined
Due Process, reversible error can be present even when all actors have attempted to perform in
complete good faith. In addition, reversible error can be present even in the face of clear guilt,
where significant rights of a criminal defendant are ignored or overridden. Yet among all the
sharply opposed forces that care deeply about the criminal justice system, I consider it to be
shared territory that conviction of the factually innocent is plainly “wrongful,” even when the fault
for the conviction cannot clearly be laid on any one actor in the system. This essay has protection
of the factually innocent capital defendant centrally in mind.
3. Cf. Gross et al., supra note 2, at 536 (asserting that “nobody, it seems, seriously pursues
exonerations” for “routine” felonies and misdemeanors, and that innocents are wrongfully
convicted in those categories “in numbers that dwarf” the 340 exonerations in very serious cases
examined in Professor Gross’s article).
4. See Hugo Adam Bedau et al., Convicting the Innocent in Capital Cases: Criteria,
Evidence, and Inference, 52 DRAKE L. REV. 587, 587 (2004). Professor Bedau’s immense body
of work in this area over a half century places him at the forefront of the committed group of antideath penalty researchers in the United States. See, e.g., Hugo A. Bedau, Murder, Errors of
Justice, and Capital Punishment, in THE DEATH PENALTY IN AMERICA 434 (Hugo Adam Bedau
ed., 1964).

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

103

even address, every possible aspect of the idea. But let’s see how serious
we really are about totally eliminating (rather than just reducing) wrongful
convictions, by imagining a total ban on their known evidentiary sources—
indeed, a ban that is immediately judicially enforceable—when the death
penalty is at issue.

A. THE KNOWN EVIDENTIARY SOURCES OF WRONGFUL CONVICTIONS,
AND CURRENTLY PROPOSED REMEDIES .
Controversy regarding convictions of the innocent is not new.5 But in
the wake of newly developed DNA science, beginning less than two
decades ago,6 undisputable evidence of “actual innocence” convictions on
America’s death row is now well-accepted.7 After more than a decade of
5. The early history of “actual innocence” concern in this country has been largely
forgotten. Almost a century ago, the Journal of Northwestern University’s American Institute of
Criminal Law and Criminology reported that a year-long survey of wardens of “every prison in
the United States and in Canada,” conducted by the American Prison Congress, had revealed “not
a single case” of “capital punishment wherein there was reasonable doubt as to the guilt of the
victim.” Robert H. Gault, Find No Unjust Hangings, 3 J. CRIM. L. & CRIMINOLOGY 131, 131
(1912). In response, Edwin Borchard, then the Librarian of Congress, asserted “[t]hat there have
been numerous cases of this kind besides the recent Toth case in Pennsylvania . . . there is no
doubt, notwithstanding the unauthentic returns from wardens collected by the American Prison
Congress . . . .” Edwin M. Borchard, European Systems of State Indemnity for Errors of Criminal
Justice, 13 J. CRIM. L. & CRIMINOLOGY 684, 706 (1913). Borchard’s article was reprinted,
together with an introductory editorial by the famous evidentiary Dean Wigmore, to accompany
the Senate introduction of a “Bill (S. 7675) to Grant Relief to Persons Erroneously Convicted in
Courts of the United States.” EDWIN M. BORCHARD, STATE INDEMNITY FOR ERRORS OF
CRIMINAL JUSTICE, S. DOC. NO. 974, at 1 (3d Sess. 1912). Twenty years later, Professor
Borchard published his famed Convicting the Innocent, which is now often, if ahistorically,
credited as being the “first” discussion of actual innocence in the United States. EDWIN M.
BORCHARD, CONVICTING THE INNOCENT: SIXTY-FIVE ACTUAL ERRORS OF CRIMINAL JUSTICE
(1932); e.g., Daniel S. Medwed, Looking Foreward: Wrongful Convictions and Systemic Reform,
42 AM. CRIM. L. REV. 1117, 1117 n.1 (2005).
6. “[T]he period starting in 1989 has seen repeated exonerations of convicts under death
sentences, in numbers never imagined before the development of DNA tests.” Kansas v. Marsh,
548 U.S. 163, 207-08 (2006) (Souter, J., dissenting). As Professor Daniel Medwed describes,
while the specter of wrongful convictions has haunted the literature for decades, it is “[t]he
emergence of post-conviction DNA testing in the past sixteen years” that “has provided the
scientific arrow to the scholar’s theoretical bow . . . .” Medwed, supra note 5, at 1117.
7. See supra note 2; Accord Richard A. Leo et al., Bringing Reliability Back In: False
Confessions and Legal Safeguards in the Twenty-First Century, 2006 WIS. L. REV. 479, 526
(2006) (DNA and the “innocence revolution has rocked the entire criminal justice system”). This
essay does not seek to resolve the debate regarding whether an innocent capital defendant has
actually been executed—the fact of a number of death row exonerations before execution is
sufficient to give strong concern. The longstanding debate about the execution of innocents, see
supra note 5, has continued in recent times. See, e.g., Stephen J. Markman & Paul G. Cassell,
Protecting the Innocent: A Response to the Bedau-Radelet Study, 41 STAN. L. REV. 121 (1988)

LITTLE M ACRO.DOC

104

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

revelatory jurisprudence,8 six evidentiary sources of wrongful convictions
are actually well known 9:
faulty eyewitness identifications;10 false
11
confessions; jailhouse snitches;12 “junk” science (meaning unreliable or
(contesting Bedau and Radelet’s claim that 23 innocent defendants have, in fact, been executed);
MICHAEL L. RADELET, HUGO ADAM BEDAU & CONSTANCE E. PUTNAM, IN SPITE OF
INNOCENCE: ERRONEOUS CONVICTIONS IN CAPITAL CASES (Northeastern 1992) (repeating their
claim). The debate has extended recently into the United States Supreme Court. See Marsh, 548
U.S. 163, 188-92 (2006) (Scalia, J., concurring) (noting prior comments by Justices Stevens,
Souter and Blackmun suggesting that “[i]nnocent persons have been executed . . . ,” id. at 193 n.4,
and asserting in response that there is “not a single case—not one—in which it is clear that a
person was executed for a crime he did not commit”). Id at 188. See also Samuel R. Gross,
Souter Passant, Scalia Rampant: Combat in the Marsh, 105 MICH. L. REV. FIRST IMPRESSIONS
67, 71 (2006) (responding to Justice Scalia by describing four cases in which an innocent
defendant may have been executed, noting that “there is no generally available procedure for
securing an exoneration after execution,” and stating “[w]e know it happens”); Joshua Marquis,
The Myth of Innocence, 95 J. CRIM. L. & CRIMINOLOGY 501 (2005).
8. The paradigm-shifting concept of “actual innocence” in the criminal justice system—that
is, criminal defendants who have been convicted, sentenced, and affirmed on appeal, yet are later
determined to have been not guilty—was brought to the popular (as well as judicial and legal)
consciousness in undebateable form by the 2000 publication of the book of the same title,
detailing the new DNA testing science for the first time. JIM DWYER, PETER NEUFELD & BARRY
SCHECK, ACTUAL INNOCENCE: FIVE DAYS TO EXECUTION AND OTHER DISPATCHES FROM THE
WRONGLY CONVICTED (Doubleday 2000) (published in paperback as BARRY SCHECK, PETER
NEUFELD & JIM DWYER, ACTUAL INNOCENCE: WHEN JUSTICE GOES WRONG AND HOW TO
MAKE IT RIGHT (New Am. Library 2003)). With the 2006 publication of The Innocent Man by
best-selling novelist John Grisham, “actual innocence” has now penetrated the popular American
consciousness.
9. Of course, the “revelation” of wrongful convictions in the United States goes much
further back than Actual Innocence. The seminal authority remains the 1932 presentation by Yale
Law Professor Edwin Borchard of 65 cases in which innocent defendants were convicted.
BORCHARD, CONVICTING THE INNOCENT: SIXTY-FIVE ACTUAL ERRORS OF CRIMINAL JUSTICE,
supra note 5. Borchard’s writing on this topic dates back to 1912, see supra note 5, and the
history in European countries goes back even further. See Borchard, European Systems of State
Indemnity for Errors of Criminal Justice, supra note 5. Indeed, it seems safe to say that so long as
human society engages in accusation and punishment for wrongdoing, wrongful convictions will
also occur. See, e.g., RADELET ET AL., supra note 7. Today, the most comprehensive sources for
“actual innocence” examples and information are found on the web (of course), at the
Northwestern
Law
School’s
Center
on
Wrongful
Convictions,
www.law.northwestern.edu/wrongfulconvictions, and Cardozo Law School’s Innocence Project,
www.innocenceproject.org.
See
also
Death
Penalty
Information
Center,
www.deathpenaltyinfo.org.
10. See, e.g., Sandra Guerra Thompson, Beyond a Reasonable Doubt? Reconsidering
Uncorroborated Eyewitness Identification Testimony, 41 U.C. DAVIS L. REV. 1487 (2008); Calvin
TerBeek, A Call for Precedential Heads: Why the Supreme Court’s Eyewitness Identification
Jurisprudence is Anachronistic and Out-of-Step with the Empirical Reality, 31 LAW & PSYCHOL.
REV. 21 (2007); Kathy Pezdek, Kathryn Sperry & Shana M. Owens, Interviewing Witnesses: The
Effect of Forced Confabulation on Event Memory, 31 LAW & HUM. BEHAV. 463 (2007); NATHAN
R. SOBEL, EYEWITNESS IDENTIFICATION (West Group 2002) (1981); BRIAN L. CUTLER &
STEVEN D. PENROD, MISTAKEN IDENTIFICATION: THE EYEWITNESS, PSYCHOLOGY, AND THE
LAW (1995).
11. See, e.g., Brandon L. Garrett, Judging Innocence, 108 COLUM. L. REV. 55 (2008); Boaz

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

105

unverified forensic evidence); 13 bad lawyering (whether by the prosecution
or the defense)14 and general witness misconduct (whether by law
enforcement, victims, experts, or other witnesses). The great multitude of
“actual innocence” case histories, and the secondary literature, converge on
these evidentiary sources of error.15
In response, a multiplicity of procedural protections have been
proposed: improved eyewitness identification procedures and independent
corroboration;16 protective jury instructions;17 videotaped interrogations and
Sangero, Miranda is Not Enough: A New Justification for Demanding “Strong Corroboration” to
a Confession, 28 CARDOZO L. REV. 2791 (2007); Boaz Sangero & Mordechai Halpert, Why a
Conviction Should Not Be Based on a Single Piece of Evidence: A Proposal for Reform, 48
JURIMETRICS J. 43 (2007); Dale E. Ives, Preventing False Confessions: Is Oickle Up to the Task?,
44 SAN DIEGO L. REV. 477 (2007); Christopher Slobogin, Lying and Confessing, 39 TEX. TECH L.
REV. 1275 (2007). The leading authority on false confessions, among many, is professor Richard
Leo. See, e.g., Richard A. Leo et al., supra note 7; Steven A. Drizin & Richard A. Leo, The
Problem of False Confessions in the Post-DNA World, 82 N.C. L. REV. 891 (2004).
12. See Raeder, supra note 2.
13. See Roger C. Park, Signature Identification in the Light of Science and Experience, 59
HASTINGS L. J. 1101, 1101 (2008) (“The example of DNA has shown what real science can do,
and has highlighted the shortcomings of other forms of forensic science.”); Jessica D. Gabel &
Margaret D. Wilkinson, “Good” Science Gone Bad: How the Criminal Justice System Can
Redress the Impact of Flawed Forensics, 59 HASTINGS. L.J. 1001 (2008). See also DAVID L.
FAIGMAN ET AL., MODERN SCIENTIFIC EVIDENCE: THE LAW AND SCIENCE OF EXPERT
TESTIMONY 147 (2002); DAVID L. FAIGMAN ET AL., SCIENCE IN THE LAW: STANDARDS,
STATISTICS AND RESEARCH ISSUES 93 (2002).
14. See, e.g., DWYER ET AL., supra note 8, at 172-92.
15. See, e.g., Samuel R. Gross, supra note 7, at 70-71 (reacting to the Kansas v. Marsh
decision by detailing wrongful conviction cases that foundered on such evidentiary sources, and
resulted in actual execution); Peter A. Joy, The Relationship Between Prosecutorial Misconduct
and Wrongful Convictions: Shaping Remedies for a Broken System, 2006 WIS. L. REV. 399
(2006); Robert P. Mosteller, Exculpatory Evidence, Ethics, and the Road to the Disbarment of
Mike Nifong: The Critical Importance of Full Open-File Discovery, 15 GEO. MASON L. REV. 257
(2008); Robert R. Rigg, The T-Rex Without Teeth: Evolving Strickland v. Washington and the Test
for Ineffective Assistance of Counsel, 35 PEPP. L. REV. 77 (2007); Stephen B. Bright, Counsel for
the Poor: The Death Sentence Not for the Worst Crime but for the Worst Lawyer, 103 YALE L.J.
1835 (1994).
16. Noah Clements, Flipping a Coin: A Solution for the Inherent Unreliability of Eyewitness
Identification Testimony, 40 IND. L. REV. 271 (2007) (arguing for the complete removal of
eyewitness identification); Thompson, supra note 10 (arguing for strong corroborating evidence
rule); Timothy P. O’Toole & Giovanna Shay, Manson v. Brathwaite Revisited: Towards a New
Rule of Decision for Due Process Challenges to Eyewitness Identification Procedures, 41 VAL. U.
L. REV. 109 (2006); BUREAU OF TRAINING AND STANDARDS FOR CRIMINAL JUSTICE, WIS. DEP’T
OF JUSTICE, MODEL POLICY AND PROCEDURE FOR EYEWITNESS I DENTIFICATION (Sept. 12,
2005), available at http://www.doj.state.wi.us/dles/tns/EyewitnessPublic.pdf (recommending
improved eyewitness identification procedures); TECHNICAL WORKING GROUP FOR EYEWITNESS
EVIDENCE, U.S. DEP’T OF JUSTICE, EYEWITNESS EVIDENCE: A GUIDE FOR LAW ENFORCEMENT
(Oct. 1999), available at www.ncjrs.gov/pdffiles1/nij/178240.pdf (suggesting the double-blind
administration of both photo spreads and corporeal line-ups); Gary L. Wells et al.,
Recommendations for Properly Conducted Lineup Identification Tasks, in ADULT EYEWITNESS

LITTLE M ACRO.DOC

106

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

Daubert-like “reliability hearings”;18 strongly regulated forensic labs and
examiners; 19 independent case reviews; 20 limiting “anecdotal” forensic
testimony;21 better lawyer training, funding and oversight22 and stronger
lawyer ethics rules.23 Indeed, the American Bar Association, a sometimes
slow-moving but powerful representative force in American law, has
adopted 11 different resolutions advocating a large number of reforms.24

TESTIMONY: CURRENT TRENDS AND DEVELOPMENTS 223 (David Frank Ross et al. eds., 1994).
17. Tim Bakken, Truth and Innocence Procedures to Free Innocent Persons: Beyond the
Adversarial System, 41 U. MICH. J.L. REFORM 547, 561 (2008); Alexander G.P. Goldenberg,
Interested, But Presumed Innocent: Rethinking Instructions on the Credibility of Testifying
Defendants, 62 N.Y.U. ANN. SURV. AM. L. 745 (2007); WILLIAM J. PALMER, CALIFORNIA JURY
INSTRUCTIONS, CRIMINAL: A COMPANION BOOK TO CALIFORNIA JURY INSTRUCTIONS 625
(1946).
18. Leo et al., supra note 7; Lisa Lewis, Rethinking Miranda: Truth, Lies, And Videotape, 43
GONZ. L. REV. 199 (2007); Saul M. Kassin, Richard A. Leo, Christian A. Meissner, Kimberly D.
Richman, Lori H. Colwell, Amy-May Leach & Dana La Fon, Police Interviewing and
Interrogation: A Self-Report Survey of Police Practices and Beliefs, 31 LAW & HUM. BEHAV. 381
(2007); Slobogin, supra note 11.
19. E.g. Paul C. Gianelli, Wrongful Convictions and Forensic Science: The Need to Regulate
Crime Labs, 86 N.C. L. REV. 163 (2007); Brian Witte, State Legislation: Crime Labs and
Forensic Evidence Reform: Md. Puts Teeth in Bill to Regulate Crime Labs, May 7, 2007,
available at http://www.nacdl.org/sl_docs.nsf/issues/CrimeLab; 42 U.S.C. § 3797k (4) (Supp. IV
2007) (requiring that DNA laboratories receiving federal grants create mechanisms for external
independent investigations); sources cited supra note 13.
20. E.g. Raeder, supra note 2, at 1417; Jerome M. Maiatico, All Eyes on Us: A Comparative
Critique of the North Carolina Innocence Inquiry Commission, 56 DUKE L.J. 1345, 1366 (2007).
21. David L. Faigman, Anecdotal Forensics, Phrenology, and Other Abject Lessons from the
History of Science, 59 HASTINGS L.J. 979, 996-1000 (2008).
22. E.g. Daniel S. Medwed, Anatomy of a Wrongful Conviction: Theoretical Implications
and Practical Solutions, 51 VILL. L. REV. 337, 371-74; Jodi L. Short & Michael W. Toffel,
Coerced Confessions: Self-Policing in the Shadow of the Regulator, 24 J.L. ECON. & ORG. 45, 64
(2008).
23. E.g. Raeder, supra note 2, at 1416-17; Kevin Hopkins, The Politics of Misconduct:
Rethinking How We Regulate Lawyer-Politicians, 57 RUTGERS L. REV. 839 (2005). In addition,
the American Bar Association recently agreed to revisions to Model Rule of Professional
Responsibility 3.8 (“Special Responsibilities of a Prosecutor”) to address actual innocence claims
and the author (Professor Little) is currently Reporter for an ABA Task Force to revise the ABA’s
Criminal Justice Standards for the Prosecution Function and Defense Function that is expressly
focusing on stronger model rules to address actual innocence issues.
24. See AM. BAR ASS’N, ACHIEVING JUSTICE: FREEING THE INNOCENT, CONVICTING THE
GUILTY: REPORT OF THE ABA CRIMINAL JUSTICE SECTION’S AD HOC INNOCENCE COMMITTEE
TO ENSURE THE INTEGRITY OF THE CRIMINAL PROCESS (Paul Gianelli & Myrna Raeder eds.,
2006).

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

107

B. THE PROPOSAL: STATUTORY EXCLUSION OF KNOWN EVIDENTIARY
SOURCES OF WRONGFUL CONVICTIONS.
There is no doubt that the problem of wrongful convictions is a
complex one, with multiple sources and no perfect resolution, given
inevitably flawed human processes.25 Moreover, although simplicity is a
virtue, solutions to complex problems are also usually complex. Certainly a
mosaic of solutions and protections are necessary to address the problem of
wrongful convictions. Moreover, the problem cannot be “solved” so much
as managed, since the challenge to fairly address crime in a free society will
never disappear.26 I offer my model statutory idea below as a supplement
to, not a substitute for, all the remedial thinking that has preceded it.
Nevertheless, this essay proposes an approach to the problem of
wrongful convictions different from those heretofore proposed, one that I
have not seen previously suggested.27 It is different from—perhaps
supplemented by, but substantively different from—procedural protections
that still allow questionable types of evidence to be used in capital
prosecutions.28 The proposal here is to categorically exclude such types of
evidence from capital prosecutions (possibly with an exception saying
“unless strongly corroborated by evidence that is independent of the
questionable sources”). The idea is to accomplish this statutorily: amend all
capital statutes to exclude, categorically, the known evidentiary sources of
wrongful conviction from capital cases. Alternatively one could propose
that capital statutes be written to explicitly exclude, from capital eligibility,
any case that is based primarily on the known sources of wrongful
conviction. Additionally, any such categorical exclusion standard should be
judicially enforceable, and immediately reviewable. To bring this idea
down to a concrete level of discussion (and criticism), a “model statute” is
appended to the end of this essay.
25. See Erik Luna, System Failure, 42 AM. CRIM. L. REV. 1201, 1207 (2005); Andrew M.
Siegel, Moving Down the Wedge of Injustice: A Proposal for a Third Generation of Wrongful
Convictions Scholarship and Advocacy, 42 AM. CRIM. L. REV. 1219, 1224 (2005).
26. See PRESIDENT’S COMM’N ON LAW ENFORCEMENT AND ADMIN. OF JUSTICE, THE
CHALLENGE OF CRIME IN A FREE SOCIETY (1967).
27. Although the title of a recent article sounds like my own, it addresses the Supreme
Court’s “categorical exclusion” of juveniles and the mentally retarded from capital eligibility, and
advances ideas quite different from those found here. See Dora W. Klein, Categorical Exclusions
from Capital Punishment: How Many Wrongs Make a Right?, 72 BROOK. L. REV. 1211 (2007).
28. Professor Leo and his colleagues propose, for example, that very stringent rules be used
to “screen out” possibly false confessions. Leo et al., supra note 7, at 512. It is possible that their
proposals would eliminate all false confessions. Nevertheless, their proposals do not amount to a
categorical exclusion of the use of confessions, when a confession is the primary evidence against
the accused, in death penalty cases.

LITTLE M ACRO.DOC

108

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

C. WHY GAMBLE WITH THE POSSIBILITY OF WRONGFUL CAPITAL
CONVICTIONS?
To date, the statutory or rules response to wrongful convictions has a
bit of a casino gambling character. What we say is: we know the sources of
wrongful conviction. But we can’t ban them entirely, because they
(confessions, eyewitness identifications, so-called expert forensic evidence,
and criminal informants) are often essential, and entirely reliable,
ingredients of many acceptable criminal prosecutions. Not all eyewitness
identifications, confessions, expert testimony, or even criminal informant
testimony, are wrong or unreliable.29 We can’t remove things like
confessions, identifications, forensic testimony, or dirty witnesses from
criminal prosecutions entirely: they are sometimes the backbone of the
overwhelming majority of criminal convictions that we believe are accurate,
and are necessary to the societal protection that criminal prosecution
provides.30
So instead (we say), let’s set up some “screens” to filter out some
questionable evidence or convictions: improved identification and
interrogation techniques, tighter evidentiary rules, aggressive jury “counterinstructions,” more stringent capital-eligibility “aggravators.”31 That is,
since we can’t ban the known evidentiary sources of wrongful conviction,
let’s instead set up “screens” and then hope—bet—that nothing “wrongful”
gets through.
Unfortunately, as experienced capital defense counsel recognize, the
heinous facts of many (indeed, most) capital-eligible murders can
overwhelm any protective screens. As Professor Andrea Lyon, who has
tried 132 murder cases, has written, “we can all intellectualize all we want
[t]o, but [in the face of heinous facts,] we vote with our hearts first and then
our heads . . . .”32 The problem of wrongful convictions thus persists: some
29. For example, the eyewitness identification made by a victim of rape, identifying her
husband of many years, is highly unlikely to be mistaken (as opposed to possibly perjurious). In
fact, proposals designed to increase our confidence in certain types of questioned evidence might,
in fact, screen out all the inaccurate instances. If we had confidence that this were true, then a
categorical exclusion statute might, perhaps, decide to exclude only “stranger” eyewitness
identification, or unrecorded confessions or those that fail a heightened “reliability” test. See Leo
et al., supra note 7. Both Professor Chris Slobogin and Professor J. L. de Wijkerslooth (of the
University of Leiden) have offered me suggestions along this line.
30. Thus, for example, Professors Radelet, Bedau, and Putnam dismiss out of hand the idea
that a defendant’s confession could, by rule, be excluded from capital trials. See RADELET ET AL.,
supra note 7, at 279 (doing so “would require such drastic revision . . . as to be impossible to
enact”).
31. See supra notes 16-24 and accompanying text; Kansas v. Marsh, 548 U.S. 163, 206-08
(2006) (Souter, J., dissenting).
32. Andrea D. Lyon, The Negative Effects of Capital Jury Selection, 80 IND. L.J. 52, 52

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

109

bad evidence will still get through, and be relied upon despite its
unreliability, no matter what procedural screens we can imagine.
Because the “procedural improvement” gamble will still sometimes
fail, and because “death is different,” the prospect of wrongful convictions
haunts the capital case landscape.33 This is because, obviously, an innocent
capital defendant who is executed cannot be revived or recompensed if
error is later detected. We can pay millions of dollars as some form of
“compensation” to a person wrongfully imprisoned for decades.34 But the
wrongfully convicted capital defendant whose sentence is carried out is, to
be blunt, dead. When errors we know are bound to occur cannot be
corrected, it is qualitatively more important to strive to prevent them.
Currently, our wrongful conviction “screens” in capital cases address
either improved investigative techniques—trying to prevent the evidentiary
errors in the first place—or more stringent eligibility criteria—only those
accused of truly heinous murders should be exposed to the death penalty.
But if known sources of error (confessions, eyewitness IDs, junk science,
and snitches35) still arise (indeed, arise more often, some would argue36) in
particularly heinous cases, these screens do not prevent the death penalty
from being wrongfully imposed, once the screens are passed. They may
lessen the risks (at least in the case of improved investigative techniques),
but where a heinous offense is death-eligible, and involves a confession,
eyewitness identification, a criminal informant or (just for example) expert
“bite mark” testimony,37 then they do not eliminate it.
The proposal here is different from a procedural screen. Rather than
saying “you may use known evidentiary sources of wrongful conviction if
you pass our screens,” this proposal says “you may not seek (let alone
impose) a death penalty at all, if your case is based primarily on the known
(2005). Accord Kennedy v. Louisiana, 128 S.Ct. 2641, 2661 (2008) (barring death penalty for
even aggravated child rape because, inter alia, the facts of such a rape will describe “a crime that
in many cases will overwhelm a decent person’s judgment”).
33. The “death is different” rationale was first coined by Justice Stewart in Furman v.
Georgia, 408 U.S. 238, 306 (1972) (Stewart, J., concurring), and adopted by the Court in
Woodson v. North Carolina, 428 U.S. 280, 305 (1976).
34. See BORCHARD, STATE INDEMNITY FOR ERRORS OF CRIMINAL JUSTICE, supra note 5;
see also John Martinez, Wrongful Convictions as Rightful Takings: Protecting “LibertyProperty,” 59 HASTINGS L.J. 515 (2008).
35. See supra notes 10-15 and accompanying text.
36. So argue Professor Gross and his co-authors. See Gross et al., supra note 2, at 532 (for a
number of reasons, “false convictions are more likely to occur in murder cases, and [are] much
more likely in death penalty cases, than in other criminal prosecutions” (emphasis omitted)).
37. See, e.g., Bite Mark Evidence Disputed in Murder Cases, ASSOCIATED PRESS, Feb. 29,
2008, available at http://www.ottawamenscentre.com/news/20080229_evidence.htm (last visited
Sept. 1, 2008) (DNA exoneration of two men on Mississippi death row, whose rape and murder
convictions were based, in part, on testimony from an alleged “authority on forensic odontology”).

LITTLE M ACRO.DOC

110

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

evidentiary sources of wrongful conviction.” This would be true even if
“we” had little doubt about the defendant’s actual guilt. My proposal is not
a screen—it is a categorical exclusion.
When, in 2004, the Massachusetts Governor’s Council proposed its ten
“bold recommendations” for a Model Death-Penalty Code,38 its authors
asserted that “the statutory list of ‘aggravating circumstances’ is the one and
only place, in the entire death-penalty system, where substantive limits can
be imposed on the death penalty that are not discretionary.”39 Not so. This
essay proposes a different type of substantive limit on death penalty
prosecutions: exclusion of reliance on the types of evidence known to,
sometimes, lead to wrongful convictions.40 This would place a statutory,
nondiscretionary, substantive limit on those cases in which a death penalty
may be sought, based on the type of evidence the prosecution proffers. It
would categorically exclude the type of murder prosecutions that have been
most troubling in the wrongful conviction context: those based primarily on
eyewitness identifications, confessions, criminal “snitches,” and unreliable
(“junk”) forensic evidence.
D. WHAT ARE THE OBJECTIONS TO THIS PROPOSAL?
The arguments in favor of my proposal are relatively simple: it would
reduce or eliminate wrongful capital convictions, while still allowing clear
instances of capital guilt to go forward. Because capital cases seem
particularly prone to the dangers of wrongful conviction evidence, they are
an appropriate category in which to try a categorical exclusion remedy.
But what objections are likely to my proposed categorical evidentiary
exclusion statute? There would be many, no doubt. In fact, the response of
prosecutors at Southwestern’s Wrongful Conviction Symposium was
basically “are you kidding?” and (rhetorically) “what is left?” Here I will
try to sketch and briefly respond to the objections that seem most apparent.
Please review the Model Statute presented at the end of this essay before
continuing.
38. Symposium, Report of the [Massachusetts] Governor’s Council on Capital Punishment,
80 IND. L.J. 1 (2005) [hereinafter Massachusetts Report].
39. Id. at 9.
40. Note that this proposal does not ban the known evidentiary sources of wrongful
conviction entirely. That, too, would be a substantive limit on capital prosecutions. However, it
seems too over inclusive to prohibit all eyewitness identifications or confessions, for example,
even in death penalty prosecutions. When such evidence seems particularly reliable and/or is
strongly corroborated by other reliable evidence that is independent of the questionable evidence,
the normal evidentiary presumption of all criminal trials (i.e., that all reliable, probative, and
nonduplicative evidence should be admitted) is too strong to ignore.

LITTLE M ACRO.DOC

2008]

1.

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

111

It Effectively Eliminates the Death Penalty.

Interestingly, a number of folks who have considered my proposal have
voiced the initial objection “but that wouldn’t leave any death penalty cases
at all!” When a seemingly solid two-thirds of Americans still support the
death penalty, despite the now-common knowledge of actual innocence
exonerations, such a result could be described as, at least, undemocratic.41
I find this initial objection revealing. Are we so dependent on
questionable evidence in capital cases—eyewitness “stranger”
identifications, unrecorded confessions or confessions from the mentally
deficient, unproven “science,” or jailhouse snitches—that we cannot
prosecute capitally without it? That some intelligent observers would think
so seems telling in itself. But I doubt that it is so. Obvious capital
prosecutions—Timothy McVeigh, the Oklahoma City Federal Building
bomber is my favorite posterchild42—would go forward without hindrance
from my proposal.43 The most compelling cases for the death penalty, with
guilt clearly established in reliable ways, would remain. Meanwhile,
eliminating those capital cases that lack independent reliable evidence is the
very point of the proposal.

2

It Would Dramatically Reduce the Number of Capital Prosecutions.

I do not know the empirical impact that my proposal, if enacted, would
have. But of course some reduction in capital prosecutions would result:
that is the very point of the idea, to eliminate those capital prosecutions that
are based primarily on evidentiary sources we know to have sometimes
caused wrongful convictions. So, would such a categorical exclusion mean
that some accurate capital convictions would be “lost” by virtue of their
evidentiary basis? Yes. But that is, fundamentally, not a persuasive
objection to a categorical exclusion system enacted for prophylactic
purposes.
First, of course, a categorical exclusion from capital eligibility avoids
only the death penalty, not conviction and lengthy imprisonment.44 But
41. See Kansas v. Marsh, 548 U.S. 163, 187 (2006) (Scalia, J., concurring) (many European
countries would “still have it [the death penalty] if the democratic will prevailed.”); Rory K.
Little, Why a Federal Death Penalty Moratorium?, 33 CONN. L. REV. 791, 798 (2001).
42. See Little, supra note 1.
43. Other prominent examples would include the D.C. sniper murders from 2002 and the
Atlanta courthouse murders in 2006, see infra note 73.
44. Note that this response exposes a different objection to my proposal, in the opposite
direction: why limit it to just capital cases? If we know that certain evidentiary sources
sometimes (albeit perhaps not frequently) produce wrongful convictions, why not categorically

LITTLE M ACRO.DOC

112

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

more fundamentally, when a risk of error cannot be entirely eliminated, the
question becomes on whom should that risk fall. A capital case solution of
categorical evidentiary exclusion simply chooses to never allow the risk of
questionable evidence to fall on the capital defendant. As Justice Harlan
wrote in Winship,45 even though the “beyond a reasonable doubt” standard
for criminal cases results in some “lost” convictions that would have been,
nevertheless, factually accurate, we accept the rule. The unusually high
standard of proof is a constitutional requirement, for criminal cases,
because it is “a fundamental value determination of our society” that a
number factually guilty criminal offenders should go free, rather than risk
even a few wrongful convictions.46
The point is that, ultimately, our concern about wrongful convictions
has to do with substantive guilt, not merely the fairness of the process. This
is why the Supreme Court has always reserved an “actual innocence”
exception for capital cases.47 No matter how procedurally fair and
evidentiarily unobjectionable a trial and conviction may appear, we will not
allow an actually innocent defendant to be executed. Thus procedural
screens and stringent capital eligibility criteria are not enough. We need, I
propose, categorical capital exclusions for the known evidentiary sources of
wrongful convictions.

exclude all criminal prosecutions based primarily on those sources? Other than the shibbolethic
response that “death is different,” I do not consider further responses to this argument here. Given
the difficulty of actually enacting anything like this essay’s proposal even in the capital context, it
is premature to consider its pros and cons outside that context. But it is a fair question, as all the
literature surrounding the controversial evidentiary sources, not limited to capital cases, attests.
45. In re Winship, 397 U.S. 358, 369 (1970) (Harlan, J., concurring).
46. Id. at 372. Justice Harlan also pointed out that we do not similarly unbalance the scales
in civil cases. Id. Yet it is immediately obvious that some civil cases are more “serious” than
some criminal cases—a civil effort to take away someone’s home, children, or life savings, can
seem more serious than a misdemeanor criminal prosecution that will result in probation. When I
ask my criminal law students every year whether they would take a million dollars in return for
one night in jail, they unanimously say “yes” (such is the force of large student loans)—at least
until I remind them that they might not get to choose which jail. The point being that we apply the
high Winship standard to all criminal cases, even non-serious ones, for shared policy reasons.
47. See House v. Bell, 547 U.S. 518, 520 (2006) (“the Court assumed without deciding”);
Herrera v. Collins, 506 U.S. 390, 417 (majority opinion) (1993) ; Herrera, 506 U.S. at 419, 427
(O’Connor, J., concurring). It is true that the Supreme Court has never actually applied this
exception to block an execution, and that the two newest Justices (Chief Justice Roberts and
Justice Alito) have yet to speak on the subject. Chief Justice Roberts did, however, also note the
exception and reserve decision on the question, in House. 547 U.S. at 555-56 (Roberts, C.J.,
concurring in part and dissenting in part).

LITTLE M ACRO.DOC

2008]

3.

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

113

Such Categorical Evidentiary Exclusion is Unprecedented.

Some may argue that in no other context do we categorically exclude
the filing of criminal charges based on the character of the evidence the
prosecution proposes to use. This is close to accurate; it is difficult (though
not impossible) to come up with similar categorical exclusions in the
criminal context. However, originality is hardly a substantive objection.
All new ideas have to start somewhere. Originality may often signal
crackpot-edness, but not always. The idea needs to be examined on its
merits.
Moreover, upon brief consideration, at least two similar and wellaccepted, evidentiary exclusions do present themselves in the criminal
context. First, as we all know, in virtually all U.S. jurisdictions, evidence
from so-called “lie-detector” or “polygraph” devices is categorically
excluded from all criminal prosecutions.48 Why is lie detector evidence
generally excluded? Because it is viewed as unreliable.49 Even though a lie
detector machine may sometimes (even often) be accurate, we distrust it
sufficiently to categorically exclude it.50
In addition, the U.S. Constitution enshrines a substantive, categorical
ban on a particularly serious type of criminal prosecution, unless the
government can produce particularly reliable evidence. In Section 3 of
Article III, the Framers placed a categorical evidentiary ban on certain types
of Treason prosecutions:
No Person shall be convicted of Treason unless on the Testimony of two
51
Witnesses to the same overt Act, or on Confession in open Court.

As Professor Akhil Amar notes, this Constitutional command creates
evidentiary requirements for Treason cases “above and beyond those of all
other [criminally] accused persons . . . .”52 Interestingly, this constitutional
48. The famous Frye case excluding polygraph evidence was, in fact, one of the leading
precursors for our current suspicion of un-validated “junk” science. See Frye v. United States,
293 F. 1013, 1014 (D.C. Cir. 1923) (holding that the lie detector test at issue had not gained
enough “scientific recognition,” or general acceptance among the scientific community to justify
admission). See also People v. Kelly, 549 P.2d 1240 (Cal. 1976) (holding that experts testifying
must be qualified to do so and when experts testify, the correct scientific procedures must be used
in the particular case).
49. James R. McCall, Misconceptions and Reevaluation—Polygraph Admissibility After
Rock and Daubert, 1996 U. ILL. L. REV. 363 (1996); see also Martin A. Hewett, A More Reliable
Right to Present a Defense: The Compulsory Process Clause After Crawford v. Washington, 96
GEO. L.J. 273 (2007).
50. See also Kennedy v. Louisiana, 128 S.Ct. 2641, 2663 (2008) (noting that the “special
risks of unreliable testimony” should be considered in evaluating the constitutionality of death
penalty statutes).
51. U.S. CONST. art. III, § 3, cl. 1.
52. AKHIL REED AMAR, AMERICA’S CONSTITUTION: A BIOGRAPHY 244 (2005). Accord

LITTLE M ACRO.DOC

114

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

direction evinces a suspicion of two of the evidentiary sources often found
in the modern actual innocence cases: confessions (excluding all
confessions, in treason cases, except those made “in open Court”), and
perjury (requiring two witnesses to the overt act required to prove treason).
When the Constitutional Convention revised the Treason clauses to
require the two-witnesses rule, James Madison’s notes indicate that “Docr
[Benjamin] Franklin wished this amendment to take place – prosecutions
for treason were generally virulent; and perjury too easily made use of
against innocence.”53 As Professor Gross and his co-authors indicate,
perjury (particularly perjurious eyewitness accounts) is a major evidentiary
source of wrongful convictions.54 Ben Franklin and his Framing peers
anticipated this problem—questionable evidence used “against
innocence”—and sought to address it, by a categorical ban on the evidence
of single eyewitnesses to treason’s overt acts.55
Treason was, of course, a capital offense then (and it remains so
today).56 If the Framers saw fit to categorically prohibit certain capital
Hamdi v. Rumsfeld, 542 U.S. 507, 560 (2004) (Scalia, J., dissenting) (constitutional Treason
clause “establishes a heightened proof requirement”); Willard Hurst, Treason in the United States,
Part II: The Constitution, 58 HARV. L. REV. 395, 399 (1945) (noting “the framers’ preoccupation
with limiting the scope of the crime.”). See also JAMES WILLARD HURST, THE LAW OF TREASON
IN THE UNITED STATES: COLLECTED ESSAYS (1971). Recent examinations of the Treason clauses
appear in Carlton F.W. Larson, The Forgotten Constitutional Law of Treason and the Enemy
Combatant Problem, 154 U. PA. L. REV. 863 (2006), and Benjamin A. Lewis, Note, An Old
Means to a Different End: The War on Terror, American Citizens . . . and the Treason Clause, 34
HOFSTRA L. REV. 1215 (2006).
53. Hurst, Treason in the United States, Part II: The Constitution, supra note 52, at 403
(emphasis added). Professor Hurst also notes that James Wilson, who had been on the defense
team in the Pennsylvania treason trials in 1778, “recognized the dangers of putting an impractical
burden on the prosecution” when the two-witness rule was discussed at the convention.
Nevertheless, he apparently voted for it. Id. at 404-05.
54. Gross et al., supra note 2, at 543.
55. Of course, Dr. Franklin and the Framers were operating on a strong common law and
British statutory tradition, incorporating a two-witness rule for treason (and, at times, other serious
offenses) that can be traced back to the thirteenth century. See L. M. Hill, The Two-Witness Rule
in English Treason Trials: Some Comments on the Emergence of Procedural Law, 12 AM. J.
LEGAL HIST. 95, 96 (1968). But such an historical lineage does not provide a rationale to
undercut the rule—rather, it supports it.
56. An Act for the Punishment of Certain Crimes Against the United States, ch. 9, § 1, 1 Stat.
112 (1790); see Rory K. Little, The Federal Death Penalty: History and Some Thoughts About the
Department of Justice’s Role, 26 FORDHAM URB. L.J. 347, 362-63 (1999). Today: 18 U.S.C. §
2381 (2008); accord Kennedy v. Louisiana, 128 S.Ct. 2641, 2659 (striking down death penalty for
aggravated child rape, but noting that “[o]ur concern here is limited to crimes against individual
persons” and “[w]e do not address, for example, . . . treason, espionage, terrorism, and drug
kingpin activity”). It is still within the memory of the living that two persons convicted of
espionage (a form of treason) against the United States (Julius and Ethel Rosenberg) were
executed in 1953. See United States v. Rosenberg, 109 F. Supp. 108 (S.D.N.Y. 1953) (reaffirming
death sentences).

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

115

criminal prosecutions unless based on particularly reliable evidence, we
should at least consider doing no less today.

4. We Don’t Need It, Since the Existing System Catches All Wrongful
Convictions.
In his remarkable 2006 concurrence in Kansas v. Marsh, Justice Scalia
asserts that the recent revelation of actually innocent exonerees
“demonstrates not the failure of the system but its success.”57 This
characterization is, of course, debatable. But even assuming its accuracy,
the proposal here would block wrongful capital prosecutions, not just
convictions. It seems self-evident that protecting a defendant, particularly
an innocent one, from the special anxieties of a death penalty prosecution
and years or even decades on death row, before innocence is finally
detected, is a better system.58
But this objection also suggests that there is an absence of empirical
demonstration that any wrongful capital prosecutions have occurred since
1989, when DNA-supported actual innocence evidence brought the issue to
popular, and legislative, attention. That is, aren’t all the “actual innocence”
convictions old, pre-dating the adoption of newer reforms? Why not give
all the recent reforms time to “kick in,” before implementing a more radical
idea that might not, in light of other recent improvements, be necessary?
An absolute rebuttal to this suggestion is impossible, because perhaps it
is correct—most wrongful conviction cases are not recent. So we may not
know for 20 years. Perhaps recent reforms will eliminate all future
wrongful capital convictions. On the other hand, perhaps not—and then we
can reverse the rhetorical objection. Why wait? Why wait, when we know
what the evidentiary sources of wrongful conviction have been? As noted
above, none of the recent reforms purport to entirely eliminate these
evidentiary sources (eyewitness identifications, confessions, un-validated
science, and criminal informants). Why do we want to continue to allow
them in death penalty prosecutions? The only answer is that we want to
avoid a “windfall” (that is, not giving a deserved death penalty) to actually
guilty capital defendants. This goes back to objection number 2 above, and
is rebutted only if Justice Harlan’s “fundamental value determination” is
57. 548 U.S. 163, 193 (2006) (Scalia, J., concurring).
58. Relevant here is the concern expressed repeatedly by Justices Stevens and Breyer that
spending many years on death row might, by itself, be unconstitutional under the Eighth
Amendment. See Foster v. Florida, 537 U.S. 990, 991-93 (2002) (Breyer, J., dissenting from
denial of certiorari) (citing prior dissents).

LITTLE M ACRO.DOC

116

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

accepted here.

5.

Why Not at Least Narrow the Proposal?

The fact is, we don’t actually question all eyewitness accounts, or
confessions, or science or even criminal informants.59 Why not narrow the
proposal to allow at least those capital prosecutions based on identifications
made in blind, sequential lineups? Why not allow at least videotaped
confessions not involving deception or lengthy interrogations? Validated
science? Fully corroborated criminal informants?
Of course, any adopted “categorical exclusion” capital statute could—–
and, realistically, probably would—allow some such evidence. But this
objection is really not so much an objection as it is an endorsement of my
proposal, with suggestion of “friendly amendments.” Certainly if we are
convinced that more reliable types of evidence are not going to lead to
wrongful convictions, we ought not exclude them.60 Thus the appended
Model Statute presents an option that would allow evidence that has “strong
and independent” corroboration, as well as validated science (which is by
definition, not “junk” science61).
However, it is important to note how vital the “independent”
requirement for such evidence is. It means something like “entirely
unconnected to the questionable evidence or the sources for that evidence.”
And if you have such evidence, query why you would also need the
questionable evidence at all? Because of the recognized “interlocking
corroboration” phenomenon in wrongful conviction cases—where one type
of questionable evidence is used to “corroborate” another (a jailhouse snitch
“corroborates” a single eyewitness, for example)—any corroborating
evidence used to support an exception to the statute must be truly
independent of the presumptively excluded evidentiary sources. The
meaning of “independent” would have to be worked out over time, but the
59. See supra note 2 (“undisputable” that “not all convictions are ‘wrongful’”) and supra
note 29 (noting possible reliable sources of such evidence).
60. There is some recent evidence, however, that we really do not know what types of
evidence within the questionable categories are “more reliable.” See Gary L. Wells, Field
Experiments on Eyewitness Identification: Towards a Better Understanding of Pitfalls and
Prospects, 32 LAW & HUM. BEHAV. 6 (2008) (asserting that sequential lineups may be,
empirically, less reliable than contemporaneous lineups). But cf. Wendy Alberts, Steering in the
Eyewitness Identification Procedure 25 (May 2008) (unpublished Masters Thesis, University of
Leiden, Law Faculty, on file with author) (criticizing the Wells study because it failed to control
for “blind” officers as it compared sequential versus contemporaneous line-ups).
61. See Faigman, supra note 21; David E. Bernstein, Junk Science in the United States and
the Commonwealth, 21 YALE J. INT’L L. 123 (1996).

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

117

concept is vital to preventing wrongful conviction evidentiary sources from
infecting capital cases.
Thus, in general, if more sophisticated, narrowing, definitions of the
known evidentiary sources of wrongful conviction are possible, and leave
us with great confidence in the accuracy of the evidence, then by all means
they should use them in the statute. Section 2 of the model statute has been
modified since my initial draft to try to capture this idea. If the legislature
enacts other statutes requiring eyewitness identifications and confessions to
be obtained only in recognizably reliable ways, then that evidence could be
used in a capital prosecution. However, the model statute might also
reasonably be endorsed without the exception provided in Section 2.

6. The Proposed Statutory Language is Unmanageably Vague and
Manipulable.
“Primarily” based on? “Strongly corroborated”? Come on! What
prosecutor or judge can’t manipulate that language toward the unjust
conviction? Of course, this is the objection to virtually any written statute
that seeks to limit governmental discretion. It objects generally to the
limited power of language to effectively regulate, rather than to the
substance of this particular statute.62 Lawyers are semanticists, and it is
human nature to stretch language in the face of hard cases.63 As in all
statutory contexts, we will have to depend on the good faith of interpreting
courts, and prosecutors, and accept that human failings will occur.
Nevertheless, suggestions for more clear, precise, and unavoidable language
are welcome. I am merely putting the ball in play. Others will have to put
it in the hoop.

7. The Prosecution Shouldn’t Have to Expose Its Evidence Before
Charging.
This objection seems somewhat quaint, in today’s world of relatively
generous pretrial criminal disclosure requirements and the intensive review
that death penalty prosecutions now receive in most jurisdictions.64 The
62. See Little, supra note 56, at 490 (discussing “The Inevitable Manipulability of
Language”).
63. Thus Justice Holmes’ famous aphorism about hard cases making bad law. N. Sec. Co. v.
United States, 193 U.S. 197, 400 (1904) (Holmes, J., dissenting).
64. “For example, some prosecutors’ offices employ committees of experienced prosecutors
to make death penalty decisions.” Bruce A. Green & Fred C. Zacharias, “The U.S. Attorneys

LITTLE M ACRO.DOC

118

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

fact is, most effective capital defense lawyers today know the government’s
evidence before charges are filed, and certainly well before trial begins (if
they don’t, they are ineffective, which is its own constitutional objection).
A prosecutor who seeks to take on the large challenge of a capital
prosecution, with such serious implications and at such great public
expense, should not be fearful of exposing his or her primary evidence to
public inspection.
Moreover, my sketched model statute would allow the prosecution
alternatives: the government could either review its evidence internally, and
then file charges if it believed it was compliant with the statute, or seek precharge judicial approval. Thus a prosecutor could file charges without
exposing his or her primary evidence to pre-filing examination; but the
evidence would be subject to speedy judicial scrutiny soon thereafter. If
you accept the idea of categorical evidentiary exclusions, then this is as it
should be: we ought not endorse capital cases going forward, at great public
expense, if based primarily on types of evidence we know have led to
wrongful capital convictions in the past. See Objection number 2, above.

8. It’s the Jury’s Job to Evaluate the Prosecution’s Evidence, Not a
Judge’s.
More powerful, perhaps, is the objection that it is the jury that
evaluates the probative value of evidence in the American criminal justice
system, not lawyers. The defense is permitted to attack the reliability of all
evidence upon which the prosecution relies. Specific evidence that is
shown to be unreliable can certainly be excluded, upon a particularized
motion and proof in an individual case. But the defense cannot exclude
evidence simply by asserting that it “might” be unreliable, let alone that it
has been shown to be unreliable in other cases, but not in this one. The jury
is entitled to view all the relevant evidence, once it passes muster under the
evidentiary rules of general application.
Of course, the “evidentiary rules of general application” are, in fact,
categorical exclusions based on experienced evaluation of unreliable types
of evidence. For example, we seek to generally ban hearsay evidence,
because it can be unreliable, even though it may be entirely accurate in an

Scandal” and the Allocation of Prosecutorial Power, 69 Ohio St. L.J. 187, 195 n.35 (2008) (and
citing authorities). See also Little, supra note 56, at 407-28 (detailing federal Department of
Justice death penalty review process). This is increasingly necessary given the skyrocketing costs
of capital cases. See infra note 73.

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

119

individual case.65
But the serious and unusual impact of a rule that categorically excludes
some serious criminal prosecutions based on a type of evidence cannot be
denied. At bottom, there is no doubt that endorsing this essay’s proposal
requires a “fundamental value determination” ala Justice Harlan.66 If we
are unwilling to make the decision that capital cases are too serious, and
fundamentally different, from other criminal cases, then different
evidentiary rules probably ought not apply.67
However, the evidence that we do consider death penalty cases to be
“different” is increasingly abundant. There is no American jurisdiction that
has considered the death penalty that does not have a special statute, or set
of statutes, to address capital prosecutions.68 Moreover, many of those
statutes are under intensive review in many jurisdictions, since Illinois’
Governor commuted the death sentences of all 170 of Illinois’ death row
occupants in 2000 69 and Massachusetts’ Governor created a blue-ribbon
Commission to propose a “model” best-practices death penalty statute in
2003.70 As Justice Souter has recently observed, in light of the recent and
undisputed DNA exonerations in capital cases, “[w]e are . . . in a period of
new empirical argument about how ‘death is different.’”71 This essay
simply presents one more possible statutory revision that could serve to
reduce the number of wrongful convictions in the capital arena.

9.

This Won’t Eliminate All Wrongful Capital Convictions.

No doubt this is true. So long as bad prosecutorial and defense
lawyering, law enforcement corruption and witness misconduct persists to
any degree, all the statutes in the world will not prevent all wrongful
65. And just as with the two-witness Treason rule, see supra notes 51-56 and accompanying
text, the hearsay rules are ultimately constitutionally based in the due process concerns of the
Framers, as expressed in the Confrontation Clause of the Sixth Amendment. See Crawford v.
Washington, 540 U.S. 964 (2003).
66. See supra note 46 and accompanying text.
67. Although this point just raises the broader question that was raised, and then put aside in
note 44, supra.
68. Rather than provide another horribly long footnote to support this assertion, the author
swears that the results of a fifty-State survey performed by his research assistant Todd Daloz
(Hastings Class of 2009) are on file with the author.
69. See Marquis, supra note 7, at 503 (presenting a somewhat uncharitable account of
Governor Ryan’s actions).
70. See Massachusetts Report, supra note 38.
71. Kansas v. Marsh, 548 U.S. 163, 219 (2006) (Souter, J., dissenting) (quoting Gregg v.
Georgia, 428 U.S. 153, 188 (1976)).

LITTLE M ACRO.DOC

120

10/6/08 11:23 PM

SOUTHWESTERN UNIVERSITY LAW REVIEW

[Vol. 37

convictions. This proposal seeks only to reduce, in a categorical way, such
convictions, and eliminate the spectacle of death penalty prosecutions going
forward based on evidence that objective observers know is of questionable
value. Yet some ineffective defense lawyers (if sleeping, for example 72)
will not enforce even the proposed statute, rendering it a substantive nullity
in some cases.
On the other hand, if adequately compensated, experienced capital
defense lawyers were required in every case—although how we will pay for
that is a difficult question, see note 73 below—then bad prosecutorial
lawyering, police corruption, and witness misconduct are all far more likely
to be detected and rebuffed. And as noted above, an effective remedial
regime to the problem of wrongful convictions is a mosaic, not a onesolution-beats-all situation.

CONCLUSION
Unless the death penalty disappears in the United States because it is
too expensive to defend,73 capital cases will continue to remain a
particularly disturbing subset of all “actual innocence” cases. According to
various sources, over 70 capital defendants have been sentenced to death in
the past 20 years, only to be later exonerated.74
This essay presents a simple idea: since we know what types of
evidence have led to most “actual innocence” convictions, why not
statutorily ban those evidentiary sources from capital prosecutions (except,
perhaps, where other independent evidence strongly corroborates guilt)? Of
course, as the Supreme Court once said of the Treason clauses, an idea with
a “superficial appearance of clarity and simplicity” can nevertheless be
“packed with controversy and difficulty.”75 Perhaps that is the case here.
72. Burdine v. Johnson, 262 F.3d 336 (5th Cir. 2001) (en banc).
73. See Baze v. Rees, 128 S.Ct. 1520, 1548 (2008) (Stevens, J., concurring) (noting “the
enormous costs that death penalty litigation imposes on society”); Kozinski & Gallagher, Death:
The Ultimate Run-On Sentence, 46 CASE W. RES. L. REV. 1 (1995). For one example, see Jeffrey
Toobin, Death in Georgia: The high price of trying to save an infamous killer’s life., NEW
YORKER, Feb. 4, 2008, at 32 (detailing the Brian Nichols capital case, in which the defendant
undoubtedly killed a judge, court reporter and sheriff’s deputy, but whose offer to plead guilty was
rejected in order to pursue the death penalty, and in which the judge suspended all proceedings in
late 2007 when Nichols’ indigent defense costs rose to over $1.2 million and the County ran out of
money to pay). Nichols seems undeniably guilty of the murders (whether his mental condition is
eligible for, or deserves, the death penalty, remains to be seen). But the County at issue is, so far,
not able to pay for the effective defense he is entitled to under the Sixth Amendment.
74. Gross et al., supra note 2, at 529 tbl.1 (74 death sentence exonerations); Kansas v. Marsh,
548 U.S. 163, 209-10 (Souter, J., dissenting); but cf. Liptak, supra note 2.
75. Cramer v. United States, 325 U.S. 1, 46-47 (1945).

LITTLE M ACRO.DOC

2008]

10/6/08 11:23 PM

CATEGORICAL EXCLUSION OF EVIDENCE

121

But the seriousness of the question suggests that the idea should at least be
fairly considered.
MODEL STATUTE FOR CATEGORICAL EXCLUSION OF KNOWN
EVIDENTIARY SOURCES OF WRONGFUL CONVICTIONS
1. No capital charge may be [filed] [prosecuted] in any case where the
prosecution’s case in chief rests primarily on
(a) eyewitness [stranger] identification testimony;
(b) a confession;
(c) a criminal informant; or
(d) unvalidated forensic evidence,
2. The only exception to Section 1 above is when there is independent
evidence that strongly corroborates the evidence listed in (a) through
(d) above [or perhaps “if the evidence has been reliably obtained under
statutes governing eyewitness identifications and confessions”].
3. Absent such strong, independent corroboration, as specified in
Section 2 above, a combination of evidence drawn from (a) through (d)
above is also an insufficient basis on which to [file] [prosecute] a
capital charge.
4. Upon request by counsel for any capital defendant, the prosecutor’s
case-in-chief evidence shall be reviewed independently for compliance
with this statute by a separate capital case review unit within the
prosecutor’s office, and then either
(a) prior to the filing of the charge, by a judicial officer within the
jurisdiction; or
(b) immediately after filing of the charge by the judicial officer to
which the case is assigned.
5. Capital charges found to be primarily supported by evidence
specified in Section 1 above shall be dismissed.
6. Any judicial order that dismisses capital charges under this statute
may be immediately appealed by the prosecution on an expedited basis.
There is no double jeopardy objection to such an appeal. The defense
may also seek to immediately appeal a judicial order that allows the
prosecution’s capital charges to proceed, but such an appeal is
authorized only if the trial judge, or an appellate court upon proper
motion, certifies that the appeal presents a reasonably close question
under this statute.