Lacking Evidence: A Response to Jeffrey Bellin’s The Evidence Rules That Convict the Innocent

Michael Conklin*

In sum, while the numbers we have may be misleading, as of now, they point dramatically in a single direction.

—Jeffrey Bellin¹

Introduction

The preceding quote embodies much of the problem with Jeffrey Bellin’s article The Evidence Rules that Convict the Innocent. Namely, if the numbers presented by Bellin are misleading—and they are—then the direction in which they point is largely irrelevant. This Article critiques Bellin’s conflation of innocence with exoneration, presentation of misleading evidence, and failure to provide any specific solutions to the problem presented in Bellin’s article. Issues like these have plagued criminal justice scholarship for decades. This Article aims to address these issues plaguing criminal justice scholarship and provide long-overdue clarification.

“Innocent”?

Throughout the article, Bellin conflates factual innocence with wrongful conviction. For example, when discussing “factually innocent defendants,” he references exoneration data.² He claims that “DNA tests definitively established the innocence of hundreds of defendants convicted of serious crimes.”³ However, the citation provided for this claim only references DNA exonerations without definitively discussing the role of DNA.⁴ Bellin frequently refers to the National Registry of Exonerations.⁵ But even the Registry explicitly acknowledges that, while exonerees may be factually innocent, this is not required to be considered an exoneree.⁶

Bellin is by no means the first to equate innocence with exoneration. Supreme Court justices have made the same mistake. In Justice Souter’s Kansas v. Marsh dissent, he claimed that over 110 death row inmates had been released since 1973 “upon findings that they were innocent of the crimes charged.”⁷ Justice Scalia corrected him, explaining that most of these 110 allegedly “innocent” people were released from death row due to technical issues, such as inadmissible evidence, double jeopardy, or the death of key witnesses.⁸

Identifying factually innocent defendants is complicated by the subjectivity in the elements that make up crimes. For example, some have claimed that Ernest Dobbert is innocent of the first–degree murder of his wife, not because he did not kill her but because he did not deliberate long enough before doing so.⁹ It is ultimately unknowable exactly how long he in fact deliberated before killing her. And even if this were ascertainable, his ultimate innocence or guilt would still be unknowable because exactly how much deliberation is required to constitute first–degree murder is largely subjective.

Equating factual innocence with wrongful conviction distorts the tradeoff the criminal justice system makes in exonerating people: sometimes the guilty go free, to. Understanding that tradeoff is essential to considering the rules of criminal evidence. Exaggerated claims of falsely convicted “innocent” defendants distorts a legitimate issue; namely, the ideal tradeoff in the criminal justice system between letting the guilty go free and convicting the factually innocent. Understanding the tradeoffs involved in this balance is essential when considering the rules of criminal evidence but requires an accurate understanding of factual innocence.

Bellin also claims that examples of innocent defendants being convicted “tarnish” the idea that criminal trials are the “gold standard of American justice.”¹⁰ Bellin’s position falsely presumes this “gold standard of American justice” requires absolute perfection. A gold standard signifies “the supreme example of something against which others are judged or measured.”¹¹ Therefore, to justify the claim that criminal justice trials are not the gold standard, Bellin would need to show a better alternative. Simply pointing to instances in which an innocent defendant was convicted—without proposing how this could be minimized without impacting the frequency of false acquittals—falls short.

Evidence?

The irony in Bellin’s article about evidence is the scant evidence he provides to support his own claims. Bellin opines that it is necessary to “evaluate whether the evidence rules work” by performing an “accuracy-test [on] the rules of evidence.”¹² But such a test was never conducted. Instead, Bellin simply recited statistics regarding what percentage of cases with “innocent” defendants involved different types of evidence.¹³ For example, he noted that “a false confession appeared in 13[%] of convictions that led to exonerations”¹⁴ and that “22[%] of the convictions of DNA exonerees involved false informant testimony.”¹⁵

These statistics are misleading for multiple reasons. First, given Bellin’s continuous conflation of innocence and wrongful conviction—and his erroneous use of the term “innocent defendants” when introducing these statistics—readers are likely to incorrectly assume the statistics refer to factually innocent defendants. Second, the methodology for calculating these statistics is highly problematic. For example, consider the claim that thirteen percent of convictions that led to an exoneration involved a false confession.¹⁶ This statistic is the result of simply counting every case involving any confession and a later exoneration.¹⁷ This is a poor proxy for analyzing confessions for two reasons. First, it includes factually guilty defendants who were exonerated; and whose confessions were not false. Second, it excludes false confessions that did not lead to an exoneration. This methodology is further problematic because it includes cases in which the confession was never even presented in court.¹⁸ Finally, just because a trial contained a certain issue, such as a confession,¹⁹ a jailhouse informant,²⁰ or an eyewitness identification,²¹ does not mean that that issue was the proximate cause of the conviction.

Additionally, the statistics Bellin cites are not from a random sample.²² For example, the false confession statistic refers to only thirteen percent of 1,600 cases over a sixteen-year period, amounting to an average of only thirteen instances per year.²³ And again, not all of these defendants were factually innocent, nor were they all convicted because of their false confessions. Here, it is important to note that the standard for rules of evidence is not the complete absence of any false convictions. In a country where over seven million arrests are made every year,²⁴ advocating for changes to the rules of evidence based on thirteen potentially false confessions per year suggests a refusal to see the big picture.

Solution?

Given the problematic nature of Bellin’s evidence, it is perhaps laudable that he does not advocate for any specific policy proposals. Rather, he calls for “a general reevaluation of evidence rules and doctrines.”²⁵ Elsewhere in the article he is more modest, only calling for “[a] shift in attention,” which “does not mean that any particular rule must be changed.”²⁶ He concludes with a nondescript call to action for “scholars, judges, and policymakers to examine the research on false convictions and assess what went wrong.”²⁷

Bellin’s lack of any proposed solution is conspicuous. Pointing out isolated problems in a complex system and then calling upon others to assess what should be done is easy. It is far more difficult to consider potential solutions and how they would affect the tradeoff between convicting the innocent and not convicting the guilty. For example, there are already numerous safeguards in place to protect against false confessions.²⁸ Furthermore, the defendant is free to present evidence to the jury regarding the problematic nature of confessions. These solutions seem more than adequate when one also considers that we have designed criminal trials to favor the defense. For example, the defense is not required to prove to all twelve jurors that a confession was false.²⁹ Rather, it only needs to prove to a single juror that there is some reasonable doubt regarding the confession. This is not to say that the rules of evidence regarding confessions should not be altered. However, we should also consider the safeguards already in place and the probability that a change will allow the guilty to go free. Bellin exclusively relies on examples that support further restrictions on the rules of evidence to favor the defense.

Additionally, Bellin’s repeated use of exonerations to support altering the rules of evidence creates an interesting paradox. Namely, exoneration is an intentional part of the criminal justice system.³⁰ When an inmate who is factually innocent receives an exoneration, that demonstrates that the system is working—at least partially—as intended. Certainly, a preferable outcome would be that factually innocent defendants are never convicted in the first place. The system is never going to be perfect but exonerations show that the system has decent mechanisms for correcting itself already. As such, we need not alter the rules of evidence based on indicators that merely show the system already works as it should.

Conclusion

Bellin’s repeated conflation of innocence with wrongful conviction greatly undermines his critique of the rules of evidence. Likewise, Bellin’s examples only illustrate problems with convicting the innocent and do not address the issue of freeing the guilty. Properly understood, these examples do little to justify altering the rules of evidence because they distort the tradeoffs of doing so. Finally, the lack of any specific solution is a significant omission because it obscures the tradeoffs that would inevitably be involved.

Footnotes

* Powell Endowed Professor of Business Law, Angelo State University.↩︎

Jeffrey Bellin, The Evidence Rules that Convict the Innocent, 106 Cornell L. Rev. 305, 324 (2021). ↩︎
Id. at 308. ↩︎
Id. at 306. ↩︎
Brandon L. Garrett, Convicting the Innocent: Where Criminal Prosecutions Go Wrong 6 (2011). ↩︎
Bellin, supra note 1, at 319–20, 339, 341–42. ↩︎
Glossary, The Nat’l Registry of Exonerations, https://www.law.umich.edu/special/exoneration/Pages/glossary.aspx [https://perma.cc/Z2LD-N6WR]. ↩︎
Kansas v. Marsh, 548 U.S. 163, 209–10 (2006) (Souter, J., dissenting). ↩︎
Id. at 185–99 (Scalia, J., concurring). ↩︎
See generally Dobbert v. Florida, 432 U.S. 282 (1977). ↩︎
Bellin, supra note 1, at 350 (quoting Lafler v. Cooper, 566 U.S. 156, 186 (2012) (Scalia, J., dissenting) and similar language in Duncan v. Louisiana, 391 U.S. 145, 156 (1968)). ↩︎
Gold Standard, Dictionary.com, https://www.dictionary.com/browse/gold-standard [https://perma.cc/JG8Z-MTCU] ↩︎
Bellin, supra note 1, at 315, 317. ↩︎
See id. at 325, 330. ↩︎
Id. at 330. ↩︎
Id. at 338. ↩︎
Id. at 330. ↩︎
E-mail from Maurice Possley, Senior Researcher, Registry of Exonerations, to author (Apr. 27, 2021) (on file with author) ↩︎
See id. According to Possley, the coding manual for the National Registry of Exonerations states, “[a] confession is a statement made to law enforcement at any point before conviction, which was interpreted or presented by law enforcement as an admission of participation in the crime, or presence at the crime by someone who was not previously known to have been there, even if the statement was not presented in court.” Id. ↩︎
See Bellin, supra note 1, at 330. ↩︎
See id. at 338–39. ↩︎
See id. at 325. ↩︎
Id. at 321. ↩︎
The National Registry of Exonerations list referenced by Bellin for this claim contains cases from 1989–2015. The Nat’l Registry of Exonerations, supra note 6; see also Bellin, supra note 1. ↩︎
See Number of Arrests for All Offenses in the United States from 1990 to 2021, Statista (Oct. 20, 2023), https://www.statista.com/statistics/191261/number-of-arrests-for-all-offenses-in-the-us-since-1990 [https://perma.cc/829R-BSLA]. ↩︎
Bellin, supra note 1, at 349. ↩︎
Id. at 350. ↩︎
Id. ↩︎
See generally Welsh S. White, False Confessions and the Constitution: Safeguards Against Untrustworthy Confessions, 32 Harv. C.R.-C.L. L. Rev. 105, 111–17 (1997) (discussing constitutional safeguards against false confessions) ↩︎
Fed. R. Crim. P. R. 31(a). ↩︎
Glossary, supra note 6. ↩︎