CONVICTING THE INNOCENT: WHERE CRIMINAL PROSECUTIONS GO WRONG, by Brandon L. Garrett. Cambridge, MA: Harvard University Press, 2011. 367 pp. Cloth $39.95. ISBN: 978-0-674-05870-5.

Reviewed by Craig Hemmens, Department of Criminology & Criminal Justice, Missouri State University. Email: craighemmens [at]


The criminal justice system is intended to ferret out wrongdoers and punish them for their transgressions. Balancing the effective and efficient prosecution of criminals with the respecting of individual rights and the protection of the innocent is difficult, as Herbert Packer (1968) so eloquently explained many years ago. It has long been assumed that the adversarial system, with its many checks and balances, stages of review, and procedural and constitutional rights afforded to criminal defendants, by its nature reduces the chances of an innocent person being found guilty to virtually nothing. Sadly, the development of DNA evidence has revealed this is a false assumption. To what extent the innocent are wrongfully convicted is unknown, but recent high profile exonerations have made it clear to all that terrible mistakes do happen. Precisely how and why these mistakes happen is only now becoming clear.

In Convicting the Innocent, Brandon L. Garrett, a professor at the University of Virginia and a former lawyer with Innocence Project co-founders Peter Neufeld and Barry Scheck, provides the most comprehensive and compelling accounting to date of how DNA evidence has revealed the fallibility of the criminal justice system. In this book he examines the cases of the first 250 people exonerated by DNA evidence. While there have been previous studies of this phenomena, they have not been as thorough in their investigation (Scheck et al. 2000), or have been focused on individual cases (Grisham 2006; Smith 2008) and thus sometimes discounted as rarities. While Garrett’s work has much to recommend it, perhaps its greatest contribution is its demonstration that wrongful convictions are not freak occurrences resulting from the actions of evil-intentioned people. Rather, he shows that wrongful convictions occur for a variety of reasons, and frequently despite the fact that those involved have the best of intentions. Wrongful convictions, he aptly shows, are not a chance event but the inevitable result of flaws in the criminal justice system.

What are these flaws? They have all been identified in other research, but Garrett does an admirable job of showing how they all intersect in the cases of the first 250 exonorees. The major flaws are fallible eyewitness procedures, coerced confessions, use of jailhouse informants, and flawed forensic analysis.

Eyewitness testimony obviously is very damning – when a victim points to the defendant and says, without hesitation, “he is the one who raped me” it is hard for a defendant to overcome such apparently strong evidence. Sadly, Garrett demonstrates eyewitnesses are often mistaken. The fallibility of [*63] eyewitness testimony is old news to social scientists, who have been aware of it for years (see e.g. Kassin 1998; Smith et al. 1989), but the frequency with which the problem shows up in the 250 exoneree cases is astounding –over 75% of these cases involved mistaken eyewitness identifications. Why? Garrett finds widespread evidence of police contamination of the identification procedures, from conducting unfair lineups or showups to suggesting which suspect to identify. Often police either fail to follow procedures properly or, in many instances, the procedures themselves are imperfect and susceptible to error.

One might assume that an innocent person would not confess to a crime. Garrett’s study teaches us otherwise. Over 40 (16%) of the 250 exonerees in fact confessed to crimes they did not commit. Why? In many instances it appears police interrogators provided them with incriminating information and the suspect then adopted the information as their own. Often these suspects suffered from mental deficits that rendered them particularly vulnerable to manipulation by police. Others were juveniles. Clearly, the provision of the Miranda warnings does not, in and of itself, assure that confessions cannot be extracted from those who are susceptible to police pressure and intimidation.

Another problem that arises frequently is the use of so-called “jailhouse informants” – people housed with the defendant who testify, often in return for a lighter sentence or other benefit, that the defendant confessed the crime to them. Garrett finds this in more than 20% of the exoneree cases. Jailhouse informants have long been used, even though many courts have acknowledged the possibility of inaccurate testimony is higher. Some courts limit the use of jailhouse “snitches” while others include special jury instructions regarding their credibility. But still prosecutors anxious for a conviction rely on them, often to ill effect.

DNA evidence has been hailed as a breakthrough in the detection of crime. It is widely assumed to be infallible, although like fingerprints in an earlier time (Cole 2001), time may prove otherwise, at least to some degree. In any event, while the science may be valid, it still requires proper training and procedures. Garrett finds invalid forensic testimony in over 60% of the exoneree cases. These errors range from poorly run labs that contaminate the evidence to overstating the results to blatant lying. Significantly, Garrett also finds that the DNA evidence, when properly examined, identified the actual perpetrator in almost half (45%) of the cases. DNA, when properly used, does provide extremely useful evidence. But, as with all evidence, it is only as good as those who are using it.

In writing this book, Garrett conducted an astonishingly comprehensive examination of records in all 250 exoneree cases. He reviewed police reports, interrogation transcripts and recordings, prosecution files, trial transcripts, and appeals court opinions. He reviewed the trial transcripts of 207 of the 234 exonerees who went to trial [*64] (16 pled guilty). These documents provided a wealth of data, and Garrett, a lawyer without social science training, does an admirable job of sifting through it all, analyzing the data, and presenting it to the reader in a clear manner. Garrett does not simply provide a recounting of the numbers. What makes his book especially compelling are the stories he tells –accounts of some of the exonerees and their travails. These stories make real the numbers, and reveal a system much more flawed than many realize or are willing to admit.

Just how flawed? Virtually all of the 250 exonerees were wrongly convicted of murder, rape, or both. Such cases more often involve forensic evidence; Garrett notes whatever errors were found in these cases may exist without any way of determining it scientifically in other cases where forensic evidence does not exist, or in cases where forensic evidence has not been retained after the trial. I am not sure what is more frightening to a reader, that so many have been wrongfully convicted, only to later be exonerated, or that these 250 are likely just the tip of the proverbial iceberg – and the criminal justice system may well be the Titanic.

So what is to be done? Garrett proposes some reforms. None are new, but they might well help. He argues for double-blind eyewitness identification procedures (where the police officer conducting the identification does not know who the suspect is), video and audio recording of interrogations, full disclosure of prosecutor’s communications with jailhouse informants, and external oversight and auditing of forensic labs. All of these reforms have been proposed before, but perhaps never with such compelling evidence in support of them. Garrett’s work makes it clear that wrongful convictions are not isolated incidents or fluke accidents. They are the product of systemic failures.

In many ways, Garrett’s book lends support to the recent work of William Stuntz (2011), who argues that those seeking greater “justice” from the criminal justice system mistakenly focused on the creation of legal rules such as the Miranda warnings and the exclusionary rule, at the expense of ensuring that criminal justice procedures were truly fair and accurate. These two books, read in conjunction, force any conscientious reader to reexamine their conceptions of the American criminal justice system. There is much work to be done.


Cole, Simon. 2001. Suspect Identities: A History Of Fingerprinting And Criminal Identification. Cambridge, MA: Harvard University Press.

Grisham, John. 2006. The Innocent Man: Murder and Injustice an a Small Town. New York: Doubleday.

Kassin, Saul M. 1998. "Eyewitness Identification Procedures: The Fifth Rule." Law and Human Behavior 22(6): 649-653. [*69]

Packer, Herbert L. 1968. The Limits of the Criminal Sanction. Palo Alto, CA: Stanford University Press.

Scheck, Barry, Peter Neufeld, and Jim Dwyer. 2000. Actual Innocence. New York: Doubleday.

Smith, Abbe. 2008. Case of a Lifetime: A Criminal Defense Lawyer’s Story. New York: St. Martin’s Press.

Smith, Vickie L., Saul M. Kassin and Phoebe C. Ellsworth. 1989. "Eyewitness Accuracy and Confidence." Journal of Applied Psychology 74(2): 356-359.

Stuntz, William J. 2011. The Collapse of American Criminal Justice. Cambridge, MA: Harvard University Press.

© Copyright 2012 by the author, Craig Hemmens.