Investigating Wrongful Convictions

by John Nardizzi

On April 23, 1997, a Massachusetts judge ordered the release of Marvin Mitchell on personal recognizance. He walked out of jail a free man, the first man released from prison in Massachusetts after DNA evidence proved his innocence. Mitchell’s conviction was based on the following evidence. In September 1988, an eleven year old girl was abducted from a bus stop and raped. The victim was later driven around her neighborhood by police officers to see if she could find someone who matched her description of the assailant. The victim spotted Mitchell, who wore a mustache and goatee. Although the victim was confident that her assailant was clean shaven, Mitchell was picked up on an unfounded charge of public drinking and photographed. The victim later identified him from the picture.

Mitchell was then tried and convicted of forcible sexual intercourse and other counts. He would be incarcerated for seven years and three months before being released from prison. For his troubles, the state gave him funds for a bus ride back home. He was not eligible for any of the benefits normally offered to parolees because, ironically, he was innocent and so not eligible for parole.

Troubling cases continue to mount.

According to a study done by Professor Samuel R. Gross of the University of Michigan Law School [the Michigan study] between 1989 and 2003, a total of 328 exonerations occurred in the United States. The study counted exonerations conservatively: it did not account for mass exonerations in Los Angeles [the Rampart unit scandal] and Tulia Texas [rogue officer framed 39 drug defendants]. According to the study, 145 of the wrongfully convicted were cleared by DNA evidence, 183 by other types of evidence.

The sheer number of false convictions have exposed deep flaws in the criminal justice system--regarding, in particular, the use of informants, police interview tactics, and methods of eyewitness identification. The question raised by the release of these innocent people is obvious: how many more innocent people are rotting inside of U.S. prisons while the guilty roam free?

There are two fertile areas for investigators to assist clients in the wrongful convictions area. The first area involves post-conviction relief in a criminal case on a motion for a new trial or writ of habeas corpus. DNA is not a panacea. It exists in only a small percentage of cases, and even if it exists, the DNA evidence may have decayed, and so testing fails to render any results. In such cases where no forensic results will be forthcoming, an investigator is often hired to discover new or recanting witnesses, undisclosed exculpatory evidence, police misconduct, or some other type of evidence.

Second, there are civil cases (either state tort claims or federal civil rights cases) whereby an exoneree sues a police department or state official for misconduct that led to the wrongful conviction. In many of these cases, DNA testing has already led to the release of the client. The bar is high in these cases: state officials usually have immunity defenses, and the plaintiff must typically show intentional or reckless misconduct. Many states have passed (or are debating) legislation that compensates exonerees for the years they lost. Even in these statutory compensation schemes, an exoneree must often prove that they did not engage in misconduct or deception that contributed to their conviction. Thus, even here, an investigator is needed.

When evaluating whether to take a wrongful convictions case (whether it be a civil claim or criminal appeal), the following issues and tactics often come into play:

Planning the Investigation

One practical matter that arises is due to the simple fact that these cases are old; because they are old, the case file may have become contaminated with incorrect information. Numerous investigators, lawyers and law students have worked the case, sometimes for decades. Much of the information in the file may not be corroborated, and may reflect one person’s theory repeated in documents produced over time, and later assumed to be fact. For example, in one recent case, the date of birth for a missing witness was revealed to be wrong: an investigator years earlier had included the information as a possible lead, but later summations of the case included the birth date as fact. Once better information was developed, the key witness who had been missing for twenty years was tracked down.

Because the cases involve old information, consider using a variety of locator databases.

Thus, when possible, review original documents such as police reports, interview notes, etc. and draft a memo to file that includes your theories and impressions about the case. Drafting a task memo is one of the most productive tasks an investigator can undertake. Try to divide tasks into those that reexamine old ground, and those that break into a new arena. Keep updating the memo as the case progresses. When possible, consider scanning all documents into a PDF format, or use some sort of case management software, so that the voluminous case file can be searched by keyword. This allows for scanning reams of material quickly, (for example, all mentions of an address on Jones Street, or all mentions of a witness named Sam Langford).

Eyewitness Error

Eyewitness error is far and away the single greatest factor in wrongful convictions. The Michigan study concluded that in 64% (209 out of 328) of the exoneration cases, at least one witness misidentified the defendant. Eyewitness error was most apparent in rape cases: almost 90% of the rape cases involved eyewitness error.

Eyewitness error is often closely linked to police misconduct because of the manner in which a photo or live line-up was formulated. Eyewitnesses should be closely questioned about how they first participated in a line-up. Although most old cases do not meet the recommended protocol of sequential (not simultaneous) viewing of a line-up, there are other areas for inquiry. A few examples will suffice. Problematic line-ups include:

A witness picks one picture or person from a line-up only to be told “Try again,” until he chooses the suspect favored by police.

A witness is directed to one picture because of its position, color, or some other unique characteristic.

The line-up is grossly suggestive: only one person with facial hair is depicted even though the suspect was described as having a beard.

The witness sees a live show-up (ie. is shown one member of the line-up in person) just before seeing the line-up.

False Confessions

False confessions are also closely linked to the notion of police misconduct, although they may occur in lawful interrogation if certain factors exist. Factors to watch for include juvenile witnesses (44% of juvenile exonerees in the Michigan study gave false confessions compared to 13% of adults), and witnesses with a personality or mental disorder (69% of exonerees with mental disabilities gave false confessions compared to 11% of those with no disability).

Cooperating Witnesses / Snitches: Part of prosecutors’ arsenals since the beginning of English common law, the use of cooperating witnesses in criminal cases has been fraught with error. In fact, in the Michigan study, the use of cooperating witnesses ranks behind eyewitness error as the most common factor of exoneration cases. Cooperating witnesses (or snitches as they are sometimes called) provide testimony in exchange for leniency on other pending criminal cases. Snitches are often involved in murder cases, and studies show their use often dovetails with cases where police misconduct is found.

Witnesses who receive deals from prosecutors often start by giving general, vague outlines of a story. As time passes, and their own bargaining position fluctuates, they fill in holes and provide additional details.

Successfully exposing these gaps in testimony require several tactics. First, be sure sure to obtain all copies of witness interviews and compare the level of detail in each one. If later interviews are fleshed out much more than earlier ones, consider whether the witness doctored the story to match one told by another witness. In particular, review all handwritten notes taken during each interview (surprisingly, obtaining access to such notes on many old cases has been feasible). In one murder case, a witness named our client as the shooter. When the handwritten interview notes were examined, the witness was found to have named someone else as the shooter. That change was not included in the official typed version of the interview; it was only picked up by examining the original notes.

If a witness spent time in prison, obtain all incarceration records and compare times and places of incarceration (the defense lawyer must subpoena this information in Massachusetts). In a recent case, a falsified affidavit from a jailhouse snitch was exposed by proving that the snitch could not have overheard an alleged jail house confession because he was imprisoned elsewhere. during the time of the conversation.

Watch also for a snitch whose story has a physical displacement aspect: when the crime occurs, he is often not present, or just around the corner--a tactic used to insulate himself from being charged with the crime, but allowing him to testify to observing the crime or its immediate aftermath. A recent case saw a snitch who had repeatedly testified to being just out of the immediate vicinity of violent crimes--yet able to testify to seeing people running from the scene, or hearing excited utterances or confessions of people immediately afterwards (which overcame any hearsay objections). The defense team suspected that the snitch was often not just present during the shootings, but was actually the shooter. Because some prosecutors claim that they cannot make deals with shooters, some snitches, in hope of making a deal, will tell outrageous lies to hide their use of firearms.

Misconduct by Law Enforcement

One common factor in wrongful convictions is an unreasonable, almost irrational focus on one suspect by police. Such a focus may be due to personal animosities that develop, but it runs against solid investigative practice. A professional investigator does not identify the target and then seek evidence that only fits his theory. Instead, a professional initially seeks out all possible evidence and traces the evidence to the target. The approach is, at least early on, one of broad-ranging inquiries, not a blind shot in one direction. In one remarkable case in New England, a state police investigator admitted on the stand that his sole aim in investigating a criminal case was to either include or exclude the client (who was later exonerated). This laser focus on our client--perhaps done with the right intentions and a genuine belief in guilt--caused the prosecution to ignore many leads that may have led to the killer. Witnesses who failed to back up the official theory were isolated or ridiculed. Evidence that ran counter to the theory was ignored or discounted.

Besides obvious cases of physical abuse of witnesses, some psychological forms of witness interrogation that have occurred include:

Interviewing a witness while plying the witness with drugs and alcohol. In one case, the officer purchased the alcohol and let the witness drink himself into a stupor in the back of the cruiser.

Threatening to arrest or falsely arresting a witness when no reasonable basis exists for an arrest.

Threatening to take custody of the witness’s children, or hinting at some other form of government harassment (loss of a business or professional license, tax consequences, etc.

Directing threats against a witness’s family members unless the witness cooperates.

Interrogating a witness in a speeding cruiser while falsely claiming that the witness is being chased, or will be killed, by the suspect’s family.

New Evidence

One of the most difficult angles to examine involves witnesses or evidence that was never presented at trial. This is the classic form of newly discovered evidence that can be presented on a motion for a new trial. Examples include key witnesses who were never interviewed; physical evidence that was never tested such as blood evidence, DNA, fingerprints, or hair fiber. Recanting witnesses pose special problems as courts are split as to whether this constitutes new evidence. Such witnesses may have been interviewed many times over the years, making their credibility an issue for both the defense and the prosecution. Which statements are true? Why are there details in one interview that are missing from others? Even if the statements are reconciled, are there other factors that cause concern about bias or undue influence regarding the statement? If a recanting witness can show they were threatened, coerced or bribed, and evidence exists that can prove this coercion, a better opportunity exists for bringing this evidence into the case.

Public Bias: “He’s Still Guilty!”

Be prepared for a range of reactions from witnesses when you investigate old convictions--anything from disinterest to outright hostility. Criminal convictions are backed by the power of the written word, the gathered conclusion of jurors who parsed the evidence and arrived at final judgment. News stories have been written about the case, undoubtedly portraying the most heinous aspects of the case. Consider that careers have been made on your clients conviction: prosecutors have run for office on the image of prosecuting violent crime--possibly even your client; federal agents have advanced to leadership positions after proclamations of their gleaming law enforcement credentials. The suggestion that the person who was convicted of the crime is actually innocent is stunning--even offensive--to some people.

After one man was released from prison in New England after the real killer confessed, some witnesses remained irrationally convinced of his guilt. One man, pointing to a news article proclaiming the innocent man’s release from prison, yelled: “He’s still guilty. He’s still guilty. No matter what they say!”

These are just a few of the main points that arise in the investigation of wrongful convictions (many of them arise in other criminal cases). Wrongful conviction cases are as interesting, fulfilling, and frustrating as any case an investigator can take on. Without talented investigators, the conviction of innocent people--and the failure to convict the guilty--will continue to stain our justice system.


97% of the exonerations in the Michigan study stemmed from murder and rape cases. Some experts speculate that if the same intense scrutiny applied to major felony cases were applied to lesser crimes (say burglary and drug cases), a similar proportion of wrongful convictions would be discovered.

Exonerees served an average of over ten years in prison for crimes for which they were innocent.

Four defendants were exonerated posthumously after they had died in prison.

While extremely effective as evidence, DNA evidence exists only in a small percentage of criminal cases, the majority being rape cases.

The rate of exonerations increased sharply over the 15-year period of the study, from about 12 a year through the early 1990s to an average of 43 a year after 2000. From 1999 on, about half of all exonerations have been based on DNA evidence.

Other sites with information on wrongful convictions include:

Center on Wrongful Convictions

The Innocence Project -

Death Penalty Information Center -


John Nardizzi is president of Nardizzi & Associates, Inc., a Boston-based investigations firm specializing in the investigation of complex civil and criminal cases. He is handling five wrongful convictions cases at the present time.