Numerous factors lead to the wrongful conviction of innocent people for crimes they did not commit.  In cases where innocence investigations have resulted in exonerations, studies were performed to identify the cause(s) of the underlying conviction.
Factors Leading to Wrongful Conviction:

• 81% of time - Mistaken Identification by Eyewitnesses
• 51% of time - Serology Errors (ABO, protein blood typing)
• 50% of time - Police Misconduct
• 45% of time - Prosecutorial Misconduct
• 35% of time - Forensic Hair Comparison Errors
• 34% of time - Junk/Sloppy Science admitted at trial
• 32% of time - Bad Lawyering
• 22% of time - False Confessions
• 20% of time - Deliberately False witness testimony
• 19% of time - Deliberately False snitching by informants
• 7% of time - Other Forensic Science errors
• 1% of time - DNA testing errors

Scheck, B., Neufeld, P. & Dwyer, J. Actual Innocence: When Justice Goes Wrong and How to Make It Right. NY: Penguin Books. (2001).

Mistaken Identification by Eyewitnesses

Forty years ago, Justice Brennan wrote on behalf of the U.S. Supreme Court that “the vagaries of eyewitness identification are well known; the annals of criminal law are rife with instances of mistaken identification.” (United States v. Wade, 388 U.S. 218, 228 (1967)).  This observation holds true today.  The Innocence Project at the Benjamin N. Cardozo School of Law compiled statistics showing that incorrect eyewitness identification is the most prevalent cause of wrongful convictions.  Of the first 130 DNA exonerations of death row inmates in the country, 101 of the actually innocent had been convicted on wrongful eyewitness identification. 

While the accuracy of eyewitness identification is widely debated, the public at large considers eyewitness testimony to be the most reliable piece of information.  Juries often sympathize with victims of violent crimes and unquestioningly assume their eyewitness identifications are correct.  Law enforcement officials often get “tunnel vision,” focusing all of their efforts into making a conviction stick to a suspect whom the witness has identified as the perpetrator. Eyewitness testimony is regularly the most convincing part of a trial. It adds direct evidence to a prosecution that would otherwise be based on circumstantial evidence.  However, there is overwhelming evidence that eyewitness identification is often wrong. 

Research on eyewitness identification has shown there are two types of memory:  event memory and identification memory.  Event memory quantifies how well the eyewitness can describe the actual event.  Identification memory quantifies the eyewitness’ ability to recognize a person.  These types of memories are strongly affected by the vast questions and procedures that eyewitnesses must endure in order to remember and identify a suspect. 

In identifying a suspect, a witness faces numerous variables which may cause eyewitness misidentification.  A witness may make a “good faith” error by trying to cooperate with a suggestive or biased investigation/lineup.  A witness may face his/her own prejudice or perceived community prejudice.  This is especially seen in cases where cross-racial identifications apply.  Also, a witness who has witnessed or suffered a violent crime may be suffering from psychological trauma or shock.

Seventy-five percent of cases overturned based on DNA evidence were originally convicted based on eyewitness testimony.  Eyewitness testimony is extremely persuasive when presented to a judge or a jury.  However, social science research over the last thirty years has shown that eyewitness identification is a highly unreliable source of evidence.  There are two main reasons for the unreliability of eyewitness identification:  estimator variables and system variables:

Estimator Variables are the factors over which the justice system exerts little to no control.  They include items such as:

• Age of witness – children and the elderly tend to be less accurate than adults
• Facial distinctiveness – faces rated highly attractive or highly unattractive are recognized better than other faces
• Suspect disguise or changed appearance – this leads to less accuracy in identification
• Target salience – it is harder to identify a perpetrator if many people were present at the crime scene
• Exposure duration – it is harder to identify a perpetrator if the viewing time is short
• Weapon presence – when weapons are used in a crime, witnesses tend to focus attention on the weapon rather than the person wielding it
• Seriousness of crime – people tend to be less accurate in identification tests if the crime is thought to be less serious
• Stress and arousal – moderate arousal tends to lead to better acquisition of information than high or low levels of arousal
• Cross-gender bias – people are more accurate when they identify someone of their own gender
• Cross-racial bias – people are more accurate when they identify someone of their own race
• Time delay – memory declines over time
• Changes in experiential context – it is sometimes difficult to recognize someone if he or she is seen in a different place
• Post-event information – information obtained after the crime occurs can distort a person's memory.  Post-event information can also be a system variable if the information comes during an interview with law enforcement and/or justice officials.
• Crime scene variables – the lighting and/or physical layout of the crime scene could affect how well the witness views the crime
• Distance – from which the witness observes the crime

System Variables are factors over which law enforcement and the criminal justice system can exert control.  They include items such as:

• Lineup instructions – if a witness expects the perpetrator to be present in the lineup, he or she may feel obligated to pick someone, even if the perpetrator is not there. The witness should be told that it is a legitimate answer to say that the culprit is not present.
• Foil bias – the suspect should not stand out from the innocent distracters (also known as “foils” or “fillers”) in a lineup. Everyone in a lineup should wear similar clothing and have similar general appearances.
• Investigator bias – the investigator may unintentionally or intentionally let the witness know which person in the lineup is the suspect
• Presentation bias – a sequential presentation of people in a lineup is better than presenting all people simultaneously
• Post-event information – information obtained after the crime occurs can distort a person's memory
• Communications or instructions given after identification has been made – if a witness is led to understand that he/she should have identified a particular suspect, he/she may be inclined to choose that person

As previously noted, cross-racial bias is an estimator variable that may occur when people identify members of other races.  In cases where eyewitness misidentification involving cross-racial bias was the cause for a wrongful conviction, misidentification occurred as follows:
Race and Mistaken Identification

• 44% of time - Caucasian misidentifying African American
• 25% of time - Caucasian misidentifying Caucasian
• 21% of time - African American misidentifying African American
• 4% of time - Latino misidentifying Latino
• 3% of time - Caucasian misidentifying Latino
• 1% of time - Latino misidentifying African American

Scheck, B., Neufeld, P. & Dwyer, J. Actual Innocence: When Justice Goes Wrong and How to Make It Right. NY: Penguin Books. (2001).

In an effort to reduce the number of wrongful convictions, a small handful of states have implemented guidelines for police protocol in interviewing eyewitnesses.  In 2006, the California Commission on the Fair Administration of Justice released a report stating that “[m]istaken eyewitness identification…is the leading cause of the wrongful convictions that have been identified in recent research.”  Accordingly, the California State Senate enacted S.B. 171 and S.B. 1544 to provide regulations relating to eyewitnesses, mandating “double blind” lineups and sequential lineups, rather than simultaneous photographic lineups.  The National Institute of Justice (NIJ) also commented on the need to improve the accuracy of eyewitness identifications and released guidelines in 1999 addressing this problem. 

Of particular interest, the NIJ recommended guidelines address the following issues:

An implementation of sequential lineups, rather than simultaneous lineups.  Typically, law enforcement officials use simultaneous lineups.  Specifically, a witness is shown a number of mug-shots or “suspects” at one time.  Potentially, the witness will choose the person that looks most like the perpetrator, instead of trying to identify the actual person.  In a sequential lineup, a witness would be shown one photograph or person at a time.  This would allow a witness to look at each person to make an independent determination of whether the person was the actual perpetrator.

Having previously uninvolved law enforcement officials perform lineups and handle eyewitness identifications.  For the most part, the law enforcement officials who have worked on a case perform the lineup once a suspect is identified.  This gives law enforcement officials the opportunity to knowingly or unknowingly give the witness hints as to which person(s) have been identified as potential suspects.  Having previously uninvolved law officials perform the lineups would avoid the possibility of officer bias affecting the witness.

Very few states have attempted to follow those guidelines, which might lead to a decrease in wrongful convictions.

Recently, two Texas cases gained national attention in regards to eyewitness identification.  Ruben Cantu was featured by the Houston Chronicle while Carlos DeLuna was the subject of a Chicago Tribune investigative report. Ruben Cantu and Carlos DeLuna were both convicted, sentenced to death, and executed.  There was no physical evidence tying either man to the crime.  The only thing implicating Cantu and DeLuna was eyewitness testimony.  In both cases, the eyewitnesses have now stated that they identified the wrong man. 

The witnesses stated that they identified Cantu and DeLuna because of police pressure, fear of authorities, and the general way in which the lineups were conducted.  In Cantu’s case, the eyewitness admitted he fears authority and felt immense police pressure to implicate Cantu.  In DeLuna’s case, the eyewitness was shown neither a sequential lineup nor a simultaneous lineup.  Rather, the eyewitness was shown only Carlos DeLuna.  The witness was also told by the police that they had apprehended DeLuna in the general area of the crime with cash in his pocket.  In fact, during pre-trial activities, the eyewitness could not even identify DeLuna in the courtroom. 

The problems surrounding eyewitness identifications are so severe that in 2003, a commission appointed by Governor George Ryan of Illinois released 85 recommendations regarding criminal justice reform.  These recommendations included a proposal to forbid capital punishment in cases where the conviction is based solely on the testimony of an eyewitness. 

Limits of Serology
Serology is the scientific study of blood serum.  Serological tests can also be performed on other bodily fluids, such as semen and saliva.  Testing serological material in practice refers to the diagnostic identification of antibodies found in the particular serum. 
Before DNA testing, bodily fluids collected from crime scenes were routinely tested for blood type (serology). While this form of forensic science has often been reliable, it has substantial limits because it lacks the precise power of DNA testing.  Serology can only determine if a suspect is a member of the population of potential donors who could have deposited the biological evidence at the crime scene.  In many trials, serologists have failed to disclose the limited applicability of the test or have mistakenly applied statistical evidence, leading to juries being misled and innocent people being convicted. 

While serological evidence can now be tested for DNA to determine a person’s guilt or innocence, many states have not passed legislation guaranteeing the right to DNA testing.  Even in those states where this right is mandated by statute, there are stringent limits on its application and inmates have been refused testing even when the results might have determined guilt or innocence.  An inmate seeking to test biological evidence for DNA also faces procedural issues.  Some states, such as Alabama, do not supply attorneys for the entire appeals process.  In many states, attorneys lack the adequate resources for a re-investigation of the biological evidence.  Finally, inmates in some states must wait several years before an attorney is assigned to begin their appeals, increasing the possibilities that evidence can be destroyed, lost, mishandled, or manipulated. 

Mishandling of Scientific Evidence
When newly tested DNA evidence is presented by an inmate after conviction, it is typically checked several times before the inmate can be set free.  However, this careful examination of the DNA evidence is often not applied to pre-trial DNA testing that can lead to a wrongful conviction.  Pre-trial DNA testing is subject to a myriad of possible contamination points:  it can become contaminated at the scene of the crime; it can be collected improperly; it can be transported and/or stored improperly; it can be tested improperly; and it can be presented to the judge or jury improperly. 
Recently, the Harris County crime lab has undergone severe scrutiny and intense investigations for their unreliable results.  In fact, the DNA testing that had been done at this facility has been shown to be so unreliable that all of its results are being stricken from the national database of DNA profiles.  The Houston Chronicle’s investigation into the crime lab found the following problems:

• James Bolding, the founder and former head of the DNA lab, did not meet the standards for the job. Among other things, he originally failed both algebra and geometry in college and he never took statistics. Bolding held a bachelor's and a master's degree from Texas Southern University, but was academically dismissed from the University of Texas Ph.D. program. Bolding resigned from the lab after Houston's police chief recommended he be fired.

• Jobs were often given to graduates without the required degrees, such as those who had majored in chemistry or zoology. Among those hired to do DNA tests or prepare samples for testing were two workers from the city zoo. One had most recently been cleaning elephant cages. The other had done DNA research, but only on insects.

• The lab hired Joseph Chu despite a former employer's comment that he "has difficulty in speaking English," (a serious handicap when testifying in court). In his application, he wrote, "I have skilled several equipments" and "I have experience in testing animal and sacrificing them." His supervisors rated him poorly in communication. Chu was suspended for fourteen days after several errors were found in four cases, including a capital murder case. He also misrepresented his degree in a court document.

• The roof at the lab had been leaking for years, contaminating critical evidence.

L. Olsen, “DNA Lab Analysts Unqualified,” Houston Chronicle, September 8, 2003.

Sometimes, DNA evidence is presented in a deliberately false and manipulative manner.  For example, the head serologist at the West Virginia State Police crime laboratory, Fred Zain, falsified test results in as many as 134 cases from 1979 – 1989 to confirm the prosecution’s theory of the crime.   (Paul Giannelli, The Abuse of Scientific Evidence in Criminal Cases: The Need for Independent Crime Laboratories, 4 Va. J. Soc. Pol’y & L. 439 (1997).)  Zain’s falsifications were not discovered because of the prevalent lack of resources for defense expert witnesses.  This lack of resources is a chronic problem in defense representation.  Without such resources, neither rigorous cross-examination nor counterproof is present to keep prosecution experts honest.

Government Misconduct

In some cases of wrongful convictions, officials have taken steps to ensure that a certain defendant is convicted even in the face of weak or even clear proof of that defendant’s innocence.  While most law enforcement officers and prosecutors are honest and trustworthy, criminal justice is still a human endeavor with potential for corruption.  Misconduct can include:  withholding crucial evidence from the defense; deliberate mistreatment or destruction of evidence; deliberate suggestions given during identification procedures; coercion of false confessions; and use of unreliable informants or snitches.

Unreliable Scientific Methods

With the advent of DNA technology in the late 1980s and 1990s, the criminal justice system now had a more reliable method to determine whether someone was actually guilty or innocent of a crime.  DNA technology has called into question other inexact scientific methods of determining guilt or innocence, such as: blood typing, hair comparison, bite marks, and ballistics.  All of these inexact types of evidence can now be tested for DNA to conclusively reveal the truth.  However, these inexact methods are often still used to convict defendants in American courtrooms. 

Even today, microscopic comparison of hairs, particles, and other fibers is used in courtrooms to convict defendants.  In some instances, it is the only evidence used to convict a defendant.  However, comparison tests cannot stand up to scientific scrutiny.  DNA results have shown repeatedly that these tests, as performed, lack sufficient reliability and frequently produce erroneous results. The advent of mitochondrial DNA testing has given scientists much stronger evidence, even when comparing only a fragment of a hair.

Defense Services (Bad Lawyering)

Justice Ruth Bader Ginsburg pointed to inadequate representation as part of the problem relating to wrongful convictions, and even endorsed the idea of a moratorium on executions. "People who are well represented at trial do not get the death penalty," said Ginsburg.  "I have yet to see a death case among the dozens coming to the Supreme Court on eve-of-execution stay applications in which the defendant was well represented at trial." (“Ginsburg Backs Ending Death Penalty,” Associated Press, April 9, 2001)

The Model Rule of Professional Conduct 1.1 ethically mandates lawyers to provide competent representation, defined as “the legal knowledge, skill, thoroughness, and preparation reasonably necessary.”  Rule 1.3 gives lawyers the duty of diligence, requiring that they “take whatever lawful and ethical means are required to vindicate a client’s cause.”  Many factors work to undermine these rules, most of them stemming from limited resources, under-funding of defense services, the strict standards for reversing a case, and the stringent standards surrounding malpractice liability. 

The standard for reversal on the basis of ineffective assistance of counsel, set forth in Strickland v. Washington, 466 U.S. 668 (1984), requires a “reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.”  This restrictive standard allows most bad lawyering to remain unremedied. Similarly, in order for a convicted defendant to pursue a malpractice claim against his or her attorney, he or she must prove their innocence before the claim can be considered.  Further, some states insulate their public defenders from civil lawsuits by clients by granting them governmental immunity. 

Inadequate funding plays a significant role in contributing to bad lawyering.  The lack of funds and other resources makes it difficult for lawyers to give adequate time to investigate and prepare their cases.  It also prevents lawyers from retaining investigators and forensic experts.  This lack of funding also serves to undermine the functioning of the adversary system that the criminal justice system is based upon.  The adversary system assumes that each side has the resources with which to investigate and present his own legal and factual positions, and also to discover the weaknesses in his opponent’s legal and factual position.  However, if one side lacks resources to properly investigate, prepare, and present his case and discover the weaknesses in his opponent’s, the adversary system ceases to exist. 

One of the most important factors in determining whether or not a defendant will receive the death penalty is the quality of legal representation he or she receives.  Most capital defendants cannot afford their own attorneys.  They will receive court-appointed counsel, many of whom are overworked, underpaid, or lack the trial experience required for the complexities of a death penalty trial.  The criminal justice system lacks sufficient resources to ensure poor defendants a fair trial.  When they are represented by an ineffective, overworked, or incompetent defense lawyer, their chances of a thorough investigation into their case and a competent defense presented at trial become increasingly slim.  Ineffective, overworked, or incompetent lawyers have failed to investigate cases, call witnesses, or properly prepare for trial, all leading to the conviction of innocent people. 

Some of the worse cases have included lawyers who fell asleep in the courtroom during the trial, failed to investigate alibis, failed to consult forensic experts, failed to appear at hearings, abused alcohol or illicit drugs during the trial, or were disbarred shortly after a capital trial. 

“Expert” Evidence & Testimony
Properly qualified and admissible expert testimony can be powerful evidence. Experts can identify a potential suspect, excluding all others.  They may also assist the fact finder to appreciate specific facts and circumstances that are outside of own general knowledge and understanding.  However, tainted, manipulated, and unsubstantiated expert testimony/evidence, presented in a scientific and authoritative manner or based on unreliable science and fact, can easily persuade a fact finder to wrongfully convict an innocent person. 
Forensic evidence is offered by “experts” during a trial, often resulting in jurors giving it more weight than other evidence.  Because forensic science can mean the difference between life and death for some defendants, fraud and other types of misconduct in the field are of particular concern.  In some instances, crime labs or their personnel are not impartial because they are too closely tied to the police and/or the prosecutors.  In other cases, personnel lacking the requisite background knowledge or training will embellish their findings, thinking that he will not be caught since the lawyers, judge, and jury members lack training in the relevant forensic science.  In still other cases, critical evidence is destroyed or contaminated so that re-testing is impossible; these wrongful convictions can never be overturned because the evidence can not be re-tested. 

The risk of misconduct or mishandling of forensic evidence starts at the crime scene, where evidence can be mishandled, destroyed, or even planted.  This evidence is then sent to a forensic lab where it can be contaminated, unnecessarily consumed, mislabeled, or poorly tested.  Next a report is generated where technicians or their supervisors can misrepresent the results.  The identification, collection, testing, storage, handling and reporting of any piece of forensic evidence involves a number of people and a number of stages where evidence can be deliberately or accidentally mishandled.

False Confessions

Innocent defendants have been convicted because they made incriminating statements, gave confessions, or pled guilty to a crime.  These confessions are not always based on actual guilt, but are sometimes motivated by external influences. 

Some innocent people have confessed to crimes they did not commit due to their mental states.  Confessions given by children or juveniles are often unreliable because they can be easy to manipulate and may not fully understand their situations.  They may believe they can go home as soon as they admit their guilt and their innocence can be proved later.  Confessions given by mentally disabled or mentally ill people can be unreliable because they may feel compelled to conform and accommodate authority figures.  Many law enforcement officials who are conducting interrogations lack any special training to question suspects with mental disabilities or mental illnesses. 

Still other people have confessed because of the sheer burdensome length of their interrogation, exhaustion, or because they feel that confessing is the most beneficial way to handle the situation at that time.  Some law enforcement officers will use certain tactics to elicit a confession from an innocent person that can include telling a person they will be convicted with or without a confession and that their confession could result in a more lenient sentence, telling a person that confessing is the only way to avoid the death penalty, or implying that physical harm could come to them if they do not confess. 

Why would an innocent person confess?  An innocent person may confess for a variety of reasons including: diminished capacity, ignorance of the law, fear of violence, duress, coercion, mental impairment, actual infliction of harm, intoxication, threat of a harsh sentence, or misunderstanding the situation.