Frequently Asked Questions of Gary L. Wells
Definitely not. Only a very small fraction of crimes can be solved with DNA testing. The vast majority of perpetrators in assaults and murders and virtually all robberies, drive-by shootings and other major crimes do not leave behind definitive biological trace evidence that can show that the eyewitness was mistaken. The need for reliable eyewitness evidence is not significantly diminished by the advent of forensic DNA tests. The Gary Graham case in Texas, in which he was executed based largely on a single (and very questionable) eyewitness identification, is more representative of the actual state of affairs. Like most all eyewitness cases, there was no biological evidence in the Graham case to be analyzed. Every year, over 77,000 Americans become criminal defendants after being identified by an eyewitness. Perhaps a few thousand of these can now be resolved one way or another using DNA tests. The remaining tens of thousands rely on the diagnosticity of the eyewitness identification procedure to produce the appropriate result.
The DNA exoneration cases are almost exclusively cases of sexual assault because those (sexual assault cases) are the cases in which definitive biological evidence tied directly to the perpetrator is left at the scene of the crime. As my former student and current friend and colleague John Turtle said: "Unless the guy robbing that 7-11 store gets pretty damned excited, he's not going to be leaving behind any biological fluids."
The New York Supreme Court recently ruled that testimony by eyewitness experts can be permitted at trial. Isn't expert testimony the answer to this problem?
No. Although many have suggested that the appropriate answer to the eyewitness problem is to permit testimony by eyewitness experts in cases that rely on eyewitness identification evidence, there are several reasons why this is not a significant answer to the eyewitness reliability problem. First, there are too few experts to actually service this problem and expert testimony is expensive. Also, although an expert can cite various factors that make eyewitnesses more or less reliable, no one, even an eyewitness expert, can tell whether this particular eyewitness made a mistake. Once an eyewitness identification error has occurred, it has most of the same appearances that an accurate identification has. Borrowing a quote from the brilliant Boston defense lawyer James Doyle "The scientific eyewitness literature can tell you that something happens 80% of the time, but it cannot tell you whether a given case is one of the 80 or one of the 20."
It is true that eyewitness identification evidence is less reliable than jurors believe it to be. However, instead of bringing jurors' beliefs about the reliability of eyewitness identification evidence down to the current level of eyewitness reliability, we need to bring the level of eyewitness reliability up to the level that jurors expect it to be. The best policy for the legal system to adopt is one of prevention of eyewitness identification error, not catching eyewitness identification errors after they have occurred.
Your research shows that the methods that law enforcement uses for conducting lineups is contributing to mistaken identifications and false confidence by eyewitnesses. Since the scientific evidence is now clear that sequential lineups are better and that blind testing should be used, why isn't the legal system adopting these new procedures?
That is a simple question, but the answer is compound. First, keep in mind that some jurisdictions are now (finally) changing. The most prominent example is the state of New Jersey. New Jersey has now adopted blind testing and sequential lineups for all police in the state (see New York Times article). However, New Jersey is clearly the exception for two reasons. First, unlike other states, the Attorney General of New Jersey has authority over all police in the state of New Jersey. Second, New Jersey AG John Farmer has an excellent staff (I particularly commend Lori Linskey and Debra Stone) who have pursued this matter vigorously and AG Farmer had the good judgment to exercise his authority to issue instructions to NJ police requiring them to use sequential lineups, blind testing, and other safeguards that we have been advocating. This method of adopting better lineup procedures cannot happen in other states. The Attorneys General of Kansas, or Pennsylvania, or New York, for instance, have no authority to order police in their states to adopt particular procedures for gathering evidence. In some places, such as Philadelphia and New York, there have been concerted attempts to have judges order the police to conduct lineups in ways that we have shown to be more reliable, but these efforts have been stalled by opinions that say that judges have no authority to order these things.
In some jurisdictions, prosecutors are openly opposing any declarations that would suggest that sequential lineups and blind testing are needed. They seem to be driven by a concern that such declarations might raise doubt about people they have prosecuted or open up lines of appeal for people who have been convicted using the riskier procedures. It seems similar to the way some prosecutors have been opposing people's calls for post-conviction DNA testing. I believe that the opposition by subsets of prosecutors will ultimately fade, especially when they begin to appreciate the way this can make the evidence they use at trial more credible. But this opposition accounts for some of the slow progress.
The fact is that we just do not know at this point how to effect change in the methods by which lineups are conducted. Some change is happening at the level of a few individual police departments. In Clinton, IA, for instance, the police chief ordered his officers to follow the NIJ Guide for conducting lineups. There are, however, over 13,000 police departments in the U.S. that operate largely autonomously. How to affect large numbers of these departments in a short period of time (like is happening in New Jersey) is an open question. Influential people (such as judges, legislators, police chiefs, county prosecutors) are going to have to step up to the plate.
How often do eyewitnesses identify the wrong person from lineups in actual cases?
No one really knows the answer to this question. However, recent field studies show that eyewitnesses select a known-innocent "filler" from a lineup 20-25% of the time in actual criminal investigations. [A filler is a member of the lineup who is not a suspect. Fillers are used in lineups merely to "fill out" the lineup so that the eyewitness is not shown just one person.] It turns out that the 20-25% rate of filler identification is probably a conservative estimate because police often record filler identifications as "eyewitness could not make a positive identification of suspect," which tends to get counted as "no identification."
Click Here to go to Wells' Eyewitness
Identification Home Page