Guilty until Proven Innocent – The Problems with Eyewitness Misidentification

Serving nearby areas by Palm Beach and West Palm Beach, Florida

Across the country, more than 230 people have been convicted of crimes that they did not commit.  After some or all of their sentences, often more than a decade in prison, they were exonerated through DNA or other forensic testing.  Of those wrongful convictions, 75% were based in large part on misidentification by an eyewitness.  The actual number of wrongful convictions is actually much larger, and very easily runs into the thousands, as the exonerations have come only in recent years with the availability of newer technology and forensic capabilities.  In this article, Romano Law Group attorney Eric Romano highlights the dangers of eyewitness identification.

You Can’t Hide Your Lyin’ Eyes (or can you?)

Understanding the Dangers of Eyewitness Identification

Eric Romano

Romano Law Group

West Palm Beach, Florida

            “Eyewitnesses frequently play a vital role in uncovering the truth about a crime.  The evidence they provide can be critical in identi­fying, charging, and ultimately convicting suspected criminals. That is why it is absolutely essential that eyewitness evidence be accurate and reliable. One way of ensuring we, as investigators, obtain the most accurate and reliable evidence from eyewitnesses is to follow sound protocols in our investigations.

            “Recent cases in which DNA evidence has been used to exonerate indi­viduals convicted primarily on the basis of eyewitness testimony have shown us that eyewitness evidence is not infallible. Even the most honest and objective people can make mistakes in recalling and interpreting a witnessed event; it is the nature of human memory. This issue has been at the heart of a growing body of research in the field of eyewitness identifi­cation over the past decade. The National Institute of Justice convened a technical working group of law enforcement and legal practitioners, together with these researchers, to explore the development of improved procedures for the collection and preservation of eyewitness evidence within the criminal justice system.”


            -A Message from the Attorney General, U.S. Dept. of Justice – “Eyewitness Evidence, A    Guide for Law Enforcement”, Oct. 1999 (


            A 2009 report issued by the Innocence Project (, “Reevaluating Lineups: Why Witnesses Make Mistakes and How to Reduce the Chance of a Misidentification” (, provided some startling statistics:

            -Over 230 people convicted in U.S. courts have been exonerated through DNA testing

            -Of those 230 wrongful convictions, 75% involved eyewitness misidentification

            -In 38% of the misidentification cases, multiple eyewitnesses misidentified the same innocent person

            -Over 250 witnesses misidentified innocent suspects

            -53% of the misidentification cases involved cross-racial identifications

            -Of 175 misidentification cases, 25% involved 2 witnesses misidentifying the same innocent person

            -Eyewitness identification is by far the leading cause of wrongful convictions in the U.S.

            Every criminal defense attorney’s worst fear is the innocent client who gets convicted. Tragically, that fear is realized more often than most people think, largely due to faulty eyewitness identifications.  In their book, Actual Innocence, Barry Scheck, Jim Dwyer, and Peter Neufeld estimated that of the first 100 wrongful convictions determined by DNA testing, more than 80% relied heavily on eyewitness identification.  Because such identifications are inherently subjective, imprecise and subject to misperception, unquestioned reliance on them often leads to unjust results.  The circumstances surrounding the crime and the witness’s perception of the perpetrator are not always enough to call into question the reliability of the identification.  Often, an erroneous identification will withstand rigorous cross-examination and the scrutiny that comes with litigation, largely because the witness, although wrong, genuinely and firmly believes she is right.


Anatomy of an Identification

            An eyewitness identification generally encompasses three distinct, yet related, steps.  First, the witness must observe or perceive the criminal.  Second, the witness must commit that perception to memory.  Third, the witness must recall the memory in order to make the identification.

            The first stage, observation, is the most critical, and there are many factors that determine how well the witness perceives the criminal.  You must carefully examine the conditions under which the observation was made.  Jurors have a natural tendency to want to believe and trust an eyewitness to a crime, as such witnesses are often perceived as neutral, objective, sincere, and honest.  They are also usually quite sympathetic, as they will often be the victim of a traumatic or terrifying ordeal.  The goal of the criminal defense attorney is to overcome this natural tendency and to persuade jurors to view eyewitness testimony with critical analytical judgment.  The circumstances surrounding the event affect the witness’s perception.  Such factors are commonly referred to as “event-related variables”, and they include the following:

            -Time of day

            -Lighting conditions

            -Race/ethnicity of witness and defendant

            -Distance between defendant and witness

            -Length of observation time

            -Stress of situation

            -Presence (or lack thereof) of weapon

            -Any relationship between witness and defendant

            -What defendant was wearing

            -Any visual obstructions

            -Witness’s vision acuity (wear glasses/contacts, etc.)

            -Nature of crime/event

            -Whether witness or defendant were moving or stationary


            In the second stage, memory storage, the witness must commit the information to memory.  This often involves creating a mental picture of the witnessed event, and storing it in memory.  The witness’s ability to do this depends on many factors, including the witness’s age, gender, cognitive abilities, mental status, and general health.  For example, women tend to pay closer attention to detail, while men tend to see the bigger picture and overlook the details.  Women tend to be more empathetic, while men tend to be more analytical.  Men generally have stronger spatial skills and are better at judging distances, while women tend to have better visual memory.  Younger people tend to have better short and long-term memories.  During this stage, the information is also being filtered through the witness’s mental frame, a process often referred to as “framing.”  Each person has a unique frame of reference, which is influenced by the factors listed above as well as education, marital status, religion, cultural background, nationality, race, and more.  This “framing” is usually a subconscious process, meaning the witness does not give any cognitive thought to it.  This often results in inaccurate information being stored through no fault of the witness.

            In the third stage, memory recall, many of the same factors are just as important.  However, and added factor at this stage is the amount of time that has elapsed since the event.  As a general rule, the more time that has gone by, the more unreliable the memory will be, for two reasons.  First, memories tend to fade and lose detail with time.  Second, the witness’s experiences between the event and the identification will often influence the memory.  For example, a witness will often hear other information about the event from law enforcement, the media or from other witnesses.  Such information can influence the witness’s memory and can often cause the witness to integrate such information into her own memory of the event, forever altering the witness’s own memory.  This leads to a “false memory”, which is usually innocent and unintentional (in fact, the witness often doesn’t even know that her memory has been slightly altered or influenced by second-hand information), but it is unreliable regardless.

Once the witness’s memory is altered, the original memory cannot be restored. (James Doyle, True Witness: Cops, Courts, Science and the Battle Against Misidentification, 91).


Identifying the Problems


Forced Choice


            Lineups, both in person and photographic, are inherently flawed because they often present the witness with a forced choice.  Despite instructions to the contrary, witnesses feel a psychological pressure to choose somebody.  Even when the guilty person is absent from the lineup, the witness will often choose the person who looks most like the real culprit.  This tendency is referred to in the psychological literature as “relative judgment”. See From the Lab to the Police Station: A Successful Application of Eyewitness Research; Wells, et al., American Psychologist, June 2000.  Apparently in recognition of this tendency, the U.S. Department of Justice guide, Eyewitness Evidence: A Guide for Law Enforcement (1999), explicitly states “Scientific research indicates that identification procedures such as lineups and photo arrays produce more reliable evidence when the individual lineup members or photographs are shown to the witness sequentially – one at a time – rather than simultaneously.” (p. 9).  Even when the witness is told that the suspect may or may not be present and that the witness does not need to select anyone, the witness still often feels tremendous pressure to pick someone (usually the one who looks most like the suspect) in order to bring closure, to solve the case, to help the police, and to avoid a feeling of failure or helplessness.


Investigator Influences


            Another common flaw in witness identification procedures is that the investigating officer usually knows the identity of the suspect.  Although often unintentional, this knowledge alone can cause the officer to subtly influence the witness’s choice, by subtle changes in body language, demeanor and tone of voice.  The manner in which the lineup or photo array is conducted can sometimes be suggestive or easily manipulated.  Such factors are known as “procedure-based variables”.  Such variables can have a harmful effect – instead of helping to reveal the witness’s true memory, they replace it with a new one.  Event-based variables influence the witness’s initial perception, and procedure-based variables influence the manner in which the witness’s perception is converted into evidence.  While the witness and the court system generally have no control over event-based variables, problems resulting from procedure-based variables can be identified and prevented, or exposed when they occur.

            Witnesses have a natural tendency to want to help the investigating officer find the guilty person.  This often leads a witness to choose somebody from the lineup in order to be “helpful”, rather than to disappoint the officer by not choosing anyone.  At this point in the investigation, the officer has usually established a rapport and trust with the witness.  The witness may view the lineup as a sort of test, and may feel pressure to “pass” by making the right choice.  Knowing that the officer probably already knows who the “right choice” is, the witness will look to the officer for guidance.  Many times, sensing frustration or hesitation by the witness, the officer will try to help the witness make “the right choice”.   Such guidance is often innocent and done without the investigator even realizing it (but not always).




            Another important factor in witness identification is the confidence with which the witness makes the identification.  In general, confident people are perceived as more credible.  A witness who testifies that he is 100% sure of his identification will generally receive a greater level of trust than a witness who says he’s 50% sure.  The witness’s confidence level is also affected in large part by outside influences.  One of the most significant factors in a witness’s confidence level is feedback regarding the identification.  If a witness makes an identification, but is hesitant and unsure, and is then told that another witness also identified the same person, the witness’s confidence in the accuracy of his own identification will be significantly increased.  Even if both witnesses identified an innocent person, both will feel confident in their choice upon learning that another witness identified the same person, thereby increasing the weight given to their testimony and increasing the chance of a wrongful arrest and/or conviction. The same result occurs when the witness receives any positive feedback about her identification (for example, when the investigator says “great, that’s who we thought it was” or “he has a lengthy record of doing the same thing to others.”).  Similarly, the more opportunities a witness has to identify the same person (for example, in a physical lineup, then in a photo array, then in a subsequent photo lineup), the more it reinforces to the witness the correctness of his choice, and the more it strengthens his confidence in his decision.  The more times the witness sees the suspect’s face in a lineup, the more recognizable the face becomes, and the more likely a wrongful identification will take place.  This is because the witness will recognize the face from the previous lineup, which may replace or alter the memory of the face at the crime scene.




            A showup, in which the police show a particular suspect to the witness at or near the crime scene, is inherently suggestive.  Lineups, both in person and by photo, can also be suggestive, depending on how they are administered.  Even a well-intentioned investigator can subtly influence the process without even realizing it.  For example, during an in-person lineup, the witness may notice that the investigator is staring intently at person #3.  In a photo lineup, the investigator’s finger may be pointing to photo #2 as he places the sheet in front of the witness.


Suggestions for Preventing Misidentification


            Based on a comprehensive review of various professional studies, wrongful conviction cases, and psychological research, the Innocence Project has recommended the following guidelines to help prevent misidentification.  These suggestions are also consistent with those recommended by the DOJ guide referenced above, as well as by other law enforcement agencies and criminal defense organizations:

            1.         A lineup should include only one suspect.

            2.         In addition to the suspect, a photo array should include at least five fillers, and a physical lineup should include at least four fillers.

            3.         Fillers should generally resemble the description of the perpetrator provided by the witness, including any unique features (tattoo, dreadlocks, scar, weight, build, height, etc.), yet be discernible enough for the witness to tell them apart.

            4.         If a witness has previously viewed a lineup or photo array to identify another person suspected in the case, the fillers used should be different from those used in the previous identification procedure.

            5.         In a photo array, the suspect’s photo should be contemporary, should resemble his appearance at the time of the crime, and should not significantly stand out from the fillers.

            6.         If there is more than one witness, the suspect’s position in each lineup/array should be changed with each witness, and the witnesses should be segregated and instructed not to discuss the case or procedure with each other or with the media.

            7.         Prior to and during the identification procedure, the witness should not be provided any information about the suspect’s background or connecting the suspect with the crime.

            8.         The person who conducts the lineup should not know who is a suspect and who is a filler.  This is known as “blind administration”.

            9.         Describe the photo array only as a collection of photographs.

            10.       Instruct the witness that the perpetrator may or may not be present.

            11.       Assure the witness that the investigation will continue regardless of whether or not the witness makes an identification.

            12.       Instruct the witness that the investigator must ask the witness to state in her own words how certain she is of any identification she makes.

            13.       Advise the witness that the administrator does not know who the suspect is.

            14.       The administrator should not provide any feedback to the witness, and should immediately document any statement the witness makes regarding her level of confidence in her selection.

            15.       Video record the identification procedure.

            16.       Use sequential procedures, rather than showing all lineup members simultaneously.


A brief legal framework

To determine if an out-of-court identification of a defendant is admissible at trial, the courts employ a two-part test: (1) Did law enforcement use an unnecessarily suggestive procedure, and if so, (2) considering all the circumstances, did the procedure create a substantial likelihood of irreparable misidentification?  Manson v. Brathwaite, 432 U.S. 98 (1977); Grant v. State, 390 So.2d 341 (Fla. 1980); Fitzpatrick v. State, 900 So.2d 495 (Fla. 2005).  To determine whether an identification is reliable, which involves assessing the likelihood of misidentification, courts should consider the following factors:  (1) the opportunity for the witness to view the defendant at the time of the crime, (2) the witness’s degree of attention, (3) the witness’s prior description of the defendant, (4) the level of certainty demonstrated by the witness at the time of confrontation, (5) the length of time between the crime and identification. Neil v. Biggers, 409 U.S. 188 (1972).  Identification may be attacked as a denial of due process if the circumstances of the identification were unnecessarily suggestive and conducive to irreparable mistaken identification. Stovall v. Denno, 388 U.S. 293 (1967).  When the photo of the defendant appearing in a photo array was cropped to remove all indicia that he was wearing jail attire, the photo is less likely to be found inadmissible as unfairly suggestive. See Soanes v. State, 31 So.3d 914 (4th DCA 2010).

“[A]n identification made shortly after the crime is inherently more reliable than a later identification in court.”  State v. Freber, 366 So.2d 426 (Fla. 1978).  Testimony of a prior out-of-court identification is admissible as substantive evidence of identity if the identifying witness testifies at trial that a prior identification was made.  Id.  Pursuant to Fed.R.Evid. 801(d)(1)(C), evidence of a prior out-of-court identification is not hearsay and is admissible if the declarant testifies at trial and is subject to cross-examination.  Similarly, a police officer may testify at trial that a witness to the crime identified the defendant by name as the perpetrator immediately following the incident; such testimony is not hearsay when the identifying witness testifies at trial. Liscinsky v. State, 700 So.2d 171 (Fla. 4th DCA 1997).  In addition, an out-of-court identification of the defendant is sufficient to support a conviction.  Corroboration is not required. People v. Cuevas, 906 P.2d 1290 (Cal. 1995).

The identification of a defendant by a witness should not be confused with the identity of a defendant being discovered by a law enforcement officer.  An out-of-court identification by a witness may be suppressed if it is found to be unnecessarily suggestive or unreliable, as discussed above.  The defendant’s identity, however, is not subject to suppression, no matter how it was obtained, even if resulting from an unlawful arrest. I.N.S. v. Lopez-Mendoza, 468 U.S. 1032 (1984); Ware v. State, 679 So.2d 3 (Fla. 2d DCA 1996).  For example, if the officer learns the defendant’s identity after stopping him for a traffic violation, and the stop is later found to be illegal, the defendant’s identity is still admissible in a prosecution for driving with a suspended license.  However, at least one Florida court has held otherwise. Perkins v. State, 734 So.2d 480 (Fla. 4th DCA 1999)(identity of defendant must be suppressed when it was obtained as a result of an unlawful traffic stop).

            As a criminal defense attorney, you must always be vigilant and diligent in fully investigating every aspect of each out-of-court identification.  In most cases, you will find some flaw in the procedure or some degree of suggestiveness that calls into question the reliability of the identification.  Don’t let your client join the club of “convicted innocents”.