Massachusetts Supreme Judicial Court limits impact of landmark identification case

Issue March/April 2018 By Christopher W. Spring
Criminal Law Section Review


A trio of Massachusetts Supreme Judicial Court decisions delivered in the last three years has fundamentally changed the rules regarding the identification of criminal defendants at their trials by commonwealth witnesses. While two decisions issued on the same day in 2014 greatly limited a prosecutor’s ability to elicit an in-court, eyewitness identification of a defendant, the commonwealth won an important case earlier this month that will make such an in-court identification considerably easier. This article will review the current state of the law regarding in-court, eyewitness identifications in criminal cases in Massachusetts.

Seismic shift in Massachusetts identification jurisprudence: a witness generally must identify the defendant prior to trial before identifying him in court.
Three years ago, in Commonwealth v. Crayton, 470 Mass. 228 (2014), the Massachusetts Supreme Judicial Court considered the circumstances under which a commonwealth witness will be permitted to identify a criminal defendant in front of the jury when the police failed to conduct an identification procedure prior to trial. The defendant in Crayton was convicted of two indictments charging him with possession of child pornography after he allegedly viewed illegal images on a public computer at a Cambridge library. Two students, in eighth and ninth grades, respectively, were studying in the library when they peeked over a man’s shoulder and saw an underage naked girl on the computer screen he was viewing. The students notified a library employee who was able to determine the man had used another person’s library card to sign onto the computer. Meanwhile, the man logged off and left the library. While nobody knew the man’s identity, a description of his physical characteristics was shared with other library employees.

A subsequent investigation by library management and the police led to the discovery of videos on the subject’s computer depicting an underage girl masturbating and 210 photographs of children who were engaged in sex acts. When the defendant entered the library the following day, an employee thought he matched the suspect’s description and called the police. The defendant acknowledged being present at the library the day before and using the computer at the center of the investigation. While the defendant denied looking at child pornography, the police were able to connect his personal information to the computer at the time the child pornography was downloaded, and he was indicted.

The police never administered any type of identification procedure, such as a photo array or a lineup, to allow the students to attempt to confirm the defendant was the same person who viewed child pornography in the library. Instead, the prosecutor waited until the trial, which took place more than two years later, and asked the students to identify the defendant in the courtroom as he sat at the defense table in front of the jury. The defendant had filed a pretrial motion asking the trial judge to prohibit an in-court identification, as such a procedure would be inherently and unnecessarily suggestive. The trial judge denied the motion and the students predictably identified the defendant, leading to his convictions. 

On appeal, the Supreme Judicial Court grappled with the question of whether it is ever proper for a witness to identify a defendant in court when no out-of-court identification was previously made. The court began its analysis by noting that the Massachusetts Declaration of Rights prohibits introduction of an out-of-court, eyewitness identification when “… the identification is so unnecessarily suggestive and conducive to irreparable misidentification that its admission would deprive the defendant of his right to due process.’ Id. at 234, quoting Commonwealth v. Walker, 460 Mass. 590, 599 (2011), citing Commonwealth v. Johnson, 420 Mass. 458, 463-464 (1995), and Commonwealth v. Thornley, 406 Mass. 96, 98 (1989). The question became, then, whether the level of suggestiveness inherent in an in-court identification procedure conducted by a prosecutor in front of a jury was permissible. 

The court identified three differences between an in-court identification procedure and an out-of-court showup conducted by the police. First, when the identification procedure is conducted in the courtroom, the jurors have the opportunity to watch it and potentially assess the eyewitness’ level of confidence in the identification. Second, an eyewitness who participates in an in-court identification procedure is immediately subject to cross-examination by the defense attorney. Third, if the prosecutor intends to have an eyewitness identify the defendant in court, defense counsel has the opportunity to propose alternative, less-suggestive identification procedures. The court ruled none of these factors compelled the conclusion that identification procedures conducted for the first time in a courtroom are proper. Accordingly, “[w]here an eyewitness has not participated before trial in an identification procedure, we shall treat the in-court identification as an in-court showup, and shall admit it in evidence only where there is ‘good reason’ for its admission.” Id. at 241. What would constitute a good reason? If the identifying witness previously had a relationship with the defendant (such as a romantic partner or a neighbor), or if the identifying witness was a police officer who saw the defendant commit a crime and arrested him, a trial court could reasonably find there was a good reason to forgo an out-of-court identification procedure.

While the court’s decision was to apply prospectively, the defendant’s convictions were reversed as a result of multiple errors at his trial. The Supreme Judicial Court’s decision in Crayton reverberated across the commonwealth. All over the state, prosecutors scrambled to arrange for witnesses to identify defendants before trial dates, as identification jurisprudence was forever changed in Massachusetts. 

An attempt at an out-of-court identification procedure isn’t enough — the witness must actually identify the defendant.
The same day Crayton was decided, the Supreme Judicial Court considered a related issue. What happens if a witness attempts but fails to make a pretrial identification, but is then able to identify the defendant at trial? This question was answered in Commonwealth v. Collins, 470 Mass. 255 (2014).

According to trial testimony, the defendant had agreed to sell two kilograms of cocaine to two men for the price of $38,000. The transaction was to occur at the Dorchester apartment of the girlfriend of one of the buyers. The defendant arrived with a gun and, instead of providing the cocaine, demanded the money. When they buyers refused to turn over the cash, the defendant shot them both. One man was killed and the other was injured. The defendant grabbed the buyers’ bag of money and fled, only to be captured a couple of weeks later.

During the robbery, the decedent’s girlfriend was in her bedroom in the apartment. She saw the defendant “for a second” before the shooting started and for another second as the defendant fired his gun into the living room. Exactly one month after the crime, a Boston police detective showed a sequential photo array to the girlfriend, which included the defendant’s photograph. A sequential photo array involves showing the witness one photo at a time, whereas a simultaneous photo array involves displaying multiple photos on a single piece of paper. According to the detective, the girlfriend said “no” after viewing each of the eight photos, but she later said she was considering two of the photos (including the defendant’s). 

At the defendant’s trial, which occurred more than two years later, the girlfriend testified during the commonwealth’s case-in-chief. On direct examination, she described viewing the photo array. She testified that she picked out two photos and said, apparently for the first time, that the defendant’s photo looked more like the person at her bedroom door. On cross-examination, the defense attorney confronted the girlfriend about never having previously said that the defendant’s photo looked more like the shooter than the other photos. On redirect examination, the girlfriend said she couldn’t make a positive identification from the photo array because she had seen the shooter only from the side. The prosecutor then asked the girlfriend if she saw the shooter in the courtroom, and the girlfriend pointed to the defendant. A Suffolk Superior Court jury convicted the defendant of second-degree murder, armed assault with intent to murder, and unlicensed possession of a firearm. 

On appeal, the defendant argued he was the victim of ineffective assistance of counsel because his attorney had not objected to his in-court identification by the victim’s girlfriend. At the time of the defendant’s trial, such an identification was not impermissible. Therefore, defense counsel was not ineffective and the defendant’s convictions were upheld. However, the Supreme Judicial Court ruled that in future cases, a witness who had been unable to unequivocally identify the defendant in a pretrial, out-of-court identification procedure would not be permitted to later identify the defendant during the trial.

Collins differed from Crayton in that an out-of-court identification procedure was attempted. The court held, however, that when an in-court identification follows a failed out-of-court attempted identification, “the danger is that the jury may disregard or minimize the earlier failure to make a positive identification during a nonsuggestive identification procedure, and give undue weight to the unnecessarily suggestive in-court identification.” Id. at 262. Further, witnesses who couldn’t positively identify the defendant during pretrial identification procedures were “likely to regard the defendant’s prosecution as confirmation that the defendant is the ‘right’ person and, as a result, may develop an artificially inflated level of confidence in their in-court identification.” Id. at 262-263. Therefore, in all cases litigated after Collins was decided, a witness will be precluded from identifying the defendant at trial if a pretrial identification procedure failed, unless there exists a “good reason.” In a footnote, the court suggested a “good reason” might exist under a Collins scenario in a case where the victim knew the defendant (for example, if they were involved in a domestic relationship) but failed to identify him during a pretrial identification procedure as a result of fear or an unwillingness to cooperate with the police. This scenario seems somewhat far-fetched, because a police officer likely would not administer a pretrial identification procedure when the victim claims to know the defendant. 

The pretrial identification rule announced in Crayton is satisfied even if the out-of-court identification is a suggestive showup.
The question left open in Collins was whether a witness could identify a defendant at trial if a pretrial identification was successfully made by way of a suggestive showup procedure. In Commonwealth v. Dew, SJC-12225 (decided Nov. 6, 2017), a majority of the Supreme Judicial Court concluded a successful suggestive showup identification is sufficient to allow an eyewitness to then make an in-court identification. 

The victim in Dew was a pizza delivery driver who was robbed in a Beverly parking lot by a black man wearing a dark jacket and a red scarf that covered the lower part of his face. The robber held a knife to the victim’s throat and stole a small amount of cash. When the robber walked away, the victim called the police. Based on the victim’s description of the assailant and the direction in which he had walked, the responding police officer went to a nearby apartment to search for the defendant (who was known to the officer). The officer found the defendant in a bedroom wearing jeans, sneakers, and a red T-shirt. The defendant’s black winter jacket was nearby. The officer handcuffed the defendant behind his back, draped the black jacket over his shoulders, and brought him outside to conduct a showup identification procedure. The defendant was placed next to another man, who was described as Hispanic with a “tannish” complexion. Two uniformed police officers stood with the men — one on either side. The victim was driven to the defendant and identified him “fairly immediately” from a distance of about 15 feet. The victim said he recognized the defendant’s eyes and the unique shape of his nose. The defendant was arrested and charged in Essex Superior Court with masked armed robbery and with being a subsequent offender. The victim was permitted to identify the defendant at trial and a jury found him guilty. On appeal, the defendant challenged the denial of his motion to suppress the showup identification and the trial judge’s decision to permit an in-court identification.

While Massachusetts appellate courts have repeatedly warned that showup identification procedures are generally disfavored as inherently suggestive, those same courts have repeatedly allowed the results of showup identifications to be admitted in court. The Supreme Judicial Court has said that a one-on-one pretrial identification procedure does not violate a defendant’s due process rights unless it is determined to be “unnecessarily suggestive.” Commonwealth v. Austin, 421 Mass. 357, 361 (1995). Good reasons to conduct these types of identification procedures include: the nature of the crime; police concerns for public safety; the need for efficient police investigation in the immediate aftermath of a crime; and confirmation that the police have accurate investigatory information and are on the right track to solving the case. Id. at 362. In Dew, the court concluded the defendant had failed to prove by a preponderance of the evidence that the showup was unnecessarily suggestive, and the out-of-court identification was therefore properly admitted in court.

The defendant argued the inherent suggestiveness of the showup identification prevented the eyewitness from offering an unequivocal identification of the defendant, as required by Collins. While acknowledging that “the defendant raises important issues regarding the potential disproportionate impact of inherently suggestive, in-court identification testimony on the jury,” the court determined that because the eyewitness’ identification of the defendant at the showup procedure was unequivocal, it was proper to also allow an in-court identification. Dew, supra, slip. op. at 19. The court noted that a trial judge has the authority, based on common law principles of fairness, to exclude otherwise-admissible, unreliable eyewitness identification testimony if its probative value is substantially outweighed by the danger of unfair prejudice. 


Chief Justice Ralph Gants, who authored Crayton and Collins, penned a concurrence joined by Justice Kimberly Budd. The chief justice argued that the victim in Dew should not have been permitted to identify the defendant in court. The in-court identification, according to the chief justice, had minimal probative value but contained a “substantial” risk of unfair prejudice to the defendant. Id. (Gants, C.J., concurring, slip. op. at 7). “Allowing an in-court identification in addition to a showup identification creates a risk that the jury will gloss over these particular aspects of the showup identification and simply accept the subsequent in-court identification.” Id. (Gants, C.J., concurring, slip. op. at 8-9). Because the showup identification here was seriously flawed, according to the chief justice, it had been an abuse of discretion for the trial judge to have allowed the in-court identification. Nevertheless, the chief justice concluded the error had not been prejudicial, and it had therefore been proper for the court to have affirmed the defendant’s convictions.


For three short years, criminal defense attorneys rejoiced at the incredibly favorable identification jurisprudence created by Crayton and Collins. Criminal cases were regularly being dismissed on the morning of trial when witnesses had been unable to identify defendants prior to the trial date. Crayton offered important safeguards to criminal defendants who advanced a misidentification defense at trial. Dew will substantially limit the impact of Crayton. A suggestive showup identification procedure can now be used as the predicate for an in-court identification, as long as the showup was not “unnecessarily” suggestive. Dew will increase the likelihood of criminal defendants being convicted of crimes they did not commit. 

Other Articles in this Issue: