Michigan v. Thomas

Brief Filed: 9/17
Court: Michigan Supreme Court
Year of Decision: 2017

Read full-text amicus brief (PDF, 194KB)

Whether the single photograph-identification procedure used by the officers created a substantial likelihood of misidentification, and thus violated the defendant’s right to due process.

Eyewitness Identification Research

The victim was robbed and shot by a hooded man whose face the victim could see for only “six or seven seconds.” Based on a police radio call describing the assailant as a “black male wearing dark clothing [with a] youthful appearance,” an officer stopped and searched the defendant 50 feet from the site of the shooting. Finding no weapon or outstanding warrants, the officer concluded that she lacked probable cause to arrest. She nonetheless took a cell phone photo of the defendant (later described as showing him “from the chest up, at an oblique angle somewhere between front facing and profile”) and, about an hour later, showed that photo to the victim in his hospital bed — asking only “was this the guy?” The victim immediately confirmed that the person in the photo was the attacker, though there were inconsistencies between his testimony at the crime scene and at the hospital (including the attacker’s size, facial hair and whether the victim had previously seen the attacker “around the neighborhood.”).

Based on the identification, the defendant was charged with armed robbery and attempted murder. He moved to suppress the photo identification as impermissibly suggestive, and also moved to suppress the victim’s subsequent in-court identification as tainted by the photo ID. The trial court suppressed both identifications and dismissed all charges. Over a dissent, a panel of the Michigan Court of Appeals reversed. (It did not reach the issue of the in-court identification, because it found the photo-identification admissible.) Recognizing that “[s]howing a witness a single photograph is considered to be one of the most suggestive photographic identification procedures,” the court held that, under the totality of the circumstances, the photo identification here was not so suggestive as to violate due process. In particular, the court highlighted that “the victim indisputably had the opportunity to observe the assailant, if only for a short time,” and that “the victim's description matched defendant,” even if it “was general and could [have] appl[ied] to many people.” The court also gave significant weight to the fact that other parts of the victim’s testimony were detailed (such as the hand the attacker used to hold the gun), and also stressed the short time that had elapsed between the attack and the ID and the victim’s high degree of certainty. The court also emphasized that the photograph the victim was shown did not “indicate that the person pictured [was] under arrest.”

This is a continuation of a line of briefs that APA has filed in recent years regarding the deep body of research on eyewitness identification and the factors that enhance erroneous identifications. Our briefs in 2011 in Perry v. New Hampshire in the U.S. Supreme Court (our brief was cited in the 2012 majority and dissenting opinions), and this year in Garner v. Colorado, directly address single suspect identification.

This was the unusual case in which the Michigan Supreme Court granted oral argument time to American Psychological Association as an amicus curiae, which argument occurred in a high school gymnasium in Detroit to a local audience of students and community members. (Video of the oral argument is available.) Less than a week after oral argument, on Nov. 1, 2017, the Michigan Supreme Court agreed with APA’s argument and concluded that a single photograph identification was sufficiently unreliable that the court of appeals order had to be reversed, the trial court’s order reinstated, and all charges against Thomas dismissed.