Weighing Reliability Against Suggestion

Weighing Reliability Against Suggestion

The fifth factor completes the list of inputs, and the remaining task is to combine them. The five factors measure how much independent memory of the crime the witness probably had, while the suggestive procedure supplies a push toward a particular person, so the court has to compare the two. Manson v. Brathwaite joined them in a single instruction: the Neil v. Biggers factors are considered, and against them is weighed the corrupting effect of the suggestive identification itself.¹ The weighing arrives only at the second stage of the inquiry. Due process concerns arise when law enforcement uses an identification procedure that is both suggestive and unnecessary, and even then suppression of the resulting identification is not the automatic consequence.²

The question the weighing answers is whether the procedure created "a very substantial likelihood of irreparable misidentification," a phrase that Biggers applied to out-of-court identification testimony after deleting the word "irreparable."³ Two outcomes are possible. If the court finds that the unnecessarily suggestive procedure created a very substantial risk of misidentification, due process is violated and the identification is excluded; if the reliability of the identification outweighs the suggestiveness, it is admissible and the jury decides how much it is worth.⁴ Because the factors feed one judgment, no single factor decides the outcome, and a court does not tally them like elements of an offense.

Weighing, rather than counting, follows from what the Due Process Clause guards against, which is a misidentification that the police procedure caused. Whether a procedure caused one depends on two quantities that vary independently: how much real memory the witness brought to the display and how much the display itself supplied. A witness who watched the culprit at close range for several minutes needs little help from a display, and a gentle display offers little help in any event, so the risk of a manufactured identification is low. A witness who glimpsed the culprit in poor light needs a great deal of help, and a display that singles out the suspect is ready to supply it. A checklist that scored only the factors would admit a corrupted identification whenever the witness happened to sound sure, while a rule that looked only at the procedure would exclude trustworthy testimony whenever the police were careless. The Court rejected the second rule in Manson for stated reasons: automatic exclusion keeps from the jury evidence that is reliable and relevant, and the totality approach still influences police behavior even though a per se rule would deter more.⁵

In Biggers the weighing had to absorb a bad factor. Seven months passed between the rape and the showup, in which two detectives walked the defendant past the victim and, at her request, had him repeat the words her attacker had used.⁶ The Court was inclined to agree that the police had not exhausted the search for physically comparable persons to build a lineup, and it still held that the evidence is not excluded merely because a confrontation was unnecessarily suggestive.⁷ It called the seven-month lapse "a seriously negative factor in most cases," and then set what lay on the other side of the scale against it.⁸ The victim had been with her assailant for up to half an hour under adequate light and had faced him directly, her description was more than ordinarily thorough, she had "no doubt" at the showup, and in the intervening months she had been shown between 30 and 40 photographs and had viewed lineups and showups without identifying anyone.⁹ That last fact did the most work against the delay, because a witness who had declined every earlier invitation to choose had "previously resisted whatever suggestiveness inheres in a showup."¹⁰ Weighing all of it, the Court found no substantial likelihood of misidentification, and the identification was properly allowed to go to the jury.¹¹

The suggestion side of the scale was nearly empty in Manson. An undercover officer who had bought heroin from a seller standing within two feet of him for two to three minutes in natural light described the seller to another officer within minutes, viewed a single photograph two days later, and said there was "no question whatsoever" that it showed the seller.¹² The Court acknowledged that identifications arising from single-photograph displays "may be viewed in general with suspicion."¹³ It found, though, little pressure on the witness to go along with the suggestion such a display carries: the detective had left the photograph at the officer's office and was absent when the officer first viewed it, there was no urgency, and because the officer examined it alone nobody's presence pushed toward an answer.¹⁴ The indicators of the officer's ability to identify accurately were therefore "hardly outweighed by the corrupting effect of the challenged identification itself."¹⁵ Biggers and Manson reach the same result by different routes. In Biggers a strong record outweighed a seriously negative factor and a showup that carried real suggestion, while in Manson the indicators were strong and the display itself exerted little force, so less had to be outweighed. The balance therefore moves with either side of the scale: more reliability offsets more suggestion, and a gentler display lowers what the reliability must offset.

A court of appeals ran the same balance on a more conspicuous display in United States v. Stevens. The wanted board showed Stevens in photographs larger than the others, in color when the others were not, and twice, features that could have drawn the victims' attention to him.¹⁶ The court assumed the procedure was unnecessarily suggestive and still found the identifications reliable, because two military police officers trained in observation had watched the attacker at close range for at least ten minutes, expressed high certainty on first seeing the photographs, and recognized them within about an hour of the attack.¹⁷ A display with several suggestive features therefore did not end the inquiry. The court asked whether those features were outweighed, and the officers' view, attention, certainty, and promptness answered that they were.

Foster v. California supplies the pole where nothing offsets the suggestion. The only eyewitness to the Western Union robbery besides the robbers first viewed a three-man lineup in which the defendant was close to six feet tall, the other two men were five feet five or six inches, and the defendant wore a jacket like the robber's, and the witness only "thought" the defendant was the man.¹⁸ A one-to-one meeting left the witness unsure, and a week or ten days later the police arranged a five-man lineup in which the defendant was the only repeat participant, and this time the witness was "convinced."¹⁹ The Court held that these suggestive elements made the identification "all but inevitable" whether or not the defendant was the robber, and that the procedure so undermined reliability as to violate due process.²⁰ In weighing terms, the only sign of reliability in the opinion is the witness's final conviction, and that conviction was itself produced by the repeated display, so the reliability side holds nothing independent of the suggestion. Foster predates the catalogue of five factors and is not a factor-by-factor holding, yet it shows what an empty reliability side looks like. Manson and Foster differ on both sides of the scale: in Manson an officer with a strong basis for a real memory viewed one photograph alone and unhurried, while in Foster the police put the same man before an uncertain witness again and again.

Because the weighing is a judgment and not a computation, judges applying it can disagree about the same display. Justice Marshall, joined by Justice Brennan, dissented in Manson on the view that a single photograph is inherently suggestive, that the police should have used a proper lineup, and that the reliability test does not adequately protect against suggestion.²¹ They would have reversed the conviction.²² The majority and the dissent thus assigned very different weight to the same display, and opposite outcomes followed. On an exam, a fact pattern near the line rewards an answer that names what sits on each side of the scale and says which side the facts favor.

The balance belongs to the judge as a question of admissibility. The defense raises it by a motion in limine, and the trial court hears proof on the five factors before the jury hears testimony about the identification.²³ If the court concludes that the unnecessarily suggestive procedure created a very substantial risk of misidentification, due process is violated and neither the out-of-court identification nor an in-court identification comes in.²⁴ If the court concludes that reliability outweighs the suggestiveness, the testimony is admissible and the evidence goes to the jury, which decides how much weight it deserves.²⁵ The screening function matters because jurors tend to give the greatest weight to a witness's confidence even though confidence is a poor gauge of accuracy, so a corrupted identification that reaches the jury may persuade it for reasons unrelated to memory.²⁶ Later in this chapter, the next section covers the hearing and the remedy of suppression, and the safeguards that remain after admission.

The most frequent error is counting factors. A court that tallies three favorable factors against two unfavorable ones has not weighed anything, because the factors differ in force from case to case and the display on the other side of the scale carries a weight of its own. Biggers shows a seriously negative factor outweighed, and Foster shows an identification excluded even though the witness ended up convinced, since that conviction was the display's product. A second error treats suggestiveness alone as enough to exclude. The Court declined to exclude evidence merely because a confrontation was unnecessarily suggestive, and the question stays whether the suggestion created a very substantial likelihood of misidentification once the indicators of reliability are placed on the other side.²⁷