Some background on Lineup Fairness
“According to Devlin (1976), lineups were instituted through a Middlesex magistrate’s order in the mid 19th century. They were intended as a ‘fair’ replacement for the practices of courtroom identification, and showups, which were widely used in 19th century England, but widely recognized as potentially unfair to the defendant. Their origin indicates that the notion of ‘fairness’ is their raison d’etre. They are intended to secure an identification that can potentially incriminate someone, but also is fair to those who are subjected to it, particularly those who are innocent of the crime.”
“Study of the case law in many countries, as well as recent DNA-based exonerations in the U.S., indicates that lineups are not invariably fair – many innocent people are convicted after identification from a lineup by an eyewitness. The problem is significant because eyewitness evidence is cited as the most significant source of wrongful conviction (Scheck, Neufeld, & Dwyer, 2001). The DNA exoneration cases where the false conviction is established with near certainty show that eyewitness evidence has been largely responsible for false conviction in more than 70% of cases (www.innocenceproj.org). Wrongful identifications result from a number of failures of police procedure, however many of these are minimized when the lineups presented to witnesses are fair. If witnesses are induced to make a lineup identification, a fair lineup will expose the innocent suspect to an identification risk of 1/the number of persons in the lineup. A lineup biased against the suspect (because s/he stands out in some manner) adds additional risk. A lineup in which only two members (the defendant and 1 other) fit the witness’ description of the offender increases the identification risk to 1/2. It follows that we must be able to construct fair lineups, and to diagnose the fairness of lineups.”
“The original guidelines of how to construct lineups, as they have come down to us in the English law, contained these key ideas i) to put a ‘sufficient number of men…’ in a line, ii) ‘…who bear a reasonable resemblance to each other in physical appearance…’, iii) ‘…such that the identity of the suspect is not suggested to the witness…’. In practice, the ‘sufficient number’ came to mean eight in England, and six in the U.S. The degree of resemblance proved impossible to specify precisely, but case law in England evolved to specify that the resemblance should extend to height, weight, clothing and general appearance.”
* from Malpass, Tredoux & McQuiston-Surrett, (2007) Lineup Construction and Lineup Fairness.