The most egregious cases of informant unreliability occur where an entire case turns on the testimony of a single compensated snitch. The dangers of wrongful conviction in this scenario are so obvious that numerous states have or are considering corroboration requirements. But informant testimony can produce wrongful convictions in another way, and that is by making weak cases look stronger than they are. For example, Florida Today ran a story last week (updated link) on the probable innocence of Gary Bennett. Bennett was convicted based on a now-discredited dog sniff expert and the testimony of a jailhouse snitch. Similarly, in the high profile case of Cameron Todd Willingham, the Texas man was convicted and executed for arson based on a combination of poor forensic science and the testimony of a jailhouse snitch who later recanted. See previous post.
Such cases are not accidents. Jailhouse snitches are infamous for fabricating information about homicide and other high-profile cases, and offering the information to law enforcement without any solicitations or promises on the part of the government. In other words, the very existence of the case generates the bad evidence because of the general expectation in the offender population that such information will eventually be rewarded. This snitch testimony, however, makes the original case look stronger than it really is. This problem cannot be solved by corroboration requirements, since the informant’s information is automatically “corroborated” by the pre-existing weak evidence. Yet another reason to restrict the use of jailhouse informant testimony.