August 17, 2010

The APD has attempted to fix the problems that led to the Johnston tragedy by reorganizing its Narcotics Unit and strengthening its Office of Professional Standards (which houses the department's Internal Affairs and Corruption Units). While these steps are positive, they continue to rely on the police to self-regulate, a task that the APD and police departments generally often do poorly. (In Atlanta's case, a report written in response to the Johnston incident found repeated breaches of APD policies and a police culture that ignored these breaches.) Indeed, if Johnston's death was the result of "rogue officers," strengthened policies are unlikely to prevent future tragedies.

Atlanta City Council, apparently skeptical of the APD's ability to self-police, reacted to the Johnston tragedy by creating a Citizen Review Board and empowering it to investigate citizen complaints against the APD. The creation of the CRB is a positive step away from unfettered police discretion and toward civilian oversight.

But informant use presents a unique problem that this CRB (and others like it) is ill-suited to address. In particular, Atlanta's CRB has the authority only to respond to citizen complaints. Such complaints provide an effective oversight mechanism only for misconduct that is likely to be reported. For example, civilians who are subject to excessive use of force or who suffer false arrest or imprisonment are likely to bring complaints to the CRB. But police use (and misuse) of informants takes place almost entirely in secret, and those civilians who might have knowledge of police misconduct -- the informants themselves -- are unlikely to report it. After all, informants frequently assist the police under threat of criminal prosecution for a prior offense. And any complaint to a CRB by such an informant will likely result in the informant's incarceration, thus creating a strong disincentive to report misconduct.

For instance, prior to raiding Johnston's home, APD officers stopped a suspected drug dealer, planted drugs on him, and threatened to arrest him if he didn't provide information to incriminate someone else. That dealer falsely told police that a dealer with a significant amount of cocaine was at Johnston's home. The police then lied on an affidavit to secure a warrant, thus leading to the fatal raid. As this recounting shows, a CRB would not have prevented the Johnston incident, both because there was no time for a citizen complaint to be filed and because the informant was unlikely to complain for fear that the police would follow through on their threat of prosecuting him for the planted drugs. Instead, police misuse of informants will likely come to light only in situations like the Johnston tragedy, where an innocent civilian is injured, killed, or falsely imprisoned as the result of police misconduct. Meanwhile, the more run-of-the-mill misconduct involving false or fabricated informant testimony (used to incriminate the actually guilty), informants permitted to continue to engage in low-level criminal conduct, and informants coerced into engaging in dangerous activities with little to no training will continue unchecked.

Of course, bad actors intent on misconduct are difficult to deter completely, but two approaches would work better than those taken in Atlanta. First, affirmatively obligating the police to report on informant use would allow for insight into systemic problems before they lead to incidents like the Johnston tragedy. For instance, it is likely that the events leading to Johnston's death were not the first time that the officers involved had fabricated informant testimony. Informant use data might have identified them as wrongdoers due to informant use at a rate or in a manner inconsistent with their fellow officers.

Second, forbidding the police and prosecutors from trading pre-conviction leniency for informant assistance would make tragedies like the Johnston incident less likely. I've argued that many such bargains violate the Thirteenth Amendment, but regardless of the basis for the prohibition, such a ban would have two positive effects. The first is that it would place the informant-state relationship under the supervision of the courts, which could guarantee that informants are used appropriately. The second benefit is that a ban would make police hesitant to offer such deals and civilians would be less likely to accept them, knowing that they are prohibited and thus unlikely to be honored. In the Johnston case, such a ban may have prevented the police from receiving the false information tying Johnston's home to drug dealing and thus prevented her death.

The problem of informant misuse is a difficult one, but tragedies like Johnston's death deserve solutions that are tailored to the problems that cause them.