I’ve gotten some great comments on my posts (here’s the last one) on Conviction Integrity Units (CIUs) for prosecutor’s offices. One group of these comments contains an important criticism of my support for CIUs.
The thrust of the criticism is that CIUs, as units within the same prosecutor’s offices that may have produced wrongful convictions in the first place, are not independent. They are, instead, part of the very agencyiesthat they investigate. Because of that, they may not do thorough, complete, and unsparing investigations. To the extent that the lawyers in CIUs investigate their own colleagues, they will always be tempted to make their fellow prosecutors or the agencies look good. Because the lawyers in the CIU are ultimately beholden to the elected prosecutor they serve, they may not feel free to follow the facts where ever they lead, without fear or favor. And at the very least, because they are not independent, the public may not view CIUs or their investigations as legitimate, even if they do every thing right.
This is an absolutely legitimate criticism. One can say the same thing about internal affairs units within police departments. (In fact, one critic of my proposal to require CIUs made exactly this analogy.)
I certainly agree that, ideally, the best arrangement would be independent CIUs: a unit outside the prosecutor’s office to investigate claims of innocence. This could be a group of independent lawyers, like a citizen’s review board to review police conduct, or a free-standing state agency, such as the North Carolina Innocence Inquiry Commission. Organizations like these, if fully independent and adequately empowered to accomplish their missions, would be the best choice. They could not be interfered with, staff would not be intimidated by owing their jobs to the prosecutor, and they would have more legitimacy in the eyes of the public.
The problem is that, in the U.S., there seems to be little political appetite for this kind of agency. North Carolina’s Commission is a counter example, but it stands virtually alone. In the U.S., where the vast majority of prosecutors are elected county officials who often wield considerable local political power, the chances that governments will create independent bodies to investigate prosecutors’ mistakes seem virtually nil. (The creation of independent police review boards in the U.S. has been very difficult, and police and their unions have fought them mightily from the start.) Should a state legislature consider the creation of independent CIUs, overwhelming political opposition by powerful prosecutorial lobbying organizations is a certainty, and would almost guarantee defeat.
The appealing thing about internal CIUs is therefore a bit counter-intuitive, but real: they can be created by prosecutors themselves, for their own agencies. No outside authority is necessary to have this happen. A prosecutor who sees the virtue of CIUs — in terms of ferreting out old injustices, learning from past mistakes, and building trust with the citizens (that is, voters) they serve — can simply go ahead and create one. This is exactly what Craig Watkins did in Dallas County after he was elected in 2006, and it is why some other DAs are doing it now.
Of course, not all prosecutors will take this step, and maybe those whose offices need it most are least likely to do it. (Surely, it is no accident that the prosecutors in Dallas, Houston, and Manhattan that have created CIUs have all been new to the office; their CIUs investigate their predecessor’s cases. Brooklyn, N.Y. prosecutor Charles Hynes, mentioned in the New York Times article last week concerning a wrongful conviction, is the exception; he has served since 1989, and therefore his CIU investigates his own cases.) That’s certainly a fair point. But right now, of course, there are no independent CIUs except North Carolina’s. By advocating the creation of internal CIUs as a best practice to be expected of any well-run prosecutor’s office, we can move the prosecutorial profession in the right direction even if it doesn’t happen everywhere, and even if internal CIUs are only a good model, but not the best one.
But this issue is important enough that I’d be very happy with a good model in many places — certainly happier than with the best model up and running, but only in one place.