« Judge Blane Michael | Main | Lie Detection and the Technological Look-Back Principle »

Sunday, March 27, 2011

Federalism and Corruption: Which states effectively self-police themselves?

This last Friday, I had the pleasure of appearing on a panel with some very distinguished public officials in federalism and the fight against corruption in the United States. The panel was part of a conference on Policing, Regulating, and Prosecuting Corruption, organized by NYU's Center on the Administration of Criminal Law, a think tank run by Tony and Rachel Barkow. Ron Goldstock, the moderator of our federalism panel (as well as former director of the NY task force on organized crime and New York State Commissioner of the Waterfront Commission of New York Harbor), adopted the unconventional approach of posing to the panel hypothetical instances of corrupt behavior over which both state and federal prosecutors had jurisdiction. Imagining that we were a blue-ribbon commission tasked with figuring out rules for divvying up jurisdiction over corruption offenses, he asked us which level of government ought to handle the hypothetical offenses.

Being unhampered by responsibility, experience, or any practical knowledge, I had the temerity to press my strong prescription for subsidiarity: The feds ought to prosecute corruption cases involving state or local officials only if either (a) subnational authorities request such help, (b) the crime has significant interstate effects (say, on TARP money) that state officials would have insufficient electoral incentives to monitor, (c) the crime in question involves violations of national civil rights entrusted to the federal government's care (e.g., the police beating of Rodney King), or (d) objective criteria, developed by Main Justice in advance of specific disputes, indicates that state or city officials were too closely linked to the allegedly corrupt political figures to investigate their behavior in an independent manner. Absent one of these factors, the feds ought to back off and leave the cure for the corrupt behavior to non-federal prosecutors. On this theory, New York City's officials should never be subject to federal corruption investigations for garden-variety corruption -- bribes, extortion, gratuities, and conflicts of interest -- because the City's Department of Inspections, Conflict of Interest Board, very independent-minded district attorneys, and the NY Attorney General all provide sufficiently robust correctives for dishonest politicians. Needless to say, the practicing public officials on the panel were much too responsible and experienced to embrace my formula whole hog. But, to my surprise, Nora R. Dannehy and R. Blake Chisam (the two panelists most famous for their experience as federal watchdogs) seemed open to the idea of imposing pretty strict limits on the federal role.

So what is the chief obstacle to my theory of federalism in policing corruption? The problem is developing a good metric of subnational jurisdictions' capacity for controlling corruption. Such a metric would take into account factors like the independence of subnational watchdogs -- inspector generals, DAs, etc -- as well as the competitiveness of political parties, transparency of the governments' procedures, robustness of the local press, and any other political facts relevant to the monitoring and policing of corruption.

Here's my question: Does anyone out there know of academic work shedding light on states' capacity to police themselves? I'm looking for an index analogous to Heather Gerken's "Democracy Index" -- a sort of "corruption control index" -- that Main Justice could use in evaluating whether U.S. Attorneys should be permitted to go forward with corruption cases because subnational anti-corruption systems have, in some objective sense, "failed." Anything out there? Many thanks in advance for your tips.

Posted by Rick Hills on March 27, 2011 at 06:55 PM | Permalink

TrackBack

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8341c6a7953ef014e60247a88970c

Listed below are links to weblogs that reference Federalism and Corruption: Which states effectively self-police themselves?:

Comments

ohwilleke writes:

The existence of credible evidence of corruption, almost per se, is proof of inadequate ability to self-police.

But this statement rests on two implicit assumptions, both of which are manifestly incorrect. First, it must logically assume that the optimal level of corruption is zero, and, second it must logically assume that the level of corruption is purely a function of anti-corruption measures. But, like any other evil, corruption should be tolerated to the extent that the costs of fighting it exceed the benefits: We do not want to eliminate all corruption is the loss of effective government in other respects. (Imagine assuming that the presence of litter assumed some inadequacy in a city's sanitation system!) Moreover, corruption levels are a function of much more than a jurisdiction's system of rules and their implementation. NYC has an excellent system for detecting and punishing conflicts of interest -- but it also has a lot of "risk factors" for corruption (e.g., a very big budget, a lot of discretionary decision-making, a political culture that encourages deal-making, a lot of private money sloshing around, very valuable real estate, etc). That NYC's system of anti-corruption rules do not always succeed hardly means that they are inadequate or that federal prosecutions' marginal benefits will exceed the marginal costs. (Imagine arguing that, because a chronically overweight smoker had a heart attack, therefore the patient's heart surgeon was inadequate and that a second federal doctor would improve matters!)

Posted by: Rick Hills | Apr 1, 2011 9:31:46 AM

One index is here: http://www.corporatecrimereporter.com/corruptreport.pdf but it focuses largely on prosecuted cases of corruption, so there may be some bootstrap issues.

Posted by: ohwilleke | Mar 30, 2011 6:59:18 PM

The existence of credible evidence of corruption, almost per se, is proof of inadequate ability to self-police. It happens because someone has found a flaw in the self-policing system locally.

A very weak version of subsidiarity, that could call for a "heads up" notice to the appropriate state and local authorities naiively has some attractions, since an adequately self-policing entity could respond to those concerns. But, since someone who can thwart or hinder the investigation is likely to receive that "heads up" the case against doing so is strong.

Posted by: ohwilleke | Mar 30, 2011 6:53:39 PM

Hmmm. I like the idea, but after you get through the factors that are at least somewhat easier to pin down, here's the toughest. Often, if not most often, partisanship will be the biggest factor driving willingess/unwillingness for state and local actors to investigate or prosecute.

But given the ideal of ignoring partisan politics, as opposed to "accounting for and counterbalancing," I'd be loath to see a DOJ manual that says "only go after the mayor if the DA is of the same party." US DOJ already has the burden of avoiding appearances (or reality) of being partisan for the president's party, so openly accounting for party -- even if to negate it -- is playing with fire.

Also, it seems to me that countermanding partisanship might have a partisan disparate impact. On one hand, under-enforcement most likely occurs with party differences between state and local, e.g., a Republican State Attorney General goes after local Democrats. It can occur with differences among local officials, or among statewides. (It also occurs intra-party, where someone will go after rival Ds or Rs but not allied ones. Trying to define that would be even harder.)

While both parties have local monopolies, the larger one-party jurisdictions are the Democratic big cities. DOJ won't go after petty theft by rural GOP county commissioners, even if in a GOP state. We've already seen accusations for decades of DOJ's racism, because of the impact on big-city officials. Openly invoking party adds to that.

On the other hand, any system that does not account for it is not being realistic.

Posted by: anonny | Mar 28, 2011 12:23:26 PM

Post a comment