The Flawed NRC Report: What Incentivizes Prosecutors?

(This is Part 8 in my criticism of the recent National Research Council report on incarceration. Here are Part 1 (drug war), Part 2 and Part 3 (longer sentences), and Part 4,  Part 5, Part 6, and Part 7 (admissions and prosecutors).) 

As my earlier posts have made clear, that the NRC report does not really focus on prosecutors is disappointing, given the apparently outsized role they have played in driving up prison populations. As a result, the report’s policy recommendations do not target the real causes of prison growth. Given how generally understudied and under-regulated prosecutors are, this represents a truly lost opportunity.

Compounding this error is the fact that when the report does talk about prosecutors, it does so poorly. In this post, I want to examine the report’s analysis of the political incentives of prosecutors. The motivations it highlights are likely not the core ones driving prosecutor behavior, and it worth considering both why those factors aren’t so important and what some important ones could be.

The bulk of the report’s discussion of prosecutors is the following: 

Incentives for supporting certain kinds of crime-related initiatives also tend to be misaligned across different levels of government. For example, it is relatively easy for local government officials to advocate increased sentence lengths and higher incarceration rates that state government officials are typically responsible for funding (including the building and running of state penitentiaries). Yet, despite taking hard-line positions on crime control, local governments often hire too few police officers (since cities and counties are responsible for paying nearly all local police budgets)….

In the United States, most prosecutors are elected, as are most judges (except those who are nominated through a political process). Therefore, they are typically mindful of the political environment in which they function. Judges in competitive electoral environments in the United States tend to mete out harsher sentences (Gordon and Huber, 2007; Huber and Gordon, 2004). In contrast, prosecutors and judges in many European countries are career civil servants who have evolved a distinctive occupational culture with a less punitive orientation, partly as a result of differences in legal training and career paths between the United States and European countries ….

As I noted in my earlier post, the first paragraph gives far too little attention to the quite-significant moral hazard problem that arises when a county official can spend unlimited state dollars; I won’t belabor that point here. It’s the second paragraph I want to turn to now.

The report argues that as local officials, prosecutors and judges are sensitive to local demands to be tough on crime. The only papers the report cites are those by Gordon and Huber, which look at judicial behavior in Kansas (2007) and Pennsylvania (2004). These are both well-designed, insightful papers,* but they focus on judges, and on a subset of judges at that. (The report’s statement that “most judges [are elected]” elides over the fact that the way in which judges are elected varies widely, with likely implications for their behavior.)

But what about the effect of elections on prosecutors? The Gordon and Huber papers are not particularly relevant, though the report does not mention this. The paper on Pennsylvania bases its model on the fact that Pennsylvania judges, facing retention elections once a decade, operate in low-information environments; in such settings, judges are most concerned with a single bad high-profile case. That simply doesn't describe the political world of prosecutors, who are elected more regularly and likely in higher-information settings (although just how much higher is a valid empirical question to consider).

The paper on Kansas focuses on conditions more like what a prosecutor would face—shorter-term and thus higher-information elections—but it addresses a policy issue not available to prosecutors: how the choice between partisan elections and appointment/retention elections shape behavior.We can change how judges are selected, and have done so repeatedly; elections are here to stay for prosecutors.

That the report only relied on papers about judges is particularly distressing because there is at least one paper directly on point for prosecutors. Ronald Wright has published a paper bluntly titled “How Prosecutor Elections Fail Us.” He points out that DA turnover is low (40% of the DAs in his sample have held their positions for over 12 years), and in a 10-year, 10-state survey of electoral outcomes he finds that DAs win 95% of the elections in which they run (and they ran for re-election in 75% of the races). More tellingly, in 85% of the races in which they ran, incumbents ran unopposed.** Such lack of competition can be seen anecdotally: when Joe Hynes recently lost his re-election campaign after 23 years in office, he became the first sitting DA in Brooklyn to lose a re-election bid in over 100 years.

In other words, district attorneys appear to be relatively political aloof. Of course, low turnover could reflect a powerful electoral accountability: DAs are so afraid of angering the electorate that they are near-perfect agents and thus get easily re-elected. But this seems unlikely. Regardless, Wright’s point is one that demands attention, and the report  omits it.

So what does motivate DAs? It’s a good question, and one that has received very little analysis. One colleague, though, suggested an idea that is to me quite interesting: what if what matters isn’t election to the district attorney’s job, but to the next one? What if DAs are punitive because it makes it easier to run for judge, or AG (and from there to governor), or Congress, etc., etc., etc.? And what if the political ambitions of DAs have grown stronger as, say, crime rates rose and the DA became a more high-profile, and seemingly more important, job? 

The report points out that elected DAs act differently than their more-bureaucratic counterparts, implicitly suggesting that we should shift toward a more-European system. (Although another article by Ron Wright, written with Marc Miller, gives some pause, given that its title is "The Worldwide Accountability Deficit for Prosecutors"). But while that could matter for line prosecutors, if DAs are still looking to the next higher office, then it isn’t really the electoral pressures of the DA’s office that is driving punitive behavior, and thus we need to think about how to regulate DA behavior in a much different way.

This point is clearly hypothetical, one that merits closer attention but hasn’t received any (yet). But it indicates that the simple analysis of the NRC report clearly does not do the complexity of this issue justice.

 

* The paper on Kansas exploits the fact that about half of Kansas judicial districts rely on appointment and retention elections, while the other half rely on partisan elections. The paper on Pennsylvania takes advantage of different judges in the same district being elected at different times. A paper by Jason Czarneski, uncited in the report, identifies similar punitive-right-before-an-election behavior by elected state supreme court justices in Wisconsin.

** The more-relevant Gordon and Huber study on Kansas looks at the implications of facing a viable competitor. But apparently few DAs face such, again reducing the applicability of that paper to DAs, however valid it is at explaining judicial outcomes.