ALLEGANY, N.Y. -- In the election season to come, if past practice prevails the district attorney races will be relatively overlooked. North Carolina is one of the 47 states that use popular election to determine its prosecutors. Since 1998 there have been contests 22 percent of the time in general elections and primary challenges in another 21 percent of the elections.
Prosecutions by district attorneys in the United States account for 95 percent of all felony convictions. The D.A.s are given a significant amount of discretion. They choose whether to file charges, which charges to file and whether to engage in plea bargaining or to pursue a conviction at trial. Their decisions affect the accused along with the victims, their families and the community. We all have a stake, because we benefit from the deterrence of crime and because our tax dollars fund the criminal justice system.
How election pressures affect prosecutors' decisions is important. My recent research uses data from North Carolina to shed light on this impact.
We merged conviction data from each district with election returns over the period 1997 to 2010. We found that when an incumbent is challenged in a re-election bid in North Carolina, there are on average 13 percent more jury trial convictions in the district in that year relative to years without re-election campaigns. The distortion begins earlier. In the year before an incumbent runs for re-election, the average number of jury trial convictions increases by 7.2 percent compared to years without election concerns.
If re-election challenges lead to more cases taken to trial, do D.A.s in secure districts take fewer cases to court? Yes. Districts in which voters did not, even once, see a contested race in any election in the last 16 years experienced 11.6 percent fewer felony convictions per year from jury trials relative to those areas that had at least one contested election.
We found all of these results to be robust when considered in light of changes in the severity of the crime, changes in sentencing by the judges and the socio-economic characteristics of the area.
The expanded use of the jury trials is not necessarily due to resources available to the prosecutors or increased effort in election years. We found that the average maximum number of months in prison received by those sentenced to prison decreased by 2.6 months when an incumbent district attorney is challenged. The average sanction is reduced when less serious offenses, with shorter sentences, are prosecuted. Additional resources and effort would have increased the sentences as more would be invested in securing a longer conviction, discouraging pleas for lesser offenses.
Similarly, a re-election campaign leads to fewer convictions where community punishments (community service, probation, etc.) are handed out. The proportion of convictions which resulted in community punishments decreased by 10.5 percent when a challenger entered a re-election contest, and is 12.2 percent higher when a prosecutor is in a district that has been immune from contests.
Again, this is evidence of prosecutors in North Carolina being tougher when election pressures are greater. More criminal cases go to trial when election pressures are great. The evidence suggests trials, which typically generate longer sentences and garner publicity, are a vehicle used to gain support from voters. By acting hawkish, D.A.s attempt to signal to voters their skill and worth.
While a serious debate between the opposing effects of more severe sanctions - reduced crime versus increased costs - needs to be undertaken, it is discomforting if the courts are being used to extend time in political office. Either educated voters need to be wary of the distortions contested re-election campaigns can create, and not reward incumbents for inflating their statistics, or policymakers need to consider alternative mechanisms for appointing and retaining public prosecutors.
Bryan C. McCannon is associate professor of economics at St. Bonaventure University. He formerly taught at Wake Forest University.
