National Study Prosecutor Elections

The Prosecutorsand PoliticsProjectNational Studyof ProsecutorElectionsFEBRUARY 2020

Table of ContentsAbout the Prosecutors and Politics Project . 3Executive Summary . 4About the Study . 8State Reports . 10Alabama . 11Alaska . 18Arizona .19Arkansas . 22California . 26Colorado. 35Connecticut . 39Delaware . 40Florida . 41Georgia . 45Hawaii . 53Idaho. 55Illinois . 62Indiana . 77Iowa . 90Kansas . 104Kentucky . 117Louisiana .125Maine . 132Maryland .134Massachusetts.138Michigan . 141Minnesota .152Mississippi . 163Missouri . 167Montana.1831

Nebraska . 190Nevada. 202New Hampshire . 205New Jersey . 207New Mexico.208New York . 211North Carolina . 220North Dakota . 227Ohio. 234Oklahoma .246Oregon .251Pennsylvania . 256Rhode Island . 266South Carolina. 267South Dakota . 270Tennessee. 278Texas . 283Utah . 306Vermont .310Virginia . 313Washington. 329West Virginia . 335Wisconsin . 343Wyoming . 353Appendix . 357Acknowledgments . 3612

About the Prosecutors and Politics ProjectThe Prosecutors and Politics Project is a research initiative at the University ofNorth Carolina School of Law. Founded in 2018, the Project studies the role ofprosecutors in the criminal justice system, focusing on both the politicalaspects of their selection and their political power. The Project endeavors tobring scholarly attention to the democratic accountability of electedprosecutors, to increase our understanding of the relationship betweenprosecutors and politics through empirical study, and to publicly shareresearch in order to increase voters’ knowledge about their electedprosecutors and broader criminal justice issues.For more information about the Prosecutors and Politics Project, its mission,and its research, please visit prosecutors-and-politics-project/Questions about this report should be directed to the author, ProfessorCarissa Byrne Hessick (chessick@email.unc.edu).3

Executive Summary*American prosecutors wield significant power in the criminal justice system.They must decide when to file charges, which crimes to prioritize, and howlenient or harsh to be in plea bargaining. Prosecutors are entrusted with thispower, in part, because they are accountable to voters.In most states, that accountability comes in the form of direct elections. Fortyfive states elect prosecutors on the local level. These local elections provide apowerful check on the power that prosecutors wield—at least in theory. Buthow does that check operate in practice? Put differently, how much of a choicedo voters have about who will make important criminal justice decisions intheir communities?This report presents the results of a nationwide study of prosecutor elections.The first of its kind, the study gathered data from every jurisdiction that electsits local prosecutors in a recent election cycle. The study showed greatvariation in elections across the country. Some elections gave voters choicesin both primary and general elections to choose their local prosecutor. Butother elections were entirely uncontested. And some elections did not evenhave a single candidate on the ballot.Whether an election gave voters a choice seems to depend on two differentfactors. The first of those factors is the population in the district where theelection was held. Communities with large populations tended to have morethan one candidate in their elections, while communities with smallpopulations tended to have uncontested elections.Please note that this report updates and supersedes a similar report that was published in January 2020. Sincethat report was released, the underlying dataset has been updated to reflect new and corrected information froma small number of jurisdictions. Those corrections caused minor changes to the national analysis in the executivesummary.*4

CONTESTED ELECTIONS BYPOPULATIONContested ElectionsUncontested ESS THAN 100K1M There are far more communities with small populations than there arecommunities with large populations. There are more than 1,700 communitieswith fewer than 100,000 people who elect their prosecutors, as compared toonly 43 communities with more than one million people. Consequently, thereare many prosecutor elections that do not give voters any choices. Of themore than 2,300 jurisdictions that elect their prosecutor, fewer than 700presented voters with more than a single candidate on the ballot in theelection cycle we studied.Although most prosecutor elections are uncontested, most voters tend to livein jurisdictions that are more likely to give them a choice. More than a quarterof the entire U.S. population lives in only 35 prosecutor districts. And morethan half of the entire U.S. population lives in only 147 prosecutor districts.The largest of those districts is Los Angeles County, California, which has apopulation of nearly ten million. * In contrast, the smallest prosecutor districtis Arthur, Nebraska, which has a population of 460.The second factor that affects the availability of choice in prosecutor electionsis whether the sitting prosecutor decides to run for reelection. If it is an openA full list of the largest 147 districts can be found in the Appendix. All population data is drawn from the 2010census.*5

seat election—that is, an election in which the incumbent does not run—thenvoters are more likely to have a choice between candidates. But when theincumbent prosecutor runs for reelection, he or she often is the only candidatein the election.CONTESTED ELECTIONS BYELECTION TYPEContested ElectionsUncontested Elections47%75%53%25%INCUMBENT RUNNINGOPEN SEATThese two factors reinforce one another. If an incumbent runs for reelectionin a community with a small population, he or she is only likely to face achallenger 22% of the time. In contrast, open seat elections in communitieswith large populations were contested 100% of the time.6

ELECTION TYPE AND POPULATION100%82%68%56%45%22%LESS THAN 100K34%30%100K-249K250K-1MIncumbent Running1M Open SeatWhile our study clearly demonstrates a relationship between contestedprosecutor elections and both population and election type, this report mayraise more questions than it answers. Why do the number of contestedelections vary so much based on population and election type? How can weexplain the variation in contested election rates across states? What would ittake to meaningfully increase the number of contested elections?These are all important questions that are worth asking. We hope that thisreport will spark further study and exploration of prosecutor elections. Andwe hope to provide more data on this and related questions in future reports.7

About the StudyBeginning in February 2018, the Prosecutors and Politics Project begancollecting information about the most recent election cycle in each state thatelects its local prosecutor. For most states, that meant we collected data fromthe 2014 or 2016 election cycle. But in some states the most recent electionhad occurred as far back as 2012 and as recently as 2017. In total, we collectedelection results for 2,314 districts across 45 states.The dataset includes the name, party affiliation, and incumbency status ofeach candidate who ran for local prosecutor. The dataset also identifies whichcandidate won each election and the percentage of total votes each candidatereceived. These variables allow us to determine whether each district had anopen seat, whether the district held a contested election, and whether theelection resulted in the unseating of an incumbent. *We compiled our data from public sources. The decentralized nature ofprosecutor elections made it extremely difficult and time consuming to collectthis data. When available, the data was obtained from official election records.When those official records were not available or when they did not includethe relevant information, data was obtained from websites or media accounts.When those additional sources did not contain the data, we endeavored toobtain the information by emailing or calling county officials. But we were notable to obtain all variables for every election.We define open seat elections as those in which the incumbent did not seekreelection. We define a contested election as an election in which there aretwo or more candidates in either the general election or at least one primaryelection. We chose to aggregate the general and the primary elections whenclassifying contested elections, rather than to report contested primaryelections and contested general elections separately, because we sought toanswer the question whether voters have a choice about who will serve astheir prosecutor. The partisan composition of some districts may be solopsided that all important political fights occur in the primary; in others, thosefights may occur in the general election. In only a few districts are both theprimary and the general elections significant. Of course, when contested*The full dataset is available at Id doi:10.15139/S3/ILI4LC8

elections occur only in primary elections and not in general elections, it mayresult in some voters being excluded from the decision about who shall serveas prosecutor.Our findings make clear that two factors—the size of a community’spopulation and whether the sitting prosecutor runs for reelection—arecorrelated with whether voters have a choice in prosecutor elections. But itwould be a mistake to discuss prosecutor elections using only aggregatednational data. The states, and sometimes even counties, have significantcontrol over how prosecutor elections are conducted. While most states holdelections every four years, others hold elections every two, six, or eight years.Some states hold non-partisan elections. Several states do not hold primaryelections for prosecutor when the election is uncontested, and some do nothold a general election when it is uncontested. Three states allow theircounties to decide whether to elect prosecutors or whether to have themappointed. And one state imposes term limits.As this report demonstrates, not only do states have different rules about howprosecutors are elected, but they also have very different rates of contestedelections. For example, only 6% of the prosecutor elections in North Dakotaoffered voters a choice between at least two candidates. In contrast, 100% ofelections in Hawaii were contested. It is not clear what drives these differencesacross states. The differences do not, for example, map neatly onto thedifferent legal rules that states have adopted in selecting their prosecutors,such as the length of term or whether the office is partisan or non-partisan.Why population and the presence of an incumbent candidate might affect thecontestation rate of an election is not immediately obvious. And thedifferences between states are not explained by the different methods usedto select prosecutors. We do not offer explanations for these findings here.But we hope that our data will prompt further attention and study.9