Association for Public Policy Analysis & Management

Since the fateful 2000 election, scholars have examined many aspects of  election administration including administrative burden (e.g., Burden, et al., 2012), the effects of partisanship on administration of election reforms (e.g., Kropf, Vercellotti, Kimball, 2013) and the effectiveness of voting technology in the counting of ballots (e.g,, Kropf and Kimball, 2012). However, for the most part, budgets for election administration have received little examination due to the lack of data at the level at which elections are administered (counties in most states; townships and some cities in others). Because elections are primarily funded at the local level, local level financial reporting (Marlowe, Kumawala, and Neely, 2014) has tremendously limited the ability to estimate accurately the full cost of elections. So, not only is little known about how much is spent but even less is known about how local economic, administrative, and political conditions might influence the amount of money being spent at that local level and whether this influences the outcomes of elections.

Yet arguably, the amount of money an election jurisdiction is able to spend on the nuts and bolts of election administration (e.g., voter education, training pollworkers, renting locations for polls, upkeep and maintenance of voting equipment) has tremendous potential effects on whether or not a voter’s ballot will count. Thus, we ask the question of what are the effects of the budgets on program administration? In this case, we examine election program administration, taking into account economic, administrative structure and political conditions.

We were able to locate some of the very limited data within North Carolina. Using these data, unique and detailed county-level budget data from North Carolina counties from 1996-2010, we are able to examine the effects of election spending on election administration program outputs in both presidential and midterm election years. Program outputs herein include the number of uncounted ballots over time—these uncounted ballots are known as “residual votes”—the difference between turnout and the total number of votes cast for a particular office. Another election administration outcome is provisional votes cast and counted. Provisional votes are failsafe ballots; if a prospective voter is not on the voter registration list when she goes to vote, she may cast a provisional ballot which is segregated from other ballots. Election officials will check the voter’s eligibility after the election. Provisional votes have become especially important because those with adequate photo identification at the polls can only cast provisional votes in some states (including North Carolina), and they must bring an identification later.