Local Democracy in America
How Access, Competition, and Place Shape Turnout in Mayoral Races
Melissa Marschall (Rice University) & John Lappie (Plymouth State University)
Judging by rates of turnout and contestation in mayoral races across the roughly 2,500 municipalities examined in this study, it seems safe to conclude that there is room for improvement when it comes to the health of local elections in the United States. While turnout varies considerably, across the more than 10,000 mayoral races we investigated, it averaged only 43 percent, and almost half of all races (48%) were unopposed. Though these statistics seem rather pessimistic, our study confirms prior research, which finds that the most powerful predictor of turnout in mayoral elections is when the election is held (Hajnal & Lewis 2003; Marschall & Lappie 2018). We find that irrespective of size or location, municipalities with off-cycle elections have turnout rates about 25 percentage-points lower than municipalities whose mayoral elections are concurrent with presidential elections. The good news then, is that a very promising means by which to significantly increase turnout in local elections is not only quite simple, but also clearly within the province of policymakers: changing the election date.
Our study of turnout in local elections also shed important new empirical light on the effects of city size. With the large and more diverse sample of localities employed in this study, we found significant, though non-linear effects of city size on turnout in mayoral races. Compared to the smallest towns, medium-sized cities have significantly lower turnout, ranging from about 3.5 to 8 percent-points on average. In contrast, we found no difference in turnout between the biggest cities and these small towns.
Beyond the effects of election timing and city size, this study broke new ground as well. In particular, we looked at how election laws, namely partisan elections and state laws seeking to restrict voting access, shape turnout in mayoral elections. While we uncovered evidence that both matter for local turnout, effect sizes were attenuated with the introduction of state dummy variables. However, the restrictiveness of state voting laws had a consistent and statistically significant negative effect on turnout. Municipalities in states that scored highest on the index of state voting restrictions saw significantly lower turnout rates (about 2.5 percentage-points) than those in states with the least restrictive laws. When we looked at the component parts of this index separately, our findings suggest that all three laws- those prohibiting early voting, requiring government-issued photo IDs to vote, and disenfranchising felons- operate independently on local turnout. This preliminary evidence indicates that what critics have labeled “Jim Crow 2.0” laws are associated with significantly lower turnout at the local level, suggesting that further exploration and analysis are warranted. Indeed, the number and type of other restrictive voting laws is on the rise (Brennan Center for Justice 2021), and we do not know what, if any, additional impact these laws have on local turnout. It is possible that restrictive laws operate even more strongly on turnout in off-cycle elections or in municipalities with larger shares of low-income or minority voters. After all, in presidential and midterm years parties, candidates, and civic-minded NGOs exert a great deal of effort telling voters what forms of voter ID are acceptable, how to cast provisional ballots, when registration and early voting deadlines are, etc. These things are less likely to occur in off-cycle elections for local office, where campaigns are typically much more limited. In addition, since some laws disproportionately affect low-income and minority voters, who already face greater barriers, there is good reason to believe that the effects of these ‘Jim Crow 2.0 laws’ may be felt more strongly in municipalities where these voters are more concentrated. Future studies could fruitfully explore these and other questions.
Finally, while existing research has examined the relationship between competition and turnout, typically finding that more competition leads to higher turnout, few studies have focused on the other extreme: the lack of any competition at all. In this study we took a closer look at how unopposed mayoral races affect turnout. Our results show that these races are associated with significantly lower turnout than those with two or more candidates. In fact, turnout in these races was on average nine percentage-points lower than turnout in contested mayoral races. This effect is large enough to nearly wipe out the boost in turnout associated with holding mayoral elections concurrently with midterm elections rather than off-cycle. Our study also found that uncontested elections are not uncommon. Across this large and diverse sample of mayoral races, nearly half were unopposed. The role of contestation and the impact of unopposed elections on local turnout are substantial and deserve more scholarly attention.
In sum, our study makes several important contributions to the literature on local elections. Having access to a dataset that included a representative sample of municipalities in the U.S. provided greater variance in many key variables, including election timing, partisan elections, city size, and importantly, contestation, thereby enabling us test for how both traditional and more novel factors shape turnout in local elections. At the same time the data allowed for some important descriptive insights about the state of local elections in the U.S. For example, while the vast majority of Americans live in medium- and small-sized cities, these are the places prior research has studied least. Our data revealed that it is actually in America’s smallest towns, which make up 84 percent of all municipalities in the U.S., where turnout in mayoral elections is highest. At the same time, uncontested elections are most common in these small towns. Fully 56 percent of mayoral candidates in small towns ran unopposed. This compares to only eight percent of mayoral races in large cities and 22 percent of mayoral races in medium-sized cities. Importantly, our study also found that independent of city size and a host of other variables controlling for socio-demographic and institutional features, there are significant differences in turnout across rural and suburban municipalities. While the broader scope of the present study allowed us to take a closer look at the state of local elections in the places where most American live, clearly even more work is needed to better understand how electoral access, competition, and place shape turnout across the diverse, local landscape of the U.S.
Melissa Marschall is professor of political science at Rice University in Houston, TX. Her research and teaching focus on local elections, urban politics, educational policy, political behavior, and representation, mostly in the U.S., but also in Turkey.
John Lappie is associate professor of practice at Plymouth State University in Plymouth, NH. His research and teaching interests include United States political institutions, campaigns, and elections, especially at the local level.