Low turnout in local elections: when a small and unrepresentative electorate decides

Low turnout in local elections means that mayors, city councils, school boards, and local tax measures in the United States are often decided by a small share of eligible voters. Research on California cities finds that on-cycle November elections “generally double local voter turnout” and that moving to them produces an electorate that is “considerably more representative” by race, age, and partisanship. Election timing is the largest single factor, but it is not the only one: whether races are actually contested also matters.

In plain words: many people vote for president, fewer vote for governor, and far fewer vote for mayor or school board. That is partly because many local elections are held on different days from the big national elections. When few people vote, the ones who do are usually older and more settled than the community as a whole. Their choices then shape schools, roads, policing, and local taxes for everyone.

The problem of low local turnout, defined precisely

Low local turnout is easier to discuss with precise terms, because local elections vary enormously in form and timing.

Local election. An election for an office or measure of a city, county, school district, special district, or similar unit. See Jurisdictions for how layered districts determine what is on a person’s ballot.

On-cycle election. A local election held on the same day as a federal general election in November of an even-numbered year.

Off-cycle election. A local election held on any other date, such as in an odd-numbered year, in the spring, or at a special election.

Nonconcurrent election. California’s statutory term for an election “other than on a statewide election date”.⁠⁸

Consolidation. A rule requiring or allowing local elections to be held on the same dates as state or federal elections.

Turnout denominator. Local turnout is usually reported as a share of registered voters, while national turnout is often reported as a share of the voting-age or eligible population. The MIT Election Data and Science Lab notes that the voting-eligible population is the preferred denominator, followed by the citizen voting-age population and then the voting-age population.⁠⁵ Comparisons across these denominators must be read with care. See Measurement.

Contestation. Whether voters have a real choice: a race with one candidate, or an incumbent without a serious challenger, gives little reason to vote.

Scale and evidence of low turnout in local elections

Turnout in off-cycle elections, as described by the MIT Election Data and Science Lab, tends to be “much lower than elections held to coincide with federal elections,” and the demographics of those who do vote are more skewed relative to nonvoters.⁠⁵ Hajnal and Lewis observed in 2003 that while low turnout in national elections “has garnered considerable attention and concern,” the “much lower turnout in municipal elections has often been largely ignored”.⁠²

Statewide data from Texas show how large the timing effect can be among the same pool of registered voters. According to the Texas Secretary of State:⁠⁶

These odd-year elections are statewide, not local, and they carry no candidate races. They illustrate the general pattern: the same registered voters participate at a fraction of the rate when no major candidate is on the ballot and the date is off the federal cycle.

Research on local elections specifically confirms the pattern and adds nuance. Marschall and Lappie studied “more than 1,000 California mayoral elections” held between 1995 and 2014. They found that “on-cycle elections are associated with appreciably higher levels of voter turnout,” but also “considerable variation across space and time,” and that “contestation can reduce or even eliminate the differences in turnout” between schedules.⁠³ Hajnal, Kogan, and Markarian, combining election timing with detailed voter data for California cities, found that on-cycle November elections “generally double local voter turnout,” with the largest gains “when these local elections coincide with a presidential election”.⁠¹

National survey data on local turnout are thin. The Census Bureau’s Voting and Registration Supplement, the standard national source on who votes, is fielded after federal elections and asks about the latest national election.⁠¹⁰ No national survey reviewed for this page measures turnout in local elections held at other times. That data gap is itself part of the problem.

Causes and competing explanations

Low turnout in local elections has several identified causes, and researchers weigh them differently.

Election timing. Hajnal and Lewis found that “moving local elections to coincide with the dates of national elections” would “have by far the largest impact on voter turnout” among the institutional changes they studied.⁠² Later work confirms a roughly twofold effect.⁠¹

Stakes and institutions. The same study found that changes that “raise the stakes of local elections also increase turnout,” specifically “less outsourcing of city services, the use of direct democracy, and more control in the hands of elected rather than appointed officials”.⁠² When local officeholders control more, voting for them matters more.

Competition. Marschall and Lappie’s finding that contestation can erase the timing gap.⁠³ suggests that some low turnout reflects races with no real choice, not only inconvenient dates.

Information. Voters learn about national races from national media; local races depend on local coverage, which has contracted in many places. That mechanism is plausible and is treated on the Local news collapse page; this page does not attempt to measure it.

Habit and mobility. People who have lived in one place longer are heavily overrepresented among voters even in federal elections⁠¹⁰ In local elections, where the electorate is smaller, the same settled residents can make up an even larger share, though national data to confirm that are lacking.

Who is harmed by low local turnout

Low local turnout changes who decides, and the evidence on the consequences is more measured than the rhetoric around it.

Younger residents, minority residents, and partisan minorities. Hajnal, Kogan, and Markarian found that moving to on-cycle elections in California yields an electorate that is “considerably more representative” in race, age, and partisanship.⁠¹ Off-cycle electorates are therefore less representative on those dimensions.

Everyone affected by school and tax decisions. Kogan, Lavertu, and Peskowitz analyzed “more than 10,000 school tax referenda” and found that timing influences voter composition “in terms of partisanship, ideology, and the numerical strength of powerful interest groups.” Timing “has the most significant impact on voter age,” with “the elderly being the most overrepresented group in low-turnout special elections.” Residents with children in school may therefore be underrepresented in decisions about school funding.⁠⁴

Limits on the harm claim. The same study found that these compositional effects “are modest in terms of their likely impact on election outcomes,” and that the electoral and policy implications “vary between states”.⁠⁴ A less representative electorate does not always produce a different result. The harm is clearest as a matter of democratic equality and less consistently measurable as a matter of policy outcomes.

Local governments themselves. Officials elected by a small share of residents can face weaker claims to legitimacy, and separate election dates carry administrative costs. Arizona’s legislature, for example, cited both voter participation and costs to taxpayers in its finding on election dates.⁠⁹

What has been tried elsewhere to raise local turnout

The main reform tried for low local turnout is moving or consolidating election dates, either by state mandate or local choice.

State consolidation mandates. Arizona law declares it “a matter of statewide concern that all elections in this state be conducted on a limited number of days,” citing increased voter participation and reduced costs, and states that it overrides conflicting local charter provisions. Since 2014, candidate elections for political subdivisions, other than special and recall elections, may be held “only in even-numbered years,” and tax-related measures face similar date limits.⁠⁹

Turnout-triggered consolidation. California takes a conditional approach. A political subdivision “shall not hold an election other than on a statewide election date” if “holding an election on a nonconcurrent date has previously resulted in a significant decrease in voter turnout”.⁠⁸ A “significant decrease” means turnout in a regularly scheduled local election that “is at least 25 percent less than the average voter turnout within that political subdivision” for the previous four statewide general elections. The law applies to cities, school districts, community college districts, and other districts.⁠⁷ Subdivisions that had adopted a consolidation plan by January 1, 2018 could continue nonconcurrent elections until consolidating no later than the November 8, 2022 statewide general election.⁠⁸

Results. The research on California cities that moved on cycle shows roughly doubled turnout and a more representative electorate.⁠¹ Whether consolidation also changes which candidates win, or which policies pass, is less settled; in school tax referenda, the outcome effects of timing were modest.⁠⁴

Other institutional changes. Strengthening elected offices and using direct democracy are associated with higher turnout,⁠² though these are larger structural choices than a calendar change. For the history of local referendums and participatory budgeting, see Participation between elections and Comparative cases.

What has not been shown. The sources reviewed for this page do not measure how often voters on long consolidated ballots skip local races, or how much national partisanship spills into formerly separate local contests. Those are the main objections to consolidation, and the evidence on them is treated here as an open question rather than a settled fact.

The strongest arguments about local election timing across the political spectrum

Low local turnout is a rare problem on which arguments do not divide neatly by party. Each position is stated in its strongest form.

Consolidate for representation. Advocates of consolidation argue that a city should be governed by an electorate that looks like the city, and that doubling turnout and improving representativeness by race, age, and partisanship.⁠¹ is worth some ballot length. Many on the left emphasize the gains for younger and minority residents.

Consolidate to limit organized interests. Critics of off-cycle elections, including many on the right and some reformers of all leanings, argue that low-turnout elections hand disproportionate influence to groups that reliably mobilize, such as public employee organizations, developers, or homeowner associations. Kogan and colleagues confirm that timing changes “the numerical strength of powerful interest groups,” while finding the outcome effects modest⁠⁴

Keep local elections local. A localist argument, held by many conservatives and libertarians, holds that cities and districts should choose their own election dates, as home rule and charter government intend, and that state mandates like Arizona’s override of local charters,⁠⁹ centralize a decision that belongs to the community. A related argument holds that separate dates let voters judge local officials on local performance rather than on national party labels.

Turnout is not the only measure. Some argue that a smaller electorate of informed and interested residents may make better local choices than a larger one voting on a long ballot after choosing a president. Others reply that this standard would justify excluding any group judged less informed, which conflicts with equal citizenship. See Unequal participation for the general form of this debate.

Competition first. An institutionalist view, supported by the finding that contestation can erase the timing gap⁠³ holds that the deeper problem is uncontested local races. Remedies would include lowering barriers to candidacy and strengthening the stakes of local office,⁠² not only moving dates.

Unresolved questions about low local turnout

• Roll-off. How often do voters on consolidated ballots skip local contests, and does skipping cancel part of the turnout gain? The sources reviewed here do not measure it.
• Nationalization. Does consolidation bring national partisanship into local races that were previously decided on local issues?
• Outcomes. Composition changes are well documented;⁠¹ outcome changes appear modest in school tax referenda.⁠⁴ Are they larger for city offices, police, or land use?
• Special districts and school boards. The national Census survey asks only about the latest national election.⁠¹⁰ How can turnout in these elections be measured consistently?
• Legal authority. How far may a state impose election dates on home-rule cities, and how do state constitutions differ? This varies by state and is not settled by the statutes cited here.⁠^8,9
• Competition. If contestation can erase the timing gap,⁠³ which reforms produce more contested local races?
• Application to this framework. Could a local priority tool reach residents who skip off-cycle elections, or would it draw the same small group? See Evaluation and Questions.