Out of the spotlight of national politics, challengers stand a better chance.
Virginia’s election in November had major nationwide implications, thanks to an unexpected win by Governor-elect Ralph Northam after a contest that saw his challenger, Ed Gillespie, reach deep into the Trump playbook. But you wouldn’t know it mattered so much by the turnout. About 2.6 million people showed up to the polls for the November general elections—less than half of registered voters, and 1.4 million fewer than participated in the 2016 election.
Turnout was similarly low in another 2017 contest, in New Jersey, where just 39 percent of registered voters pulled the lever for Governor-elect Phil Murphy. But the low-turnout, off-cycle race might have worked to the advantage of Hoboken Mayor-elect Ravi Bhalla, who, despite enduring a racist flier campaign, persevered in an open race to become the first Sikh mayor in New Jersey’s history.
The same could be said of Virginia Delegate–elect Danica Roem’s election. In defeating the incumbent, conservative culture warrior Robert G. Marshall, Roem emerged as one of the first openly transgender officials in the nation’s history (and the author of a memorably savage epitaph for her opponent). She won the race by about 8 points, which amounted to just 1,800 votes.
Which is not to diminish either victor’s success. But it’s possible that Bhalla and Roem might have faced a steeper climb in 2016, when then-candidate Donald Trump dominated the ballot. As a rule, challengers fare far better in off-cycle elections than they do in big-ticket races.
That’s the conclusion of a new study on mayoral elections in The Journal of Politics: Incumbents hold a vital advantage in races that fall in the same year or month as national contests. In an examination of nearly 10,000 mayoral elections over the past 60 years, the study finds that incumbent mayors in on-cycle elections are nearly twice as likely to win as incumbents in off-cycle elections.
“The incumbency advantage is a hallmark of when there’s a problem of accountability,” says Justin de Benedictis-Kessner, a research associate at the Boston Area Research Initiative at Harvard University and Northeastern University. “If people are re-electing any kind of elected official at a higher rate just because they’re an incumbent, and not because they’re necessarily better at governing, that’s a problem for how we think of democracy.”
At a glance, the incumbents’ advantage in on-cycle elections is easy to explain. Presidential elections draw out many voters who may not care much about local politics. When they get to the end of the ballot, these voters may give the nod to the name they know—the cognitively safer option, the incumbent mayor. Voters who turn out for off-cycle elections are different. These may be highly motivated issue voters, whose galvanization sometimes leads to fears that off-cycle elections are vulnerable to capture by special interests or ballot referendums.
The fact that Hizzoner has a leg up on a no-name challenger is not surprising. But the dimension of this advantage is immense. Incumbency has a demonstrable effect on the likelihood that a candidate will run again for election and win, and election timing is a big factor within this probability. In cities with on-cycle elections, incumbents are 62 percent more likely to win re-election over challengers. In places with off-cycle elections, that likelihood is just 35 percent.
The incumbency boost may affect the quality of candidates who decide to enter into local politics, de Benedictis-Kessner says. On average, mayors win elections by 9.1 percent of the city’s population in on-cycle elections, whereas in off-cycle elections the margin is 4.7 percent of a city’s population on average. That means that candidates who get knocked down in on-cycle elections may decide not to try again, whereas challengers in off-cycle elections see a smaller deficit to overcome next time.
In terms of off-year elections, the worst-case scenario is when an incredibly close, open-race election is followed immediately by an election in which there is no challenger.
“Not only do they barely win an election in one year, they then go on to enjoy the huge benefit of incumbency, where challengers are either scared off or don’t get nominated to oppose them, because they’re an incumbent,” de Benedictis-Kessner says.
The study also looks at cities that chose to switch up election years (1,092 elections in 76 cities). Oceanside, California, for example, decided in the 1980s to move its elections from the spring of even-numbered years up the calendar to November, bringing them in line with national elections. “Among the cities that switch election timing, on-cycle elections give incumbents an advantage of approximately 52 percentage points, while off-cycle elections have a significantly lower advantage of only 25 percentage points,” the study reads—a large and statistically significant difference.
“It’s not necessarily bad if a really high-quality incumbent keeps getting re-elected,” de Benedictis-Kessner says. “But it’s not great if a person who isn’t really distinguishable from the challengers on the quality dimension is re-elected at a higher rate than you would expect.”
Building on a dataset for a prior project on partisanship in local elections, de Benedictis-Kessner assembled data on 9,131 mayoral elections, sorted by month and year, and featuring almost 10,000 unique candidates. The elections dataset spans 1,016 cities between 1950 and 2014, ranging from Andover, Minnesota, to New York City in size.
The research doesn’t attempt to factor in the quality of an election or administration, which would be an impossible task anyway. Rather, it focuses on close contests: By comparing candidates who barely won an election with those who barely lost, the study assumes the candidates are nearly identical in terms of quality, in order to test the incumbency advantage in subsequent elections. In an appendix, de Benedictis-Kessner accounts for other factors such as mayoral term limits and strong mayor vs. weak mayor forms of government.
“Election timing is really the only electoral institution that has an effect on the degree of incumbency advantage,” he says.
So should cities switch from on-years to off-years in hopes of embracing more competitive mayoral races? Not necessarily: Sarah Anzia at the University of California-Berkeley has shown how election timing can lead to wide disparities in policy outcomes based on organized special interests (in this case, with regard to teacher salaries). Generally speaking, it’s not in keeping with the highest ideals of democracy to say that elections with lower turnout are in some sense superior.
Yet for local contests especially, election timing has consequences.
“It depends on which representational problems you’re willing to deal with,” says de Benedictis-Kessner.