An Interview with Anthony Fowler: Analyzing the Data of Incumbent Success

Credit to Harris School of Public Policy

Here at BallotReady, we have been trying to understand the phenomenon of uncontested races.

Why do so many races only have one candidate in them — the incumbent? What obstacles prevent newcomers from running for office? Over the past few months, we’ve interviewed people who knew exactly how hard it was to run for office against a long-time incumbent: Illinois State Rep. Will Guzzardi and Minnesota State Rep. Fue Lee.

Now, in this post we showcase the perspective of someone who has studied the topic in depth, utilizing data from across the country. Anthony Fowler is an assistant professor at the Harris School of Public Policy at the University of Chicago. His research seeks to apply methods from econometrics to political science questions, including addressing problems of voter turnout and analyzing the causes of incumbent success. We sat down with him to get his thoughts on electoral competition, the unpopularity of Congress, and the importance of holding elected officials accountable.

Why are uncontested races so common?
My biggest guess would be, most uncontested races are in extremely partisan districts. Big cities that vote 80% Democrat or more and no Republican bothers running because they don’t think they have a shot. You see lots of cases of uncontested incumbents who are probably bound to win anyway, even if they had a serious challenger, so there might be a lot of qualified candidates who don’t bother running if they don’t have a shot.

It’s worth talking about the primaries a little bit, because in a lot of places, where there’s an uncontested general election, that might mean there’s an extra-competitive primary. In fact, that was the story of the one-party South for many decades, where there were almost no viable Republican candidates. Basically every general election was uncontested or virtually uncontested, but all the action was in the Democratic primary. You could argue that a lot of the benefits of competition and accountability and selection came from the primary where that was really where the race was, and so maybe it’s not such a bad thing that you get these uncontested generals when there’s a competitive primary to hold the incumbents accountable.

Is it a bad thing that races are uncontested or is it fairly normal?
It’s not obviously bad. I certainly think that electoral competition is good. One reason electoral competition is good is that when voters have a choice to make, they’re more likely to pick the better candidate and there’s more accountability. You might imagine that incumbents might work harder if they have a serious challenger and there’s some empirical work on this kind of stuff. I have some work showing that a lot of incumbent success in elections actually comes from positive selection, so it’s not just that there’s some nasty incumbency advantage where, any mediocre candidate, you barely win, and all of a sudden you’re winning by a lot; most of incumbent success is explained by the fact that elections do a decent job of selecting good candidates and if you don’t even have a competitive election to select, then you’re losing that ability.

Similarly, there’s a lot of evidence on the accountability side, that when there’s a newspaper that really covers what one member of Congress is doing, that member of Congress works harder, shows up to more committee hearings, votes more in line with the district, so you might think that competitive elections serve the same kind of purpose of, “If I’m not even contested, why bother trying hard, why bother voting in line with the district.” And there’s some nice evidence showing that governors produce better economic outcomes when they have an electoral incentive; when they’re term-limited the economic outcomes are a little bit worse in the state.

So, all of that suggests that electoral competition is a good thing, but the reason I don’t want to go so far as to say it’s a bad thing that we have uncontested races is that it could be [that] the reason why we have so many uncontested races is that in equilibrium, candidates are working hard and doing a good job, and the reason there’s no challenger is because the challenger doesn’t think they can win. But if the incumbent did slack off, the challenger would think they would win and they would run. So it could be, we’re in a really good equilibrium where we’re selecting reasonably good candidates who work hard and do a decent job, and therefore they’re not opposed because they’re virtually guaranteed to win because they’re doing a good job. In a world where they weren’t doing a good job, maybe they would be contested more.

How would you try to figure out if uncontested races are a result of a good competitive equilibrium or the more conventional worries about incumbency and institutional advantages?
It might make sense to think more about specific policies or specific reforms: what would a policy be, or a lever that somebody would pull that would create more contested elections? Would that be good or bad for political representation as a whole? So you’d want to think about that a little bit more. What lever did we pull to get more contested elections?

One lever would be: what if we redistricted such that there were less partisan districts? There have been a decent number of studies on this. The natural way that Americans tend to cluster themselves is that cities tend to be very Democratic and the suburban and rural areas tend to lean Republican but they’re not that Republican, maybe they’re 55% Republican. The way that works is that you get a lot of leaning-Republican but somewhat competitive districts out of the suburbs and rural areas and then you get a lot of really uncompetitive districts in cities. Given that’s the way people live, it’d be hard to redistrict to dramatically change that, but you could imagine taking pie cuts out of the cities, so instead of just a bunch of Chicago districts and then a bunch of farmland districts, you could have pie cuts that have a little slice of Chicago going all the way out to the farmland, and you could have a lot of competitive districts in Illinois if you wanted to. That would change lots of things. Would that be good or bad? I don’t know.

One thing that would be worse is, certain underrepresented groups like African-Americans would have less descriptive representation so there’s a trade-off there. There’s a big debate already on whether majority-minority districts are really good for minorities or not. On the one hand, they might increase the chances that those minority groups have a really strong representative of their own in Washington or in the state legislature. On the other hand, maybe they’re actually shifting the median legislator further away from their preferences. They’re getting this one guy who’s really close to their preferences but had they instead spread all of those liberal interests across many districts, maybe the median legislator would actually be closer to their preferences, so there’s all kinds of trade-offs there and it’s hard to think about all of them at once. So that’s just one thing and we could go down lots of different paths.

You can imagine public funding for elections, public financing might make it easier for a challenger to run. Right now it’s pretty costly to just come out of the woodwork and run your own campaign and maybe you’d get a lot less uncontested races, especially at the state legislative level or city council level, if you get a certain number of signatures and you get $20,000 funding and you can actually run a campaign. There are roughly five states that have pretty aggressive public funding systems for state elections: Minnesota, Wisconsin, Connecticut, Maine, and Arizona. Those states, they all have slightly different systems, but it’s if you can get enough signatures, then even a challenger can have just as much funding as an incumbent in lots of cases. So a lot of people have look at the effects of that kind of system. One thing you do find is the estimated incumbency advantage shrinks a little bit. It also looks like more of the [incumbency advantage] comes from selection in these cases where they have public financing, so voters actually have more of an opportunity to learn about both the incumbent and the challenger, maybe they do a better job selecting, but there are probably other downsides you want to think about as well.

Maybe one of the ways that elections do a good job is, we queue off donors. [Public funding] takes away that channel, removes the possibility that I can see “oh that guy’s raising a lot of money from this group, therefore he must be the kind of candidate that I like.” So you have to think of lots of pros and cons.

Is there a risk that when the incumbent has been in office for too long and uncontested for too long, that people zone out and don’t necessarily know what the incumbent is doing but they continue to vote for them?
Certainly you can imagine that kind of risk. You become entrenched and you get so many institutional advantages. But you would think, if you believe in the voters a little bit and give them a little bit of credit, there’s some reason why they keep re-electing this guy, and there’s some reason why no entrepreneurial challenger has come forward. You would think that if some high-quality challenger could come forward and win that election, they would. So there’s some reason they don’t, and it must be because either that incumbent is doing a pretty good job representing the district or that incumbent is bringing a lot of federal funding home to the district or something.

You could just descriptively imagine doing some kind of case study to see, in these cases where you get these really uncontested elections, does it look qualitatively like they’re slacking off on the job or does it look like, “Oh OK, this guy is bringing home lots of pork and he’s showing up and doing a good job.” It’d be hard to tell, but you could imagine selecting some small number of cases and comparing: does the guy seem to be doing a better job or working harder when he’s contested or when he’s not contested? I suspect it’s gotta be a little bit of the good story, because if they were doing such a terrible job representing the district and not showing up and slacking off, then some challenger would come forward either in the primary or the general and beat them.

One local example that lots of people complain about is Michael Madigan, who is this really entrenched Speaker of the House in Illinois. A lot of people think he’s really terrible for the state because he’s been there forever, he’s basically like a dictator in the state legislature, but his district keeps re-electing him, and he’s had some people really try hard to challenge him in primaries and in generals, but the district keeps reelecting him and it’s mostly because the guy’s actually powerful and he’s good for the district. You’ve got this tricky situation where, even if I don’t really like the disproportionate influence this guy has, and even if I don’t like the fact that he’s entrenched and set up these political institutions that I don’t like, conditional on that system being there, I like my guy being on top and being powerful and bringing home lots of stuff.

One of the common puzzles that people talk about is the fact that Congress universally is unpopular and you ask people, do you like the job Congress is doing? Do you like the way Congress is set up? And everyone says no. But everyone seems to like their own member of Congress. Some people just interpret that as a sign that people are irrational and they are like, “Well, my guy is better than everyone else!” But another interpretation is, they’re being totally reasonable and sophisticated, which is, “I don’t like the current system, I don’t like the incentives, I don’t like the way everything works, but conditional on the way everything is, I’d rather have my guy who is experienced and knows how it works and knows how to exploit it.”

That’s related to the answer you’re going after: why do you reelect all these people? It goes back to, what is another lever you could pull? What if you changed the nature of institutions such that there was less of a seniority system in Congress? What if committee chairs had less power and so forth? That would create more of an incentive for a new person to run and maybe that would be better, maybe that would be worse.

In our next post of the Uncontested Races series, we will be interviewing Jim Cupples, the founder of is an online database with information about how to run for every local office in your area, from filing deadlines to candidacy requirements. Make sure to check it out!

BallotReady is the only nonpartisan online voter guide that provides personalized, easy-to-use, and accessible information about your ballot. From the top of the ticket to the very bottom, BallotReady helps voters compare candidates based on stances on issues, biography, and endorsements. We make it easy to vote informed on every race and referendum.

By Eileen Li, BallotReady Intern