The American Experimenter

There’s an old saying about diligent reporting in journalism: “If your mother tells you she loves you, check it out.”

Although he’s not a journalist, Anthony Fowler seems to live by that rule as well. Three years into his appointment as an assistant professor at Chicago Harris, Fowler has made a name for himself by asking questions that seem as though they don’t need to be asked—and then publishing findings that turn the conventional wisdom on its head.

Do voter participation campaigns really improve election outcomes? Not if the voters they mobilize are the same types of people who already vote. Do incumbents gain an unfair advantage from name recognition and better fundraising opportunities? Not if they were elected in the first place because they’re actually good candidates.

Such fundamental research questions—and their unexpectedly complex answers—are helping Fowler to reframe assumptions about the inner workings of democracy, and drumming up enthusiasm in the process. Fowler was recently named to a list of top 30 thinkers under age 30 by Pacific Standard magazine. “He’s asking questions of first order of importance in our understanding of elections,” says William Howell, Sydney Stein Professor in American Politics at Harris. “He’s one to watch.”

But accolades and academic citations motivate Fowler much less than the stakes of his work. “Are we missing questions because we think we already know the answers?” he wonders. "Have assumptions about things like get-out-the-vote efforts actually diminished representation? If that's the case, we'd like to know about it, and we'd like to develop new strategies for fixing the problem."

“I do fundamentally care about what kind of democracy we live in,” says Fowler.

EXPERIMENTAL TECHNIQUE

Back in high school, Fowler took a tour of the Salk Institute for Biological Studies in La Jolla, California. The son of two restaurant employees who worked long, hard hours, he was captivated by the idea that he could work in a lab, answering fundamental questions about life and science. Fowler would work at the Salk Institute during high school and while earning a bachelor’s degree in biology from the Massachusetts Institute of Technology with the intention of becoming a lab scientist. But while at MIT, he also took several political science and economics courses, which opened his eyes to a different path.

While biology offered the opportunity to answer fundamental questions, the methods of obtaining answers were almost always the same: conducting tests in petri dishes. Political science, however, was trickier. Each new question would require a new method, and a scrappy approach to evidence.

“You’ve got to go out in the world, collect data and come up with clever ways to answer difficult questions,” Fowler says. “For instance, what really makes people vote? You can think of that essentially as a harder problem, but you can also look for a clever way to set up an experiment to see what happened.”

“These questions—whether incumbents have unfair advantages, and how power is distributed in Congress—are really big questions that affect the lives of people in really big ways,” he adds. “We should be studying those questions with the same vigor that we use to study medical questions.”

It’s a high standard, and one to which Fowler holds himself. In fact, his diligent research design has been a significant factor in the attention he’s received. Instead of merely pointing out correlations between phenomena, Fowler seeks to answer complex questions of causation.

That process played out in Fowler’s dissertation work, in which he found that the introduction of compulsory voting in Australia led to tangible increases in pension spending. Plenty of political scientists had tried to quantify the impact of mandatory voting in the past. But only after painstakingly sorting through obscure data sets could Fowler show that elections prior to compulsory voting had not accurately represented the preferences of the whole public. As he concluded in the study: “This analysis brings us closer than ever before to answering the extreme counterfactual question: ‘what if everyone voted?’’’

“These are really hard questions that no one in political science made a lot of headway with for a long time because it’s been difficult to make causal connections with the data,” says Ethan Bueno de Mesquita, Sydney Stein, Jr. Professor in American Politics and deputy dean for research and strategic initiatives at Harris. “There’s this new generation over the last 10 or 15 years who are really interested in trying to understand what is going on in a causal way. Anthony is one of two or three people of that generation who stand out because of this type of work.”

A MORE SCIENTIFIC METHOD

At a time when questionable scientific findings run rampant, a cautious outlook is noteworthy. Fowler cringes when he sees quirky studies find their way into the mass media, promising, for example, that the outcome of a football game can predict an election. Those studies are great for headlines and offer simplicity in the face of difficult questions, Fowler acknowledges, but they don’t make for great science.

He suspects that many researchers are falling victim to something called publication bias, in which negative results are rarely published, but positive results are over-reported, even when they are spurious. “The way that we do our testing, even if there is no effect, five percent of the time we might find one. We might get a false positive,” Fowler says. “That’s something that I think about a lot. I’ve been trying to come up for ways to see if we can correct publication bias.” Specifically, Fowler hopes to help the scientific community see the benefits of publishing even negative results, and to challenge journals and scientists to dig deeper if positive results seem too good to be true.

At the end of his career, Fowler says, he hopes that he can look back and see that he improved not only the knowledge in his field, but also the reliability of published research. And he believes that being at Harris sets him up nicely to accomplish those goals.

“If I wasn't here at Harris, I’d be in a political science department, most likely. There might be one or two other people who do the kind of work that I do. But Harris is this unusual place where there are a dozen of us who do this kind of work,” Fowler says. “The kinds of political scientists we have at Harris are much more careful, much more rigorous, and care a lot about research design.”

“Being surrounded by the caliber of colleagues and with the resources available at Harris, I'm excited about the opportunities I'll have to tackle difficult questions that have profound implications for the political process."