What if You Could Vote for President Like You Rate Uber Drivers?

After a recent ride in Washington, DC, the Uber app promptly asked us to rate our driver. For the eight-minute journey from Capitol Hill to Chinatown, we could rate the driver’s service, conversation, navigation, entertainment, coolness, choice of music, amenities, tidiness, and select compliments such as “above and beyond” and “late night hero.” Compare that with how much feedback we are able to share in more consequential scenarios, such as electing leaders for our local ward to Congress and the White House. In those cases, we have almost no ability to fully express ourselves.

Now imagine if we could gauge candidates for political office across as many different dimensions as our Uber drivers—rating attributes like their trustworthiness, effectiveness, policy experience, accomplishments, professional conduct, team work, civility, and authenticity. Could this change to our ballot-box choice architecture increase voter turnout and produce election outcomes that more closely reflect the constituent perspectives? Might people choose a candidate based on factors more relevant to the role, akin to how we select job applicants, automobiles, or apartments? We believe so, and preliminary research suggests that more expressive voting may increase voters’ enthusiasm for participation.

Standard remedies to improve turnout—simplifying registration, keeping polls open longer, and allowing absentee, early, and mail voting—focus on making it easier but not on making it more desirable to vote. Since World War II, voter participation in the United States peaked in 1960 at about 63 percent, slowly drifting down to around 55 percent in recent elections. This decline in participation is despite the new technologies and approaches designed to reduce costs and encourage voting. While removing artificial barriers to voting is important—particularly those that fall heavily on subsets of eligible voters—we believe that low voter turnout also depends on the perceived benefits from voting.

Standard remedies to improve turnout focus on making it easier but not on making it more desirable to vote.

A good example of a time when that perceived benefit drops is during the U.S. midterm elections. Since the presidency is not at stake, the midterms are perhaps considered not as consequential or interesting. But voting at all levels and all times can be made much more engaging and valuable. Reinforcing this need for more expressivity, a nonvoter in the recent U.S. midterms offered to the Washington Post a blunt reason for not voting: “Because my simple vote cannot adequately express the rage, fury and contempt I feel for those already in government, as well as those who seek to replace them.” Developing a voting system that’s more expressive for all could help restore true civic engagement.

Some experimental studies have begun to offer insights into the benefits of making voting methods—and the very goals of voting—more expressive. In the 2007 French presidential election, for instance, people were offered the chance to participate in an experimental ballot that allowed them to use letter grades to evaluate the candidates just as professors evaluate students. This approach, called the “majority judgment,” provides a clear method to combine those grades into rankings or a final winner. But instead of merely selecting a winner, majority judgment conveys—with a greater degree of expressivity—the voters’ evaluations of their choices. In this experiment, people completed their ballots in about a minute, thus allaying potential concerns that a letter grading system was too complicated to use. What’s more, they seemed more enthusiastic about this method. Scholars Michel Balinski and Rida Laraki, who led this study, point out: “Indeed, one of the most effective arguments for persuading reluctant voters to participate was that the majority judgment allows fuller expression of opinion.”

Additional experiments with more expressive ballots have now been repeated across different countries and elections. According to a 2018 summary of these experiments by social choice theorist Annick Laruelle,  “While ranking all candidates appears to be difficult … participants enjoy the possibility of choosing a grade for each candidate … [and] ballots with three grades are preferred to those … with two grades.” Some participant comments are revealing, stating, “With this ballot we can at last vote with the heart,” or, “Voting with this ballot is a relief.” Voters, according to Laruelle, “Enjoyed the option of voting in favor of several candidates and were especially satisfied of being offered the opportunity to vote against candidates.”

Developing a voting system that’s more expressive for all could help restore true civic engagement.

Though this research is intriguing, only through direct comparisons of different voting methods can we clearly understand their benefits and deficits, and which of those methods can desirably increase trust, confidence, and participation among voters. Consider this 2016 analysis by Balinski and Laraki based on survey from a Pew Research Center of voters to imagine what the results might have been if the U.S. had possibly used majority judgment instead of the standard vote for one process. In this survey, voters were asked to grade the major candidates—Hillary Clinton, Ted Cruz, John Kasich, Bernie Sanders, and Donald Trump—on a scale using specific terms “great,” “good,” “average,” “poor,” “terrible,” and “never heard of,” rather than merely choosing one single favorite. A rank order based on median grades showed Kasich, Sanders, Cruz, Clinton receiving an “average,” and Trump rated “poor.”

Refined methods to break ties for those with the same grade would have pitted Kasich against Sanders in the final election, with Clinton and Trump ranking last in their respective parties. Instead, primary election results and subsequent political conventions led to Clinton and Trump as the major party nominees, with the latter eventually winning the electoral college and the presidency. Thus, the final election gave U.S. voters the option of choosing between the two poorest-graded candidates in this analysis. The two finalists were selected with under 17 million votes each, out of a total of 60 million who participated in the primary elections, who in turn represented only one quarter of the 245 million voter-eligible population. Each finalist was thus favored by about 7 percent of the total voter-eligible population. The low participation in primary elections (less than half of general election turnout) provides yet another indication of how the perceived value of voting is more important than the costs of voting in people’s decisions to vote. The cost of voting is similar for both primary and general elections, but the turnouts are vastly lower in the primaries.

No voting system is perfect, but some offer far richer vocabulary than others. As a specific example, consider a primary election with six candidates. Vote-for-one provides only six possible expressions. Approval voting—selecting all candidates one endorses—offers 64 different expressions. Rank ordering six candidates provides 720 different possible expressions. Letter grading the candidates using A through F provides 46,656 possible expressions. Adding plus or minus modifiers to these letters expands this to over 34 million. Cumulative voting—spreading 20 points across candidate choices as used in corporate board elections—offers 53,130 expressions. Finally, numerically grading each candidate on a 1 to 20–point scale—commonly used when scoring wines—offers about 64 million expressions.

While popular opinion polls can help carve the political agendas, the voting process itself could provide a much richer and more reliable source of information on constituent opinions, if only the process would enable us to fully express our preferences.

Beyond influencing voter turnout and information transfer, moving to a ballot system with more expressive features could have other spillover benefits. Consider, for instance, that voting can be an essential medium for information transfer from constituents to decision makers, just as the Uber app relays feedback between drivers and passengers. While popular opinion polls—usually with the limitations of sample size and potential participation bias—can help carve the political agendas, the voting process itself could provide a much richer and more reliable source of information on constituent opinions, if only the process would enable us to fully express our preferences. Yes/no or vote-for-one binary choices have the least possible expressive content. Take the Brexit vote as a recent example, where citizens could choose only to stay or leave. An initial description of a possible range of options beyond this binary, and a method for people to rate them, might have provided a more specific roadmap for British leaders to decide what to do, and if they wished to exit the E.U., on what terms.

These opportunities for expression might increase public interest in (and engagement with) democratic decision making, encouraging more thoughtful candidate debates, more substantive election campaigns and advertisements, and richer use of opinion polling to help candidates shape their position statements (once they are aware that the public’s selection process has changed). One could even envision that the basis for funding election campaigns might evolve if funders focused on policy ideas rather than political allegiances and specific candidates. Changes such as these would ideally put the power back in the hands of the people, where it actually belongs in a democracy. These conjectures need to be tested and retested across contexts, ideally through field experiments that leverage research and expertise in engineering, social choice, and political and behavioral sciences.

Standard left-to-right political scales and the way we currently vote do not capture the true complexity of our evolving political identities and preferences. If voting is indeed the true instrument of democracy and much more than a repeated political ritual, it must allow for richer expression. Current methods seem to discourage public participation, the very nucleus of civic life. The essence of civility and democracy is not merely about providing issues and options to vote on but in enabling people to fully express their preferences. For a country founded on choice as its tenet, is it too much to ask for a little bit more choice in how we select our leaders?