Political Participation and the Cost of Abstention

February 3, 2015

This feature is part of an ongoing RSF blog series, Work in Progress, which highlights some of the ongoing research of our current class of Visiting Scholars.

During her time in residence at the Russell Sage Foundation, Susan Stokes (Yale) is writing a book investigating why people choose to participate in elections and demonstrations. She argues that the cost of abstention—or how much a person feels he or she will lose by not voting—can explain why people turn out at higher rates when the office to be filled is elevated. Stokes is also exploring how the cost of abstention may shed new insight on why low-income populations vote at lower rates than more affluent populations.

In a new interview with the Foundation, Stokes discussed how this theory complements prior models of political participation. She also detailed how the cost of abstention can help us understand what drives people to vote and to take part in political demonstrations—including, surprisingly, when violent police repression occurs.

Q. Much of the research on political participation has focused on the cost of participation. By contrast, your current research offers a model of the cost of abstention, which includes factors like the guilt or discomfort that may result from not voting. How does this complement the classic Riker and Ordeshook theory of the calculus of voting? Can we think of the costs of abstention as being similar to feeling a sense of duty when it comes to political participation?

Voting in elections and participating in social movements are both about taking part in an action that’s aimed at the collective, public good. But what drives us to participate? People by and large know that their vote for a given candidate is not going to make a difference between that person winning or losing. Furthermore, if your preferred candidate wins, you get the benefit of that win whether you participated or not—which seems like an incentive to not take part. There are costs of participation: for example, it can be a hassle to get to the polls on Election Day. The stakes are even higher when it comes to demonstrations or protests, which may seem intimidating or dangerous.

The question of why people participate at all in these political actions raises questions that social scientists have tried to deal with for a long time. The basic approach that a lot of theorizing on voting took was shaped by an influential article that came out in 1968 by Peter Ordeshook and William Riker. They set up a model of an individual’s calculus of whether to participate, which takes the form of R = (B*P) – C (where R = the reward you get from voting, B = the benefit you get if your candidate wins, P = the probability that your vote is the one that decides the election, and C = the costs of participation.)

However, once the electorate is 100 people, the probability that the outcome of the election will be a tie is so close to zero that it’s not really worth talking about as anything other than zero. So then, according to the Riker and Ordeshook equation, any costs of participation (C) are going to make the left hand side of the equation negative, and therefore nobody should turn out to vote. But since millions upon millions of people do vote, how do we account for that? In an attempt to factor in the private benefits of voting or political participation, Riker and Ordeshook added the term D (or “duty”) to the equation as follows: R = (B*P) – C + D. What they meant by duty was the feeling that we have an obligation to turn out and vote because democracy rests on participation. In the context of protests, the political scientist Dennis Chong has noted that this sense of duty can mean showing up to protest because you don’t want to disappoint your friends, or you want to save face—in other words, private benefits that have nothing to do with the movement itself.

Those are very reasonable ways of solving this paradox of participation, but they don’t quite settle the matter. A lot of researchers have questioned whether an abstract sense of duty is really what drives people to participate in elections and protests. Some have posited that people are driven more by partisanship. And for us, the idea of “duty” is limited insofar as it feels unconditional. That is to say, if we really feel an abstract obligation to vote, then we’d be voting for all kinds of offices, from dogcatcher to sheriff to president. But in reality, turnout tends to be higher when the the level of office being filled is higher. This is true both in the US and in other countries. So why is it that a sense of duty would kick in more strongly when the election is more important?

So how do we reconnect people’s decisions to vote with the importance that they ascribe to the outcome? A small, simple tweak to the Riker and Ordeshook model is to add A, or the costs of abstention. Costs of abstention can be psychological, moral, or social—in essence, they’re things that make you feel bad when you anticipate not voting. Our claim is that the desire to not feel bad actually drives people to vote and to participate in demonstrations.

Q. How does one track and measure the costs of abstention?

We borrowed a few tools from psychology to measure the costs of abstention. Psychologists have developed a technique for measuring people’s moods called the Positive and Negative Affect Schedule (PANAS) scale, which involves asking people a variety of questions about their moods in a given time frame.

In our study, we asked our participants these questions at the beginning of the survey to gauge how they were feeling. Then we took them through an experiment where we told them there was an election coming up. Through random selection, some people were told it was an important election and others are told that it was not. Then some people were told that they had voted, and others were told that they were prevented from doing so by being called out of town at the last minute. So there were four treatments: the people who had voted in an important election, the people who hadn’t voted in an important election; the people who had voted in an unimportant election, and the people who hadn’t voted in an unimportant election. After this, we surveyed them about their moods. And what we found is that the people who were told they abstained from voting in an important election showed deterioration in their subjective moods—in other words, they felt worse than they had before the experiment. Conversely, the people who were told they had voted in an important election felt happier afterwards.

We know that these changes in mood weren’t caused simply by telling people they had been called out of town, because the people who were told that they abstained in an unimportant election didn’t show the same deterioration in mood. By the same token, the people who were told they had voted in an unimportant election didn’t display the same uplift as their counterparts who had voted in an important election. So we think these studies show that there exist costs of abstention—and if people anticipate feeling rotten about not taking part in an election, they’re likely to go and vote.

Q. You’ve found that police repression of protestors can paradoxically spawn even more protests. In what ways is this driven by increased cost of abstention? In other words, why do more people sometimes join demonstrations where there appears to be a greater risk of arrest or bodily harm?

In certain cases, when the police use excessive force in trying to quell a protest, this act of repression can, in fact, cause the protest to grow in size. We looked closely at a few examples where governments had tried to suppress protests, which, as a result, not only grew in size, but became national uprisings. Through surveys and some experimental work, we’ve confirmed that there are people who are more likely to participate when they become aware of police repression of protests. These people feel empathetic anger when they see pictures of police violently suppressing protestors, which makes them more likely to then join the protests.

So far we only have detailed data for three cases, and we’re in the midst of building this theory. We’ve looked at the Ukrainian Euromaidan protests of 2013-14, the Gezi Park uprising in Turkey in 2013, and the Brazilian protests of June 2013. One of the unsurprising things we’ve found is that young people are more sensitive to costs of abstention—in a nutshell, they’re more likely to be triggered to take part in protests when the police begin to use force. We don’t yet know exactly why that is right now, but it’d be interesting to investigate further. Our studies also found that women, in general, seemed to be more likely to turn out and participate as a response to repression of protests, when they are aware of people being repressed. I’m hesitant to fall into stereotypes, but it could be that women for some reason feel more of a tug of empathetic anger than men, on average.


RSF: The Russell Sage Foundation Journal of the Social Sciences is a peer-reviewed, open-access journal of original empirical research articles by both established and emerging scholars.


The Russell Sage Foundation offers grants and positions in our Visiting Scholars program for research.


Join our mailing list for email updates.