pivotality and performativity

OrgTheory is at it again with their “I’m really not an economist, really I’m not” theme. Meanwhile, I’ve been reading up on the voting literature for a forthcoming article I’m preparing for Contexts on the sociology of voting. I came across Duffy and and Tavits’ new article in AJPS on “pivotality”: more on my reading and thoughts on it after the break.

The thorny problem is this: voters basically “shouldn’t” vote from a rational-actor perspective, since even virtually every “close” election has pivoted on more than one vote, and most by thousands or more. Thus it’s essentially always more cost-effective not to vote than to vote, since it costs something to vote and the instrumental return is all but guaranteed to be zero.  This is embarrassing to RAT-minded political scientists since it is at wide variance with actual voter behavior; voters not only continue to vote but actually seem to do so more when their vote is less likely to be pivotal, i.e., they vote more in state and national elections than in local ones. Typically the theory determines that, well, people must like voting for some reason, whether solidaristic benefits or a sense of civic duty that is assuaged by voting, and if we add the utility generated by that term to the equation, well, then it’s rational again. Whew. Now if only we knew where people got that pesky sense of civic duty from….

Duffy and Tavits take an experimental approach to the pivotality problem. In small groups (20 or so each) of college students, they asked the students to decide whether or not they wanted to “buy a token”, set up some decision rules, and demonstrated that the students (a) were more likely to vote, given the pre-set cost of doing so, if they subjectively stated that they considered their votes more likely to be pivotal; but (b) often still ended up on the “wrong” side of the equation, i.e., didn’t vote even when their expectation of pivotality was higher than their cost or did vote when their expectation of pivotality was lower than their cost.

In some ways this strikes me as a useful, if small, demonstration of perceived pivotality as one piece, albeit apparently not the only piece, of the turnout decision. But I was intrigued by the fact that the experiment asked the subjects to decide whether or not to “buy a token,” a clearly economic act. It also monetized the costs of voting by literally charging students against their expected utility, all measured in cents.

At this panel, I wondered aloud whether part of the reason for the enduring success of the voting model emerging from The American Voter was that it told generations of media and citizens how to understand voting–that is, it was not just reflective but performative as well. The audience was unreceptive, to say the least. But, returning to Duffy and Tavits, it seems to me that this was precisely a performative task: tell the students that voting is essentially economic, even consumerist, in character, and then observe the extent to which they behave in a consumerist manner.

Author: andrewperrin

University of North Carolina, Chapel Hill

3 thoughts on “pivotality and performativity”

  1. I’ve always thought the whole voting thing to be a red herring in debates over rationality. Don’t people do things all the time that are expressive or symbolic? Isn’t voting itself so low cost that the RC argument seems weird? The RC argument assumes that a 15 minute action once every four years is some huge cost. It’s not, so it’s weird to think it’ll be covered by some elaborate cost-benefit calculation. Anyway, who am I to say anything? I’m really not an economist, really, I’m not.


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: