Tim Harford’s chapter eight, a contribution from Sahar Akhtar

Harford writes that voters aren’t fooled into thinking their
votes affect the outcome and that most people vote because it makes them feel
good. These ‘expressive’ explanations
help us preserve the idea that people are rational (a great book is by Brennan
and Lomasky).  But is this an actual account of why people
vote? Until we have better survey data, anecdotal evidence will have to do.

Go to a diner, bus stop, retirement home or even a college
campus and almost invariably people will tell you that their vote counts. What
does ‘count’ mean here?  It might mean
they think their vote is important because it satisfies a civic duty to support
democracy (but why would so many think this
is the best way to discharge that duty unless they think their vote counts in
the more literal sense). Or maybe it means they think their vote somehow
encourages more people to vote (but why isn’t lying more efficient? And why
would people get influenced into voting, unless they think it matters? ) Isn’t
it possible to think that people actually believe their votes count?  But if this were true, how could we best make
sense of it? One way is to bite the bullet and accept that (a lot of) people
might just be irrational.

Voting does of course increase with education level, but
this doesn’t defeat the claim that voters might be irrational. Most of our civic/political
education in high school and college centers on the details of how democracy
functions and why voting is important, and not on the trivial impact of our

And remember the way that voting works in the U.S. at least—through some freaking inscrutable thing known as the electoral college.
On my not so good days, I still have no idea how this works and, like most MR
readers, have above average education. Does my vote count more in states with
fewer delegates, or not at all in some locations, or because this is a
republican state does this mean my vote for a democrat wouldn’t matter or would
it matter more, and where is this school? My neuronal synapses die just a little.

Also, without voter irrationality it’s hard to make sense of
the success of campaigns such as “a vote for Nader is a vote for Bush” in resonating
with potential voters—the aim of these kinds of slogans is to encourage people
to vote in a particular way. If people don’t believe their votes count,
why would these slogans be effective and why would the slogan designers
anticipate they would be effective? Harford seems to hint at this kind of
problem when he points out that while voters don’t go to the polls to impact
results, they don’t realize that what they do once they get to the booths
doesn’t matter—but I wish he gave us his thoughts on this sort of inconsistency.

The fact that people don’t simply vote, but vote for a
particular candidate, at best suggests that if people feel duty-bound it’s not
to some abstract ideal but to particular parties and groups, which raises another,
and not incompatible, potential motivation for voting.

Some might think that their votes count not individually,
but as part of a group. Harford and other economists aside (including this
one), people don’t always act on their (individual) self-interests.  (for just some examples, see Fehr and Fowler on
altruistic punishment)

There are good evolutionary reasons to think that we
frequently adopt the perspective of “what is good for us”.  You don’t have to believe in the
group-selectionist theories of people like Sober and Wilson.

If that makes you feel dirty—selfish
gene will get you there if there are enough genes shared in common among a
group. And, a la Robert Frank, what
starts out as emotional incentives to act on behalf of a fairly specified,
narrowly defined, and kin-based group gets co-opted and extends (irrationally?)
to larger, less cohesive groups. The group in this case would simply be the
class of people thought to share the same values and beliefs.

Of course, like all evolutionary explanations, this is a
just-so story and needs to be tested, but so does the rational voter idea. We still don’t have very good insight into
the motives of voters, and until we do we should remain skeptical of any one

I’m not a hater–in many (maybe most) areas of life, the rational
choice model makes damn good sense. In
some areas of politics, however, emotions run high and irrationality can be
bliss, and these may be areas where dynamic writers like Harford should resist the
model a little.

Back to TC: Readers, do tell us what you think…


I vote when I'm afraid of a certain candidate on a couple of key issues. The preservation of the right of self defense is a dominant concern for me. If I am made to feel threatened on that point, I become a voter. In the absence of a threat to a key policy, I don't vote. It is emotion that brings me to the booth - mostly I'm avoiding a sick feeling of having a scary policy imposed while I did literally nothing about it. My fear trumps my rational understanding that I'm not really contributing anything measurable even when I pull the lever.

I suspect that most voters have similar thoughts about specific issues, perhaps about abortion or, recently(and bizarrely), immigration.

Let me focus my comment on the idea of voting in the hopes of casting the winning ballot as rational or irrational:

It seems to me that using a marginal calculation to justify not voting can only be rational on the assumption that other voters are indeed irrational and will stream to the polls.

But here we have again the fundamental problem: Tim would like us to believe these other voters are rational because they can justify their actions on other grounds, such as the fact that it makes them feel good to vote. But even if we grant that people are voting for reasons beyond a strict probability calculation (which I'm not sure we can), does this really say anything at all?

It's interesting to discover hitherto hidden reasons that compel people do act a certain way, but I wonder if Tim has stretched the definition of rationality too thinly trying to fit them all in.

Ideology, group pressure, advertising that turns not voting into a civic sin creat the public good (or bad) of voting.

The main (not only) reason most people vote (at least where I live) is the same reason most people don't litter: There are strong norms against contributing to collective good problems. How often do parents use the argument "What if everyone did that?" when arguing with their children? If memory serves, very often.

I don't think voters hold any beliefs at all about whether voting is important, not in the same way they hold beliefs like "falling off a tall cliff will lead to death", which leads to taking action to avoid falling off cliffs.

I think the "civic duty" explanation can be yoked into greater service than is usually done in these discussions. Think about feasible institutional purposes of 'civic duty'. Representatives in a republic are supposed to vote in a way that their constituents would do, but this system could go horribly awry if The People aren't watching. Depending on your beliefs about the long-term stability of our Entire Political System, voting may be a way of letting Washington know that we're paying attention—and thus have more to do with keeping the mass of Watchers (Quis custodiat custodes) high than with effecting a particular political outcome. To invoke a buzzword, voting may be a rational response to a fat negative tail. (Indeed, this is what informs our civics classes: momentous or catastrophic changes are prominent objects of study in political philosophy, and some schools of historical thought believe history proceeds in jumps.) JasonL's first post reflects this type of concern. So does Donald A. Coffin's (five above mine).

Also, just because our models can't explain behavior doesn't make it irrational! A proof of irrationality should show that behavior X cannot be rational—i.e., that there is some unavoidable axiom which all models of rational agents must satisfy, which this behavior violates.

People vote because they anthropomorphize concepts such as "society as a whole", "government", "the US". We live our lives within families, firms, and institutions that operate much more in the obvious goal/action mode. And, although government failures have being documented over and over, very few people identify the problem in the fundamental disconnect that exists between government goals and government actions. It is easier to think that bad people are in charge, which leads for some to a general skepticism of politics and a tendency to vote irrationally (or seemingly so) just to bother the guys in charge, and for others to a messianic belief in voting as the supreme tool for righting all wrongs.

We simply feel better when we vote. The sense of civic responsibility that we feel, the pleasure attached to the idea of being an active member of the community, even the entertainment derived from the information gathering before voting or the tension prior the announcement of the results, all these are individual subjective rewards built up in our brain by evolution in order to compensate for the lack of more norrowly defined benefits (as everybody knows that individual votes won't influence the results). But of course the cost of voting must be kept low, especially if voting becomes a routine and the marginal utility derived from voting diminish. In Switzerland where we vote several times a year on different federal, states or local issues, participation rates are higher in jurisdictions that allow mail voting than in states where people need to go personally to the polls.

I do believe that people DO act as if their vote really counts. I have to say that I think that mine does, even if in reality it does not. I think people have to hold on to the thought that their vote does count, because otherwise, they would not even bother voting. Regardless of whether it counts or not, it makes people feel as if they have a part in decided who will be next in office. This "illusion" of power allows us to feel that, as US citizens, we play a role in governmental decisions. Even if the amount that each individual voters influence is small, together, we feel powerful.

I agree with Donald Coffin that there does seem to be some kind of instrumental reason why people vote, myself included. But the motivation to promote democracy just doesn't seem to take into account that people vote for specific candidates--if our concern is to promote democracy in general, then why not choose a candidate at random once we're in the booth?

Jeff Holmes and Colin both raise good questions about our definition of rationality. As Jeff argues, including duty type of motivations in what counts as rational starts to weaken the notion of rationality to the point of it being meaningless. Plus, contrary to what AZ writes, what's interesting about the phenomenon of voting gets pushed aside if we can easily explain it with the sorts of rational explanations typically offered (a student of mine put this point nicely to me). And I really liked Colin's suggestion--if 'irrational' behavior is needed from a collective perspective, perhaps this means we should reconsider our definition of rationality.

thanks to everyone for the great feedback...

Tim, I actually don't think economists are motivated only by self-interest. I think we believe we are! :-)

A little off-topic, but a colleague of mine at Brown (Jason_Brennan@brown.edu) has written a great paper on the immorality (not irrationality) of voting by the average voter. He kind of puts an interesting twist on things...

Comments for this post are closed