Satoshi Kanazawa

The Scientific Fundamentalist

Why Do People Vote? III

Maybe we vote because we think like cavemen

Posted Nov 29, 2009

Even after I left the field of rational choice theory of politics and became an evolutionary psychologist, the paradox of voter turnout (and my own theoretical resolution of the paradox) continued to haunt me.  Why do people act like their vote made a difference in the past electoral outcome?

From an evolutionary psychological perspective, it does make sense for people to feel personally responsible for the outcome of collective decision-making like an election.  Remember, the Savanna Principle suggests that the human brain is biased to perceive the environment as if it were still the ancestral environment, where we were hunter-gatherers living in a small band of 50-150 related individuals.  And in the ancestral environment, there were no large national elections with millions of voters or secret ballots.

Imagine you were our ancestor living on the African savanna as a hunter-gatherer.  If you say to your fellow band members, “Let’s go to the mountains to hunt wild pigs,” and your rival says “Let’s go to the forest and hunt monkeys,” and your band collectively decides to go to the mountains to hunt wild pigs, it is not at all unreasonable for you to think that your opinion (your “vote”) had an effect on the collective decision.  First, the group is small, so each adult in a 150-person band has a much larger share of the collective vote than a citizen in the US Presidential election.  Second, “voting” (collective decision making) in the ancestral environment did not involve secret ballots.  Everybody in the group knew how everybody else “voted.”  Unlike in modern national elections, your and your rival’s opinions may not have been weighed equally.  So you have all the (rational) reasons to believe that your vote had an influence on the collective outcome, and you should continue trying to influence the group’s collective decision.  Conversely, your rival has all the (rational) reasons to believe that his opinion does not carry weight in your group, and he should perhaps cease trying to influence the group decision.

In my 2001 Social Forces article, I speculated that this may be why we take the electoral outcome of the last election as a reinforcer or punisher on our last response (either voting or abstention).  Eight years later, a team of neuroscientists, led by Kevin S. LaBar in the Center for Cognitive Neuroscience at Duke University, present the first direct evidence for my speculation.

In their article “Dominance, Politics, and Physiology:  Voters’ Testosterone Changes on the Night of the 2008 United States Presidential Election,” published in PLoS One, LaBar and his colleagues took saliva samples from 57 male and 106 female voters on the night of the 2008 Presidential election, at 20:00 before the winner was announced, at the moment the winner was announced, 20 minutes after the winner was announced, and 40 minutes after the winner was announced.  As the graphs below show, women’s level of testosterone does not appear to vary as a function of which candidate they supported.  In sharp contrast, among men, those who voted for Obama maintained high levels of testosterone, whereas those who voted for McCain and other minor candidates experienced a precipitous drop in their level of testosterone 40 minutes after the winner was announced.


Earlier studies have shown that the testosterone levels of male tennis players and male chess players (yes, chess players) fluctuate as a function of the outcome of the contest.  The winners’ testosterone levels stay high after their victory, whereas the losers’ testosterone levels drop precipitously after the defeat, probably to induce submissive behavior following their defeat in the contest.  LaBar et al.’s study is the first empirical demonstration that supporters of Presidential candidates take the electoral outcome as personally as tennis and chess players take the outcome of their games.  It appears that at least the male voters take the electoral outcome very personally, consistent with my speculation.  If this result is replicable and generalizeable, we may finally know why people vote.