I am Allegra Bernstorff, a second year PPE student. Choosing a preferred area of study is challenging, but I admit that I enjoy learning about elections. Most of us are able to vote solely because our ancestors fought for that right. Elections are the cornerstone of democracy. They represent freedom, permitting us to choose how, and by whom, we are governed. Nevertheless, elections are necessary, not sufficient for democracy. While freedom to choose may be their purpose, whether this is achieved is another question. Here, political research comes in: to discover elections’ strengths, weaknesses, and how different systems impact results. These are vital questions we must ask; to understand what our right to vote actually represents, and if it is as effective as we might hope.

A paper I found particularly fascinating was, ‘Social Pressure and Voter Turnout’ by Alan S. Gerber, Donald P. Green, Christopher W. Larimer (2008), where four different letters were sent to households to see how they influenced their turnout.

Voting may be an important exercise of our freedom, but how impactful is our individual vote? The average probability of swinging the US presidential election is 1 in 60 million, (Gelman, Silver and Edlin 2012), so why do people vote?

This paper argues that we go to the polling stations because of social pressure. Anthony Downs’ model of rational voting suggests that civic duty is the deciding factor in an individual’s choice to vote. As the costs of voting outweigh the benefits, given that the impact of one vote is so small, the only factor that shifts this balance the other way, is the civic duty one feels.

But where does that come from? internally or externally? This is where ‘Social Pressure and Voter Turnout’ (Gerber, Green, Larimer 2008) comes in. The authors undertook a field experiment, of 180,000 households,  during the lead up to the Michigan primary elections in 2006. Overall, the turnout was 17.7% of registered voters. This relatively low number can be explained by the fact that the Democrat primaries where not very competitive, reducing the incentive for some voters to go to polling centers. Furthermore, in Michigan citizens are only permitted to vote in their registered party’s primary, so the Democrats who abstained could not have voted in the Republican primary instead. The researchers could not exclude all external reasons impacting turnout because they were unable to distinguish between Democrats and Republicans as there are no public lists of party registration in Michigan.

While Gerber, Green, Larimer employed a random selection process for which households they would test, they did exclude some from the list. This was principally for practical reasons, such as incomplete Zip codes which make bulk mailing harder, 60% more likelihood of voting via absentee ballot because those households would have already cast their vote by the time the research was taking place, and if the household was 60% more likely to vote in the Democrat primary because any mail sent to them would probably not influence their decision to vote or not due to the lack of competition. To avoid bias, however, the remaining households were randomly assigned to a control group (of 99,999 households), or one of four mailing groups (each containing around 20,000 households).

Bar the control group, letters reminding people to vote were sent, but the message differed between mailing groups:

  1. Reminding them of their civic duty. This tests the feeling of duty the individual experiences intrinsically (this message was also on all the other mailings).
  2. Notifying them that researchers would be studying their turnout based on public records. This tests if the threat of mere observation would influence turnout.
  3. Informing households that their turnout record would be sent to their household. This tests the impact of extrinsic duty. Does one’s likelihood of voting change with the rest of their household’s ability to see their turnout record?
  4. Informing households that they would receive their own turnout record, as well as that of their neighbors. This allowed the researchers to test the impact of maximum social pressure on a household’s turnout probability. Does the publication of a household’s turnout make them more likely to vote?

The results were as follows:

There is strong evidence that social pressure does increase turnout. The increase in turnout from those in the control group, to those that received the fourth letter, was 8.1%. Given that this percentage increase is larger than usual mailing experiments (circa 1%), we can assume that the results of this experiment depict the impact of social pressure on turnout, rather than the influence of mailing in general.

This impact that social pressure has on turnout suggests that while conformity with the norm may vary, there seems to be a consensus that voting is a norm. Therefore, the application of social pressure is additive. An external effect will positively influence turnout, whether or not the individual was predisposed to vote in the first place.

This information can be used to increase voter turnout in a cheap, effective way. According to the article, the letters applying the most social pressure had the same effect on turnout as door to door canvasing, which is far more expensive.

But is the fear of others’ judgement the right reason to vote? What about in countries where there is less social pressure to vote, such as Zimbabwe? Is their level of turnout less significant because voting is not considered a civic duty in their society?

Is there an intrinsic importance to voting, or is its value measured by what society deems to be important?

In my opinion, this is insignificant. As long as individuals are voting for what they believe in, the reasons behind why they got to the polling station in the first place are irrelevant.

However, a more pressing question is whether or not we want greater turnout. It may result in less informed voters taking part, because they didn’t take the time to inform themselves before pressured into voting, but this is an ethical question that needs to be further explored.



Print Friendly, PDF & Email