Being a longtime political junkie and occasionally involved in campaigns I suggest that, putting the ideologues aside, the driving force in voting behaviour is perceived self-interest.
I’m not so sure. Though there is a complex relationship between opinion on issues and voting behaviour, a self-interest hypothesis at best seriously under-explains opinion on numerous issues, and in some cases people hold opinions that seem contrary to their self-interest.
Self-interest under-explains opinion because there are many issues in which the voter (or poll respondent) has no personal material stake. People are passionate about the ‘sorry’ issue though, as the ‘practical reconciliation’ critics point out, it in itself will make no material difference to anyone. The republicans don’t argue that doing away with monarchy will make us richer, but that it will somehow make us feel more independent. The gay civil union/marriage issue cannot materially affect more than 2-3% of the population, and probably much less, yet most people seem to have a view on it. The government’s Tampa exercise was very popular, though most Australians live many hours flying time from where the boats come in, and most will probably never meet a refugee. The Iraq war is unpopular, even though few Australians know any soldier serving there and the cost has had no impact on daily life back home.
In the most recent federal election, many people would have struggled to identify a financial reason for voting one way or the other. The two parties offered very similar taxing and spending policies. In state elections a financial rationale for voting decisions is even harder to find. There are always targeted bribes during election campaigns, of course (pollies have to pull what levers they can), but what voting impact these have is not clear. If electoral bribery reliably converted dollars to votes, John Howard would still be in the Lodge.
Even where some potential self-interest can be identified, it is a poor predictor of opinion. In my recent article on industrial relations reform, I showed that both opposition to and support for WorkChoices greatly exceeded the proportion of people who thought that they would be disadvantaged or advantaged by the policy. In my article on protectionism (pdf) a few years back I found polling evidence that significant numbers of people understood the benefits of free trade, but still supported tariffs because they were worried about other people’s jobs. In my work in the past on the rising proportion of voters wanting more taxing and spending, I found that income had little influence on opinion – even though upper-income Australians would almost certainly end up paying more than they received in return for such an arrangement (though with the caveat that Graham makes about ‘perceived’ self-interest – they may not all realise that they would be net losers).
In the Morgan Poll series of questions about important issues, respondents give different answers to questions about which issues the federal government should be doing something about, and which issues would benefit the respondent and his/her family.
To draw on Bryan Caplan for the second time this week, there is a perfectly rational reason why self-interest might be a poor predictor of the vote. This is because (outside of McEwen at least) the chance of any one vote influencing the result is low, making voting an ineffective way of advancing material interests. But even without affecting the result, voting can express a sense of what we stand for (social justice/patriotism/the environment). People who, when arriving at the booths, ostentatiously take only one party’s how-to-vote card are making a statement, and not just casting a vote. Voting becomes an expressive act rather than a calculating, self-interested act.
As the ‘post-materialist’ theories of political behaviour have long maintained, as more people meet their material needs they will look for other things through politics. The middle-class left, consistent with this theory, is now a big consumer of symbolic politics – just at the middle class generally spends a lot on brands, so that they have the right image. (There is a nice David Brooks column in the NYT this week on this.)
One of the Coalition’s political problems is that while it is reasonably good at devising policies to meet material needs, it is weak on symbols. That is why, for a while, it started picking up working class voters, who still wanted to improve their standard of living, even as it was shedding doctors’ wives. WorkChoices was a political disaster because it upset both working class self-interest and middle-class expressive politics, which did not like its ‘unfairness’.
Self-interest is never likely to be irrelevant to politics. But there is no way a theory of public opinion and political behaviour can be built on it, and it is likely to become an even poorer guide to elections in the future than now.