Voter Behaviour: Rational Choices, Emotional Decisions, and the Occasional Coin Toss

Voter Behaviour: Rational Choices, Emotional Decisions, and the Occasional Coin Toss

A Field Guide to the 17 Million People Whose Aggregated Preferences Determine the Government, None of Whom Are Doing Exactly What the Models Predict

Political scientists have been attempting to explain why people vote the way they do since the discipline invented itself in the early twentieth century, and have produced, after a century of research, a set of theories that are individually plausible, collectively contradictory, and together constitute a reasonably honest account of the fact that human beings make political decisions for reasons that include ideology, identity, habit, self-interest, social pressure, family tradition, the performance of the leaders on the night, whether they had a good week at work, and what the candidate’s corflute looked like on the drive to the polling booth – and that no single theory accounts for all of these simultaneously.

This is not a failure of political science. It is an accurate finding about the complexity of human beings, delivered in the language of a discipline that prefers to call it "the multi-causal nature of electoral choice" rather than "it’s complicated," both descriptions being equally accurate and one being significantly more fundable as a research program.

The Theories: A Brief and Uncharitable Survey

The rational choice model holds that voters assess the policies of competing parties, determine which aligns with their interests, and vote accordingly. This model is elegant, internally consistent, and describes approximately 15% of actual voter behaviour, a proportion that political scientists call "the informed voter segment" and that the remaining 85% call "people with more time than us."

The party identification model holds that most voters have a standing psychological attachment to a party, formed in childhood or early adulthood, that filters political information and determines vote choice in all but the most extraordinary circumstances. This model explains a great deal of stable voting behaviour and considerably less of the post-2016 volatility in most democracies, which suggests that standing attachments are less standing than they used to be, for reasons that the party identification theorists are currently studying.

The economic voting model holds that governments are rewarded for good economic conditions and punished for bad ones – that when unemployment is low and incomes are rising, the incumbent wins, and when the reverse applies, the incumbent loses. This model has a robust empirical record and two problems: it works at the aggregate level but poorly predicts individual vote choices, and it requires defining "economic conditions" in a way that agrees with how voters experience them rather than how statisticians measure them, which are sometimes the same thing and sometimes not remotely.

The vibes model – not formally recognised in the academic literature under this name, but operationally dominant in how campaigns are actually run – holds that voters develop an overall gestalt impression of the leaders and parties, assembled from many data points and not reducible to any single one, that functions as their primary electoral heuristic. The vibes model predicts that a leader who feels trustworthy wins over a leader who has better policies but feels shifty. It predicts that an election campaign that produces a defining negative image of the opposition leader – untrustworthy, out of touch, risky – will succeed regardless of the policy comparison. It predicts, in short, that elections are character assessments conducted in conditions of incomplete information, which is an accurate description of virtually every federal election held since 1975.

The Swinging Voter: Democracy’s Most Important and Most Exhausted Citizen

The swinging voter – the approximately 10-15% of the electorate that does not have a fixed party alignment and whose preferences determine election outcomes – is the most studied, most targeted, and most discussed figure in Australian electoral politics. They live, statistically, in western Sydney, outer Melbourne, and a handful of regional cities. They are concerned about cost of living, housing affordability, and the performance of the health system. They are sceptical of both major parties. They decide elections.

The swinging voter is also, by definition, the person both major parties spend the entire campaign trying to reach – tailoring messages, designing policies, and calculating costings for an audience of approximately 1.5 million people while the remaining 16 million watch and wait and either appreciate the attention their concerns are receiving or find the targeted messaging slightly insulting depending on whether they live in a marginal seat or a safe one.

"Choosing a government," observes Professor Reginald Clipboard III, reaching for a cereal analogy his graduate students will find reductive, "is like choosing a breakfast cereal. Brand loyalty explains most purchases. Promotional activity targets the unconvinced. The health claims on the box bear an attenuated relationship to the product. And the decision is made, ultimately, in front of a shelf, under mild time pressure, by a person who would rather be doing something else."

For the companion study in voters whose decisions confound the models that were supposed to explain them, The London Prat’s electoral behaviour files – prat.uk – tracks British voters with similar bewilderment and similar rigour, finding broadly the same result: it’s complicated, the vibes matter, and the cereal analogy keeps coming up.

SOURCE: https://bohiney.com