GE 2015: Can electoral decision making be rational?

Colin Davies, Chair in Cognitive Psychology

Colin Davies, Chair in Cognitive Psychology

How do we go about deciding who we’ll vote for? And how should we go about deciding?

For every decision there is, in principle, a rational way to decide. Political scientists have formulated the idea of ‘correct voting’, whereby a voter can be said to have voted correctly if they have selected the candidate or party whose priorities are most closely aligned with their own. If Mr Spock were voting, this is the logical, rational approach that he would take.

Unsurprisingly, perhaps, voters are not perfect, rational decision-makers. For one thing,  voters typically have insufficient information to make a rational decision, either because politicians haven’t released relevant information (where they would make cuts, for example) or because voters haven’t availed themselves of this information.

But a more fundamental reason why voters don’t always choose rationally is that decision making in general is not rational, as demonstrated by a long history of psychological research, beginning with the Nobel prize-winning work of psychologists Daniel Kahneman and Amos Tversky. It’s important to qualify this point by noting that not being rational doesn’t prevent us from — usually — making good decisions. There may be very good reasons for us not to strive for perfect rationality in most situations: a fast, “good-enough” decision may trump a slow, optimal decision.

People make fast decisions by using heuristics — mental shortcuts that allow one to arrive at a solution without taking into account all of the relevant information. This works surprisingly well. Applying appropriate heuristics often results in the same decision that would emerge from far more complex calculations that weigh all the different alternatives and take into consideration many different factors. Researchers such as Gerd Gigerenzer have argued that, when faced with complex decisions where there is a great deal of uncertainty or a lack of information, the right thing to do is to make use of heuristics.

If the application of heuristics does tend to approximate rational voting, it may not be unreasonable for some voters to spend more time weighing up their lunch choices than deciding who they’ll vote for. Knowing that their individual vote is unlikely to decide the outcome of the election, people are not strongly motivated to expend the degree of cognitive effort that would be required to make a truly rational decision.

However, there’s good reason to challenge the notion that there are easy shortcuts that will reliably approximate rational voting. The application of heuristics often relies on unconscious processes, leaving us vulnerable to influences that are very far from rational. A heuristic that we often rely on is to evaluate options on the basis of the information that is at the forefront of our mind. But a downside of this “availability” heuristic is that it provides a basis for various external stimuli to prime our decisions, often without any conscious awareness. This reliance on the availability heuristic is probably responsible for the effect of voting location (studies in the US have shown that voters are more likely to support pro-education propositions if the polling station is in a school). Similarly, reliance on an “endorsement” heuristic leaves us profoundly vulnerable to social influences, such as polling data and the “worm” graph used in televised election debates. Another mental shortcut (“is this option better than my current favoured option”?) may explain the influence of the order of the names on the ballot (the first-named candidate has a statistically reliable advantage, which increases in size according to voter ignorance). In a close election such influences could turn out to be decisive.

Is ‘correct’ voting possible?

Research showing the influence of variables like name order, voting allocation and candidate appearance point to the downsides of relying on heuristics. Is it possible for voters to avoid being swayed by irrelevant influences? One solution might be provided by decision-making tools such as and These are independent websites that present the user with a set of questions about policy attitudes, and then compute a percentage overlap between that user’s opinions and the positions of each of the different parties. These sites also make it easy to find out where you agree with a party and where you disagree. This is an excellent way for voters to find out what their ‘correct’ vote would be. Even if you don’t ultimately vote in this way, such tools can be a valuable means of educating yourself about the policies of the different parties.

Of course, there’s another factor that interferes with ‘correct’ voting – the widespread use of tactical voting in UK elections. The first-past-the-post (FPTP, perhaps better named winner-takes-all) voting system used in the UK encourages the rational voter to engage in a strategic game akin to the prisoner’s dilemma, in which they must consider not only their own preferences, but also the likely behaviour of other voters. By motivating voters to cast their ballot for a candidate that does not best reflect their own priorities and values, the FPTP system is a considerable barrier to correct voting. For this, and other well-rehearsed reasons, a change to our voting system is long overdue – the UK should move to the proportional representation system used in the vast majority of European countries.

Given the importance of electoral decisions, it’s worth thinking about how we can minimise irrational influences and find ways to select the candidate or party whose policy positions are most closely aligned with their own. The use of tools like votematch and voteforpolicies may be a good way to achieve this aim. Beyond that, however, systemic changes to British democracy are needed to ensure that the outcome of elections is a good reflection of the priorities and beliefs of the electorate.

The views expressed here are personal views and do not reflect the views of the funders of our research.