In 2007 Bryan Caplan, a professor of economics at George Mason University, published a fine book entitled ‘The Myth of the Rational Voter’. In it, he robustly questions the ability of the average voting citizen to make reasoned choices, especially in matters about the economy.
Nearly ten years later, on 24 June 2016, professor Caplan tweeted
It is widely accepted that RT ≠ endorsement, but the likes and retweets suggest that some people do think he has a point: voters are not rational. The context is, of course, the EU membership referendum that took place in the UK one day earlier. Contrary to most polls, the option to leave the EU gained a majority of the votes. And it is particularly those who voted Leave whose rationality has been doubted, for example in this Vox article, albeit not so much in the sense that Bryan Caplan addresses in his book, but rather on the basis that they might have voted out of deeply held emotional beliefs, subject to biases and unmediated by any reasoning about the likely economic consequences.
Is that really true?
Biases don’t necessarily mean irrationality
One definition of irrationality is indeed the kind of behaviour that results from biases and fallacies, and not from reasoned consideration of objective factors That would in particular be so when conscious choices are involved — such as voting (which in most cases implies at least some degree of mindful deliberation).
All choices are, in essence, trade-offs. We must sacrifice certain scarce things in order to gain certain other scarce things that we value more. Naturally, different people have different preferences, so we seem them make different sacrifices for different gains. But if someone makes a choice that goes against their self-interest and makes them worse off — something economists would call negative utility — and they knew, or could have known, beforehand that this would be the case, that would be an irrational choice, guided by biases rather than reason.
It is sometimes argued that playing the lottery is irrational in this way, because the expected net gain of doing so is negative: in the long run, on average you lose money. Players are of course, so the assumption goes, biased: fooled by the magnitude of the jackpot and ignoring the small probability of actually winning it. Yet this is a very narrow perspective, which does not take into account any other aspects of buying lottery tickets. Few people play for the sole purpose of getting rich. The expectation, the thrill of the draw, maybe even the symbolism of using significant dates to choose your numbers — all of these rely on the lottery biases, but they also provide entertainment and pleasure. This adds to what is valued by the players, and they’re happy to pay a token amount for the experience. So the argument that they are irrational loses much (if not all) of its validity.
So what’s with the biases that presumably swayed irrational voters in the referendum, such as the status quo bias and the endowment effect? There is quite a bit of evidence that people in general tend to stick with current situation — “better the devil you know” — even if the alternative might be objectively superior. The news website Buzzfeed consulted several pollsters in April, and many saw the status quo bias as a significant factor. Peter Kellner, president of Yougov and a veteran political analyst (so a bit of an expert), expected enough undecided voters to opt for the status quo to hand victory to the Remain camp.
The endowment effect and its close cousin, loss aversion, describe how people disproportionately value what they possess and the disproportionate fear of losing it. Those who consider themselves endowed with the benefits of EU-membership would therefore be inclined to favour Remaining, scared of losing those advantages. This is what Philip Corr of London’s City University said back in May, adding that the same effect could just as well have influenced those preferring to Leave, afraid of losing (even more) jobs and the cohesion of their society to persistent immigration.
How well does the claim of irrationality, perceived in these biases, stand up when you take a broader view? The endowment effect does look irrational, but only if you ignore the emotional significance of owning something — even if only in a metaphorical manner. The emotional attachment to what you perceive as your country, its values and its customs may well be a lot stronger than the emotional and material attachment to an abstract number like £4,300 per year (the estimated cost per household of a Brexit). And emotions are what makes humans tick. Nothing irrational about trading stuff off in that way.
The status quo bias also seems irrational, until you take into account the transaction cost of going for the alternative. Changing from something that is broadly acceptable feels risky, with the prospect of a certain effort, hassle and sacrifice, in return for an uncertain payoff. Most people are satisficers for whom most of the time good is good enough. For them it is perfectly rational to prefer an easy life without the hassle of worrying about what alternative might be better, and taking the trouble of changing.
Rory Sutherland put it very succinctly on Twitter:
Not so irrational
And yet we seem rather quick to label something as irrational. This is partially because we have to believe that we ourselves are rational (how could we function if we were convinced that we are fundamentally irrational?). We make the connection between being rational on the one hand, and making the choices we make on the other, and so it is easy to simply assume that whoever makes a different choice is irrational.
If we don’t fall into this glaring trap, another one lies in wait. We see our rationality not just reflected in our eventual choices, but also in our decision process. Those who share our own approach are then also rational, even if they end up with a different outcome (because they may have different preferences). And anyone with a different approach is not. What we have here is in essence the previous mistake, only dressed up with a thin layer of pretend objective validity. We believe our way of thinking is the only rational one.
But most importantly, we may be guilty of seeing others as plain, uncomplicated beings, who follow a simple, direct path towards their choice. We tend to see these paths plotted as a straight line — a bit like herringbone diagrams in navigational rallying — blind to the many trade-offs people make on the way, like the junctions on the diagram. Our choice is sophisticated and reasoned, theirs is made impulsively and unthinkingly — and hence irrational.
This perception gives us a pixelated picture of the world, in which we see people in a small number of homogeneous groups. In its most simplistic form, Remain voters would see themselves as rational and led by the economics, and the Leave voters as led by powerful, raw emotions. Maybe every emotion would have its own cluster of adherents: a xenophobic group led by fear of further immigration, a disaffected group led by anger with the establishment and the elite, and a libertarian group with a distaste of big bureaucracies. Slightly smaller pixels, but still: pixels.
Assuming in this simplistic way that people vote out of a one unique motivation denies the complexity of thought we all possess. By and large, we are not driven by single issues. Most people are perfectly capable of considering multiple causes and consequences and of using this capability to reason in situations like this. An angry protest voter may perfectly well have considered the possibility of an economic cost to a Brexit. A committed libertarian may perfectly well have weighed up the regulatory disadvantages of a Norway option with the benefits of being part of the single market.
So, it’s not irrational to have your vote guided by something you value highly, like sovereignty or less immigration (irrespective how fuzzy the concept of sovereignty might be, or how unattainable stopping migration). These choices can provide real utility, and may be worth whatever sacrifice you have to make for them. Neither is it naturally irrational to be guided by a sense of protest against an elitist establishment that, in your eyes, has been ignoring, marginalizing and bullying you for years — especially if the polls and the media reassure you that the outcome of the vote will be to Remain anyway, and suggest the cost of your protest is minimal.
Does this not conflict with an earlier post in which I argued that referendums suck? Not really. A rational, reasoned decision process does not guarantee the right choice. If you are being misled and misinformed, and are confronted with an impossible question, you don’t stand a chance, no matter how rational you are.
This does not mean that we cannot be irrational. However, when we judge other people’s behaviour — or indeed how they vote — perhaps we ought to err towards presuming rationality, even if emotion is involved. After all, if we were truly fundamentally irrational, we must have been extraordinary lucky for millions of years to have evolved as far as we have. It is precisely our emotions that allow us to make the choices that are in our own rational self-interest.
And this is a key element in discussing rationality: how can we establish whether someone is irrational without knowing for sure whether they are acting for or against their own self-interest? As Leigh Caldwell explains, determining people’s real preferences is a deeply challenging problem, and our own projections and prejudices are useless as approximations.
Voters may have a different opinion from our own. They may have different priorities and preferences, and they may be led by strong beliefs and emotions — to the extent that they are susceptible to questionable claims and promises. They may eventually even feel regret.
But it is not because someone made a choice in which, eventually, the costs end up outweighing the benefits, that they have acted irrationally. Ignorance is often rational, particularly in politics. It is certainly not the same as irrationality — especially if what you didn’t know couldn’t really be known.
Labelling people as irrational, when there are no obvious flaws in their reasoning, and when we cannot establish whether they act against their own self-interest, is flawed. The irrational voter is, until further notice, at best a sloppy stereotype, and very much a myth.
Originally published at koenfucius.wordpress.com on July 3, 2016.
(The title to this piece was at first erroneously shown as ‘The Myth of the Rational Voter’ — now corrected.)