Do people make good choices when they vote? *Can* they?
How come most of us, most of the time, don’t do crazy things? We make hundreds of decisions every day – lots of small ones, and once in a while some more momentous ones. With each one, we have the opportunity to do something really stupid, and yet we don’t. That is pretty remarkable.
Imagine you’re trundling along, lost in thought, and you start crossing the road without paying due attention to the traffic. Suddenly you hear the blast of a horn. You don’t stop to look where it comes from and try to identify what it might be. No, you unceremoniously jump back onto the pavement. That’s your thinking System 1 at work, quick and impulsive, capable of making a decision and instructing your muscles to carry it out in a split second.
But if you are buying a house, you would not take a decision at such speed. On the contrary, you would probably take days or weeks to mull over the pluses and minuses of several possible purchases. Are there bus stops and maybe schools nearby? Is it on a busy road or in a quiet backwater, and is it close to the local amenities? That would be your System 2: slow and thorough, analytical and reasoned, capable of weighing up multiple facets different options.
The right upper hand
Of course you might, immediately after nearly having been run over, be a bit embarrassed and self-conscious about that inelegant jump back onto the pavement. But that consideration comes afterwards. It’s your System 1 that has the upper hand when life and limb might be at risk, unencumbered by worries about looking foolish. On the other hand, you might have felt an immediate attraction for one of the houses you’re considering, but even that will eventually be just one factor in the much more analytical decision making of your System 2 when determining on which house you will spend (and borrow) several years’ income.
Economics Nobel Prize winner Daniel Kahneman popularized the concept of System 1 and System 2 thought (first coined by psychologists Keith Stanovich and Richard West). In his book, Thinking, Fast and Slow, he illustrates how, sometimes, we do make impulsive decisions that had better be more reasoned. Anyone who has ever eaten a whole family bag of crisps before dinner, or drunk so much alcohol that they woke up with the mother of all hangovers the next day will understand. Sometimes we do the opposite, agonizing way too long over a decision where the possible outcomes don’ really differ that much. But by and large, our System 1 and System 2 serve us well, and help us make sure we make choices that are in our long-term self-interest. (If we systematically failed to do that, our species would have long been extinct).
You would expect that if you put this powerful decision-making capability of a whole bunch of people together, you’d end up with pretty good decisions.
Democratic elections come to mind, for parliamentary representatives or presidents, or even referendums. By sheer force of numbers, the extremes of impulsive votes (like protest votes) would be moderated by the more reasoned votes. That assumption seems to have worked for most of the post-WWII period in the western world and beyond… until recently. Political analysts will no doubt continue to study the Brexit referendum and the election of Donald Trump as US president. But meanwhile there is a broad consensus that, in both cases, protest votes contributed significantly to the unexpected outcome.
System 2 overloaded
Perhaps the surprise is not so much these recent unforeseen outcomes, but the fact that they have been so rare in the last seventy years. Do elections really distil the wisdom of the crowds – does it tap our System 2? We, the electorate, are given very complex choices. It is exceptional for any candidate or any party to stand for a programme that coincides entirely with our own preferences. So we zoom in certain aspects – what is salient to us.
But even narrowing things down to what matters to us generally doesn’t help us reason and weigh up the political options. There may be elements in one manifesto with which we strongly agree, but others with which we strongly disagree. Likewise for the alternatives. So how do we aggregate all this? Our poor System 2 has no idea what to do with it – so we revert back to our beliefs, heuristics and gut feel. How did we vote last time? How do the members of our ingroup vote? Where do our ideological convictions lie? Or indeed, which of the candidates’ looks do we like the looks of the most or the least?
The US Senate special election in Alabama on 12 December provided a compelling case study for the tensions that voters face.
It was a two-horse race in a state that had not elected a Democratic Senator since 1990. According to Wikipedia, Republican candidate Roy Moore is firmly and consistently on the far right of the political spectrum. However, he was also controversial as a judge, and most importantly, he is the subject of multiple sexual misconduct allegations.
In a polarized political climate, as the US currently experiences, the choice for the Democrats was obvious. But for the Republicans it was different. Mr Moore’s simple and clear, but extreme politics might not be to the taste of many moderate ideological Republicans. Should they have let tribalism dominate and vote for the person with the right party affiliation? Or should they have voted for Doug Jones, the Democrat candidate instead, because his politics might actually have been closer to theirs? (That is what 8% did.)
And what with the sexual misconduct allegations? Should they play a part? Some people concluded that 48% of voters were willing to elect a paedophile (one of the sexual abuse allegations involved a woman who was 14 at the time).
That’s an easy accusation to make if it concerns a candidate for whom you’d never vote anyway. But imagine a candidate who is perfectly aligned with your politics. At what point would their morality make you vote for their opponent?
Faced with such an impossible dilemma (something many also experienced at last year’s presidential election), as an individual voter you could decide to abstain and avoid going loopy. This is what 60% of registered Alabaman voters did. But as a community, this is irrelevant: someone will eventually be elected. Will that be the best candidate – the crowd’s System 2 choice – if individuals opt out?
Politics should really be the domain of our collective System 2. Elections have a huge influence over our future. Our representatives and leaders are elected for four, five, six years or even more, which is already long enough. But the decisions they make on our behalf stretch much further out.
Even if the choice we face is not between politics and morals, but between two complex sets of political preferences, we are up against an impossible task. Taxing labour or taxing wealth? More public provision of healthcare or private insurance? Free trade or protectionism? Same-sex marriage or traditional values? Legalizing drugs or three-strikes-you’re-out? Our individual System 2 can simply not cope, and so by default our System 1 takes over.
Maybe one day Google’s AlphaZero Artificial Intelligence will be capable of helping us poor bags of meat and bones with solving real-life challenges, as well as play games – however complex.
But until that time, we should concede that any hopes of System 2-driven electoral democracies are idle. We had better be ready for the consequences of many more collective System 1 decisions.