It is now conventional wisdom that many of our problems are the result of living in an environment very different from the one in which we evolved. A food-scarce environment favoured gorging ourselves on energy-dense food, life in small tribes favoured suspicion of outsiders and an environment in which threats were predominantly physical favoured a ‘fight or flight’ response to stress. These behavioural dispositions tailored to the Environment of Evolutionary Adaptation (EEA) have deleterious consequences for people in the environments which they now live, at least in the western world. Obesity, racism, violence, and the physical symptoms of stress seem at least in part a result of our adaptation to the EEA.
Our anachronistic behaviour harms us not only in our day-to-day activities, but also insofar as our stone-age minds determine our political preferences. Just as we have an intuitive folk physics and folk psychology, we also have a folk economics which enables us to do folk policy analysis. One political phenomenon I think maladaptive heuristics can illuminate is the prevalence of moral panics.
Tversky and Kahneman taught us that humans make use of various cognitive shortcuts when making decisions. One of these is the availability heuristic – the tendency to judge the frequency of a phenomenon by how easily it can be brought to mind. This obviously depends on how often we have experienced or heard about something, and also on its salience and ease of imagination. Further, it seems that our estimates of probability and specifically risk are shaped by emotional rather than simply cognitive factors. It is largely our visceral response which determines our evaluation of different risks.
A likely factor in explaining the availability heuristic is our evolution in an environment of small groups. In such an environment, the availability of first- or second-hand accounts of an event would have been a fairly reliable guide to its true probability, since people would personally experience, or at least know those who do, almost everything which happens in the knowable world. Making a statistical study of the number of deaths from spider bites in the past year would do little to improve the estimate gained from casual reflection using the availability heuristic, and would take much more effort. The heuristic would not be perfect even in the small-group EEA, but would nevertheless be preferable to precise calculation when cognitive resources and time are taken into account.
The world we live in today is very different from the EEA. Not only do we deal with many more people in our daily lives, but also get information through the mass media and see fictional events on television. This wildly different environment makes the availability heuristic a much less reliable guide to assessing risk. When we watch the news on television, the information we are presented with is not a balanced sample of everything that happened that day. We are interested in hearing about major and unusual events, rather than routine drudgery. We thus see aircraft crashing more often than cars, people dying of exotic contagious diseases more often than of heart disease and dogs on surfboards more often than dogs not on surfboards. Using the availability heuristic in this environment biases our estimates of the frequency of these events.
Situations which are easy to imagine come more easily to mind and are thus deemed more likely. For example, in an experiment in which people were asked to imagine spending a three week period suffering from a disease, those who were given easily imaginable symptoms such as muscle-aches and low energy thought themselves much more likely to catch the disease than those who were given difficult to imagine symptoms such as an inflamed liver. Fiction, especially if we see events played out in front of us on television, makes the events more available by effectively doing the work of imagination for us, we are likely to attach a higher probability to events we often encounter in fictional worlds. No rational person would treat fictional events as statistical data about the real world, but that’s what we are unconsciously doing with the availability heuristic.
The way this plays out with moral panics is pretty straightforward. A particularly salient event will focus people upon a particular issue. Media coverage of the issue will then increase to satisfy the public’s newfound curiosity. The increased availability of examples of the phenomenon will upwardly bias our estimates of its probability, and we will see a growing problem where in fact there isn’t one. It’s not hard to find examples which fit this pattern: concerns over particular types of crime at various times; the ‘spate of dog attacks’ New Zealand had a few years ago; and, most destructively, the sudden increase in the risk of terrorism following 9/11.
I think people would react irrationally to these sorts of risks under any set of political institutions, but the sort of democratic totalism we have in western countries today seems particularly apt to produce moral panics through the availability heuristic. Not only does collective choice remove the feedback we require to learn from our mistakes, it also encourages the politicization of issues which can lead to group polarization. Human decision-making is never perfect, but political decision-making is pretty much always worse. This is why I buy the positive basis of behavioural economics, but reject just about all of the purported policy consequences.