For years, I’ve been warning investors (and myself) about the countless errors of judgment, or cognitive biases, to which we are all prone. But might I have been wrong, and that what look like mistakes are in fact ways of helping us to reach better decisions? I ask because of a recent paper by Bjoern Hartig of Royal Holloway University in London.
He considers sunk costs. These are costs we’ve paid in the past and cannot recoup. Logically, they should not influence our future decisions because they are by definition past and gone. But they often do. For example, if you’ve bought expensive tickets for the theatre you’ll drag yourself there even if you don’t fancy going on the night because if you stay at home you’ll feel as if you’ve wasted your money. Similarly, investors hold onto stocks they would otherwise dump because they hope to recoup their losses – oblivious to the fact the price they paid for the stock does not affect its future behaviour. More seriously, companies and governments sometimes plough on with failing investment projects because they have already incurred big costs. And even more seriously still, governments have sometimes continued wars so that those who already died did not do so in vain.
All these are examples of what economists call the sunk cost fallacy – allowing sunk costs to influence decisions when they shouldn’t.
But this might not be a fallacy at all in some cases, says Dr Hartig. Instead, it can offset impatience and weakness of will – what economists’ call present bias – and so cause us to stick with good long-term investments.
For example, if you’ve already paid for gym membership, you’ll get out of bed and go to get your money’s worth. That looks like the sunk cost fallacy. But in fact, the 'fallacy' makes you do what you planned to do when you bought membership – keep fit. It offsets your weak will.
What we have here is an example of how one error of judgment can offset another and so produce the outcome we would have chosen had we been rational and disciplined. This is by no means an isolated example. Here are some others:
- We tend to over-value things simply because we own them. This encourages us to hold onto those assets. In itself, this is irrational – and it can be expensive insofar as it causes us to hold onto poorly performing shares. But it can offset another irrationality – the disposition effect, which leads us to sell winning stocks too soon. The endowment effect can therefore cause us to run our winners and so benefit from momentum.
- Our tendency to overreact to news can offset our tendency to cleave too strongly to our prejudices. We might therefore end up behaving like good Bayesians - updating our beliefs in light of new evidence – simply because our irrationalities cancel out.
These examples tell us that what look in isolation like errors of judgment might in fact be unconscious ways of correcting other errors.
In fact, something else tells us this – the very fact that such errors are so common. Their ubiquity tells us that they have not been eliminated by years of selection and learning, which alerts us to the possibility that they do in fact serve a useful purpose.
Of course, this is not to say that they are always useful, merely that they are good adaptations to particular environments but not others. Paying attention to sunk costs, for example, is a good adaptation to an environment in which we have a present bias, but not to other environments – just as fins and gills are good adaptations to living in water but bad ones to living on land.
It’s very easy to spot what looks like irrational behaviour – at least in other people rather than ourselves. But we should ask: is this apparent irrationality in fact useful (if not deliberate) in this particular context? Irrational behaviour might just be rational behaviour in the wrong environment.