Monday 3 April 2017

What is Unrealistic Optimism?

This post is the final one in our series summarizing the contributions to the special issue on unrealistic optimism 'Unrealistic Optimism -Its nature, causes and effects'. The paper by Anneli Jefferson, Lisa Bortolotti and Bojana Kuzmanovic looks at the nature of unrealistically optimistic cognitions and the extent to which they are irrational.

Anneli Jefferson

We know that people have a tendency to expect that their future will be better than that of others or better than seems likely on an objective measure of probability. But are they really expressing a belief that the future will be good, or should we see these expressions of optimism as hopes or possibly even just expression of desires for the future? Maybe when I say ‘My marriage has an 85% likelihood of lasting ‘til death do us part’’, what I am actually saying is ‘I really, really want my marriage to last.’ If what is expressed is a desire rather than a belief, we do not need to worry that we are systematically mistaken in our beliefs in the future and that our expectations for our future are insufficiently sensitive to the evidence we have for what is likely to happen. In the paper, we argue that expressions of unrealistic optimism are indeed what they seem to be on the surface, beliefs about what is likely to occur. The fact that optimistic expectations are frequently not well supported by the evidence is a feature that they share with many other beliefs, as we humans are not ideally rational in our belief formation.

Lisa Bortolotti

By definition, unrealistic optimism is a phenomenon that shows us to be insufficiently in touch with reality. However, establishing that we are in fact making an error when assessing the likelihood of future outcomes is surprisingly difficult. In some cases, whether an expectation is correct or not can only be established post factum. Only at the end of the Euro 2016 could we say that Ronaldo’s belief that Portugal would win the European cup had been correct (if indeed he had this belief). Things are more complicated if what we know is that Ronaldo believed that Portugal had a 95% likelihood of winning the European cup. Is this belief validated by the fact that Portugal did win? Not necessarily, as his likelihood estimate may still have been too high given some objective measure of likelihood. Furthermore, it cannot be the case that probabilistic risk estimates are proven or disproven by later outcomes. Otherwise, any risk estimate which isn’t either 0 or 1 will automatically be incorrect, it is just impossible to say whether the error lay in being too optimistic or pessimistic before the actual outcome ensues.

Bojana Kuzmanovic

But the question of whether an individual’s optimistic beliefs are false is in many ways less pressing than the question whether the individual is justified in holding that belief given the evidence available to them. Are unrealistically optimistic beliefs epistemically irrational because they do not take into account available evidence either when the individual forms the belief or when they maintain their belief?

We argue that optimistically biased belief updating provides evidence of motivated cognition. In optimistically biased belief updating people take information on board selectively depending on whether the information promotes a positive outlook or not. In a number of studies (eg Sharot et al 2011, Kuzmanovic et al 2015), it has been shown that when individuals make predictions regarding the likelihood of specific future events and are given base rate information for these events, they tend to adjust their expectations to take on board this base rate information if doing so promotes a more positive outlook. If taking on board this information would lead to a worse outlook for one’s own future, people tend to neglect the information  completely or to adjust their outlook to a lesser extent. Biased belief updating is more pronounced when individuals make judgments regarding their own future than that of others. We conclude that this biased use of information does constitute an instance of epistemic irrationality.

When looking at further factors that influence the tendency to be unrealistically optimistic, we find that individuals exhibit less optimism when they are asked to justify their expectations and that suffering adversity decreases optimism in the short term, as does mild depression. Unrealistic optimism allows us to have a rosier view of our own future than is justified by the evidence available to us, but this outlook is constrained by external pressures which push us towards a more realistic outlook.

While this post marks the end of our blogposts on the special issue, our work on unrealistic optimism continues, so watch out for further posts on this topic!


  1. Proposition: an extension of this analysis so as to account for people selecting and updating their goals reduces and perhaps removes the amount of apparent irrationality exhibited.

    Model: A person is in present state A, and wishes to achieve improved future state A'. Person concludes action X as the means to achieve this.

    However new information (or better reasoning) B becomes available, and goal A' now looks unlikely. Person can choose from various actions, from complete resignation to a new goal. Say person chooses new goal B', not as good as A', but better than current state A, now to be achieved through action Y.

    This can of course be repeated for new information, learning along the way, etc.

    It follows that any person seeking an improved future state will always appear optimistic because on the aggregate we know in practice not everyone reaches an improved desired state (whatever that might be), and there is a high failure rate. Yet if people consistently aim for improved states A' (and failing that then B' etc), rather than resigning, the aggregate will always appear unrealistically optimistic.

    It may be worth considering what a rationally optimistic population might look like. In order not to observe irrational optimism, we would need to observe, on the aggregate, goals which are almost certainly within the reach of people. They must be almost certainly within reach because information uncertainty and reasoning constraints prevent any more distant goal from being achieved reliably. What might these near-certain goals look like? They might look like Z' – hardly at all any improvement on current state A.

    Why should people prefer goal Z'? On the positive side, it is very low risk and almost certain. However is it actually rational to aim for Z' rather than A'? Is it worth the effort? Would resignation be better? Given uncertainties, a person may rationally judge that they (currently) have no reason to believe A' is not worth the effort of pursuing on a effort/reward basis. Again given uncertainties and limitations, the information required in order to more deeply judge how realistic A' is might only be discoverable in the pursuit of A'. It is rational then to prefer A' over Z' in the first instance. Then B' in the second, and Z' lastly.

    Irrational optimism is on this basis an observational artifact caused by (non-resigned) people selecting rational goals in a highly uncertain landscape of choices, rather than motivated reasoning.


    1. Thanks for your response and apologies for the delayed reaction. I think you make some very interesting comments on rational goal setting. While this is a related area, it doesn't directly speak to the epistemic rationality of expectations we discuss. It is feasible that we can rationally pursue goals which we realise to be very ambitious and unlikely to be realied. However, the findings we discuss concern the epistemic rationality, i.e. the sensitivity to evidence of beliefs about what will happen. Selectively taking into account information when it is desirable is epistemically irrational insofar as it makes biased use of available information. However, this is compatible with the assumption that adopting these goals may nevertheless be instrumentally rational, even if they are ambitious. In fact, there is literature that claims that the optimism bias and other positive illusions have this kind of instrumental rationality because they contribute to attempts to achieve certain goals.



Comments are moderated.