Daniel Kahneman (1934–2024), hereafter Danny, through his impact on those who both agreed and disagreed with the evidence and arguments in his monumental body of work, had more impact on the development of behavioural public policy than perhaps any other person. That may be his most profound legacy, although he was not initially particularly supportive of our efforts to establish this Journal, now the primary outlet in the field. In early 2016, after I asked him if he would, in principle, be willing to review the proposal for the Journal for Cambridge University Press, he replied ‘Hi Adam, I like your work, but aren't you setting up a competition with a similar journal recently started in the US? I would not be able to support your proposal without seeing what happens to that journal. Best, Danny.’ Since he wrote that he liked my work, I did not much care whether or not he would write a review, but he was wrong to suggest that we intended to compete. We had meticulously identified a gap – a gap that we relayed to those who may have felt threatened, whom we invited onto our editorial board; in any case, my prospective co-editors – Cass Sunstein, George Akerlof – and I, aligned with the view of letting a thousand flowers bloom. But with or without a written review, the journal owes much to Danny, for it would not exist without him.
Danny's work also had a profound influence on my own research. Given the longevity of his research career and the richness of his scholarly outputs, it is difficult to highlight topics that he worked on by their relative importance, but if I were to pick just two, they would be his collaborations with Amos Tversky: namely, their heuristics and biases programme of the early 1970s, and, of course, prospect theory, first published in 1979 (Tversky and Kahneman, Reference Tversky and Kahneman1974; Kahneman and Tversky, Reference Kahneman and Tversky1979). Although I do not wish to dwell upon old controversies and disagreements, these lines of work are not without their detractors. The seemingly illogical choices and decisions caused by following heuristics, for example, may have been somewhat overplayed, with this work often using highly stylised, contrived thought experiments specifically designed to demonstrate these effects. Moreover, the use of the word ‘bias’ was arguably misleading, with the implication that people were doing something wrong. It downplayed the highly plausible possibility that the heuristics evolved to help us successfully navigate our way through an uncertain world.
In prospect theory, despite them doing perhaps more than anyone else to reveal the cracks in orthodox rational choice theory, Danny and Amos did not appear to stray far from the normative appeal of the latter. They viewed prospect theory as a good description of choice under conditions of risk: that the carriers of ‘value’ for people were gains and losses around a reference point (with losses weighted more than gains of an equivalent magnitude), rather than final ‘assets’, and with outcomes weighed by subjective ‘decision weights’ rather than objective probabilities. But they did not claim that people should process their choices in these ways, and left the lingering underlying impression that they believed that reference dependence, loss aversion and probability weighting caused decision-making mistakes. Their normative outlook remained consequentialist, but their work implied that these systematic effects in descriptive decision-making offered the potential to improve the validity of instruments designed to elicit people's subjective valuation of outcomes.
More than 20 years ago, inspired by the work of Bleichrodt et al., (Reference Bleichrodt, Pinto and Wakker2001), I set out to examine whether the internal consistency of a subjective valuation method, derived from the assumptions of orthodox rational choice theory (namely, the probability equivalence method, otherwise known by health economists as the standard gamble and sometimes used to elicit health state ‘utilities’), could be improved by modifying it with the prospect theory parameters (Oliver, Reference Oliver2003). I found that the internal consistency of the instrument could be substantially improved after adjusting for reference dependence and loss aversion, but not probability weighting, which should not be dismissed lightly, since such (unadjusted) methods are used to inform decisions that are literally a matter of life and death. Unfortunately, there is not a critical mass of people with sufficient behavioural economics interest or expertise working in applied subjective valuation elicitation (in health or in other domains) for these modified techniques to be considered seriously. But the important point is that because the elicitation of subjective value is a descriptive exercise, prospect theory – or at least some aspects of it – can (perhaps) be used to revolutionise the elicitation instruments for the better. However, as intimated earlier, Danny, congruent with orthodox rational choice theory, seemingly remained committed to the normative postulate of subjective value maximisation.
His commitment to outcome optimisation in this sense showed itself again in his later work with (the great) Peter Wakker and Rakesh Sarin on experienced utility, another theme of his that had a big influence on my own research (Kahneman et al., Reference Kahneman, Wakker and Sarin1997; Oliver, Reference Oliver2008). In that work, Danny pioneered what became known as the new economics of happiness, postulating that one ought to guide policy towards that which maximises the integral of real-time (or close to real-time) moments of instant utility, rather than relying on retrospective or prospective assessment of episodes and events, where people tend to neglect duration, and heavily and, it is argued, unduly weight particular ‘moments’ (e.g. the peak and end moments of an episode – the so-called gestalt characteristics). In orthodox rational choice theory, it is assumed that such retrospective and prospective assessments (sometimes, respectively, termed remembered utility and decision utility) do not differ systematically from experienced utility. That this was falsified by Danny and his colleagues led them to again propose that the descriptive methods by which to elicit subjective value needed to be revolutionised, but that the normative postulate of maximising subjective well-being defines the sort of society in which most people wish to live (which is a contestable notion) and thus ought to be retained.
It may have been the case that, with time, Danny became a little more circumspect of the neo-Benthamite approach. It is difficult, on the whole, to pin down his normative beliefs. This difficulty may have been in some sense inevitable, partly because his major focus seemed to be on the descriptive rather than the normative and partly because he appeared to have a highly questioning character, recognising and perhaps appreciating that normative arguments are nuanced and conflicting, which is a mark of a good scholar. That said, even in his later years there are indications that he did not ‘escape’ (if that is the right term) some form of well-being maximisation as the normative ideal. In 2013, for instance, he wrote to me expressing his support for Thaler and Sunstein's (Reference Thaler and Sunstein2008) then topical well-being approach, noting his belief that ‘nudges guide people to do what they already wish they would do – they provide therapy for weakness of the will’ (Daniel Kahneman to Adam Oliver personal correspondence, 31 May 2013).
Danny's character, judging by the accounts of those who knew him, was ‘complex’ (also for want of a better word), and, via email, I did at times get a little taste of that myself. For instance, in 2016, he berated me for my failure to observe what he thought was the most profound implication of prospect theory – the so-called reflection effect – in a study that I later published, stating that I had used inappropriate methods (i.e. within rather than between-subject responses) (Daniel Kahneman to Adam Oliver personal correspondence, 31 December 2016; Oliver, Reference Oliver2018). His message was a little perplexing to me, partly because I used methods not dissimilar to those that he used in his original prospect theory paper, and partly because I reported results that were, by and large, pretty consistent with the reflection effect (incidentally, Danny also told me in that message that in the early 1980s he resolved never again to review a paper that made him angry).
But Danny was generous to me too (indeed, I thought he showed generosity by writing to me at all). For instance, after asking him to write a recommendation for the first book that I edited on behavioural public policy, he initially declined by writing that ‘I would like to help, but I do not know enough about the book to endorse it, and endorsing the motivation for it sounds tepid indeed. It is difficult to endorse motivation without knowing something about the product. I am sorry to be difficult, but I have always been very careful with endorsements’ (Daniel Kahneman to Adam Oliver, personal correspondence, 31 May 2016; Oliver, Reference Oliver2013). Although I was initially a little disappointed with his response, I soon thought that it was the entirely proper thing to do. I would not recommend a book without seeing it either. But later that same day he wrote again to say that he had not noticed the Introduction to the book that I had sent to him, that he had now studied it, that he was impressed by the depth and diversity of the book, and that he would write a recommendation for it (which he did).
Although I communicated with Danny occasionally over email, I met him face-to-face only once, when I sat next to him at a dinner at the LSE soon after he published Thinking, Fast and Slow (Kahneman, Reference Kahneman2012). I was a little awestruck at first, but he was engaged and engaging, modest, warm and kind. His conversation was deep, serious and yet not without humour. We spoke as if we were the only two people at the dinner, and I quickly felt comfortable in his company. For me, it was an uncommon occasion when I was glad that I had met one of my heroes. He struck me as self-aware at that dinner, and in his magnum opus, the publication of which the dinner was, in part, celebrating, admitted that he, with respect to prospect theory, was not immune to what he called theory-induced blindness. This is one small reason among many why everyone in behavioural public policy and indeed in behavioural science more broadly, should have read (or if not, should read) what is, in effect, not a mere trade book but is the intellectual autobiography of the man who shaped our field.
When a great scholar dies, hagiography is, I think, not the correct way to honour them. The correct way is to engage critically with their ideas – to do unto them what we expected of them in their lifetime. The essays that follow are in honour of Danny. Some of the essays are written by people who knew Danny personally; all of them are written by people who, one way or another, were influenced profoundly by him. They are short essays, and they stand on their own, so I will make no effort to summarise them here. My advice to the reader is simply to read them all. Will Danny go down in the pantheon of post-Enlightenment great thinkers? He certainly had a long, rich intellectual life, of which any scholar in history would presumably be proud. He was blessed in that respect. But to me, whether or not one agrees with them, he did not offer the overarching normative social and/or economic vision of, for example, a Smith, Mill, Hayek or Sen (although, again, this lack of a committed vision might actually reflect well on him). But that is just a personal view, and I would not protest if he were invited to sit at their table.
Danny was to attend the launch of my book, A Political Economy of Behavioural Public Policy, in New York in 2023 (Oliver, Reference Oliver2023). Before the event, he caught COVID and could not come, but when I initially invited him to speak at the launch, he replied ‘Hi Adam, I plan to attend, but I should not give formal remarks. I will be 89, an age at which silence is advisable’ (Daniel Kahneman to Adam Oliver personal correspondence, 9 February 2023). Sadly, his silence is now, of course, eternal, and no doubt, many will miss him terribly. But I will not miss him, for two reasons. First, I did not know him well enough personally to be hit hard by his passing and therefore to miss him on a human level. Grief, as the late Queen Elizabeth II used to say, is the price we pay for love. But the second, more important, reason why I will not miss him is because I do not need to. I have his great body of work – his life's work – and that will never leave me.
Open access
