About 50 years ago, Daniel Kahneman and Amos Tversky started to take on the neoclassical doctrine of rational, self-interested behaviour. It turned out to be a destructive process. The two psychologists showed how humans systematically behave differently than homo economicus1 would in a range of situations.
Choose a setup, see what the prediction of economic theory is, and prove it wrong. Kahneman and Tversky demonstrated that humans are not perfect Bayesian agents, rather we typically ignore base rates, fail to evaluate scenarios on a stand-alone basis and instead have in mind some (arbitrary) reference point, have strong preferences to maintain the status quo, and consider sunk costs to be important in our individual decision making. Later, young and maverick economists of the time, such as Richard Thaler and later Ernst Fehr showed up on the stage and called into question the foundations of some of the most prominent theorems in economics. Thaler showed how the endowment effect can get in the way of the Coase Theorem and gathered evidence against the Efficient Market Hypothesis. Ernst Fehr did a lot to show that humans care way more about their fellow man than homo economicus does.
Classical economic theorists, especially Milton Friedman and his followers, were not particularly fond of these behavioural scientists swinging their sledgehammers at neoclassical theory. They were probably much more delighted by other experimental economists, most notably Vernon Smith, who showed that economic theory does make good predictions in some settings. Most famously, he showed that markets can lead to efficient allocations, even if the assumption of perfect information is violated (however, that was before Thaler pulled the endowment effect out of his hat).
From today’s perspective, many of the discoveries made by Kahneman, Fehr, and their contemporaries seem to belong in the category “things that my mum knows but economists don’t”. Consider the Ultimatum Game: The proposer can offer the responder a split of some amount of money, say 10€. The responder can then accept the offer in which case both players receive the amount as offered by the first mover. The responder can also reject the offer, in which case both players receive nothing. Go and try it yourself: Ask your mum what she thinks a second mover will do in the Ultimatum if she is confronted with an unfair offer, say a split where she gets 1€ and the proposer get 9€. She will probably give you a better prediction than standard game theory. Within the economic profession, the findings of the first generation of behavioural scientists generated excitement on the one hand, and fervent opposition on the other, which served as a testimony of a discipline so blindly devoted to and so deeply entrenched in the concept of homo economicus that it might have gotten a little bit out of touch with reality. Behavioural Economists jolted the discipline back to reality.
Behavioural Economists did not stop after demolishing of classical economic theory. After the dust had settled, they dropped their sledgehammer and started looking for patterns in the violations of economic theory they had observed in order to develop alternatives to homo economicus. The most notable theory that emerged is Kahneman and Tversky’s Prospect Theory, which embraces some of the most distinct features of human decision-making. It states that humans evaluate a possible scenario with respect to some reference point rather than evaluating its desirability on a stand-alone basis, that humans are loss averse, i.e. loosing 10 euros brings more pain than the gain of 10 euros brings pleasure, and that probabilities are weighted according to a heuristic. Other notable contributions to economic theory are models where agents have limited cognitive capacities and self-control problems (brought forward by Thaler), as well as theories of pro-social behaviour, such as the Fehr-Schmidt model of inequity aversion.
Nowadays Experimental and Behavioural Economics is far from being a field for outsiders and mavericks. Vernon Smith, Richard Thaler and Daniel Kahneman have all received their (well deserved) Nobel prize and I would bet good money that Ernst Fehr will receive his in the years to come. Richard Thaler, who once had trouble publishing his articles in major economic journals, served as president of the American Economic Review in 2015. Behavioural Economists, once the punks of their profession, have become mainstream. The most striking testimony of the acceptance of Behavioural Sciences into the mainstream is probably its growing influence on policy. Since the successful establishment of the Behavioural Insights Team (BIT) under the British government in 2010, nudge units have sprung up in several countries, proving that Behavioural Science can inform more effective and more “human” policies.
Moreover, behavioural economics provides some alternatives to homo economicus and grants valuable insight into individual decision-making and is informing policy around the world. Models with agents of limited self-control can help to understand why people save less for their pensions than they intend to. Fehr-Schmidt preferences can explain behaviour in a range of settings with social interaction. Prospect theory can explain a whole host of phenomena, from the endowment effect to status quo bias to the observation that investors tend to cut winners short and let losses run (instead of doing the reverse).
Paradoxically, a major insight from behavioural economics is explaining and understanding why some of the most prominent theorems of economic theory do have considerable explanatory power even though their assumptions are often violated. Consider for instance the First Welfare Theorem. It states that a free market will lead to a Pareto efficient allocation under some assumptions, such as no transaction costs, completeness of contracts, and perfect information. In reality, search costs can be high, contacts are usually far from complete, and asymmetry of information is rather the norm than tan exception. Still, everyday economic activity goes ahead relatively smoothly and this is not despite but due to the fact that humans are not selfish maximizers of utility. As Kenneth Arrow once put it, every economic transaction requires a modicum of trust — and countries, in which citizens do not have this minimum of trust in the good intend of each other tend to do very poorly economically. This, of course, can only be understood by embracing Behavioural Theory.
Behavioural Economics has, however, failed to provide a better understanding of the greater picture, of how economies work as a whole, let alone a theory that could replace that of homo economicus. In this regard, the often-voiced criticism that Behavioural Economics is resembling a hodgepodge of models rather than a coherent and complete theory is probably both right and wrong. The obvious advantage of homo economicus is that its basic defining concepts can be adapted and applied to a broad range of situations. Such models can then be expanded and refined, practically ad infinitum (and sometimes ad absurdum). This is hardly possible for behavioural theory, where there only exist few theories that cover a wide range of settings. However, the criticism here is misguided in the sense that human behaviour depends very much on specific circumstances. The undifferentiated application of a “one size fits all” approach can therefore easily turn into a “one size fits none” debacle. The criticism is, however, right in the sense that behavioural theory does not (yet) allow for a comprehensive theory of the economy as a whole, from the single individual to the world economy.
The onus will be on the future generation of Behavioural Economists to translate the insights gained into individual decision-making into a better understanding of the functioning of the economy as whole through the creation of a comprehensive, behaviourally-inspired theory. It has however, proven difficult to weave the different bits and pieces into one such theory, simply because such a theory could never account for all the quirks of human behaviour. An alternative approach might be that of a more hybrid theory, one that is indeed composed of various bits and pieces. This option might be less elegant than a complete and comprehensive theory. But it might also be the way forward.
by Moritz Loewenfeld
The TSEconomist 