The case against rationality
For economists and behavioural scientists trying to understand and describe human behaviour, the word ‘rationality’ is a hinderance rather than a help. I recommend it be dropped.
What is rationality?
“ ‘When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean — neither more nor less.
‘The question is,’ said Alice, ‘whether you can make words mean different things.’
‘The question is, said Humpty Dumpty, ‘which is to be master — that’s all.’ ”
Through the Looking-Glass, Lewis Carroll
The Oxford English Dictionary definition of ‘rationality’ is: “The quality of being able to think sensibly or logically” so that decisions made are “based on or in accordance with reason or logic”.
By this definition, there are two interpretations one can make: rational can mean to think and act sensibly or it can mean to think and act logically. While logic can be clearly defined objectively, sensible is a subjective term and tied up therefore with subjective value judgements. Unfortunately, in my view, all too often rationality has been defined by whomever is using it to describe the behaviour that they see as sensible (e.g. consistent, optimal by their world view, or self-interested), without thought or sufficient scrutiny to how their value system is embedded in that definition. Despite the ambition to be an objective science, economics has a subjective value system at its core. Furthermore, the variety of definitions ‘rational’ has makes it a difficult term for communicating a shared meaning. This is deeply important. Those who are the master of the definition impose their value system on society via their economic models. I believe this suggests that we should deconstruct the terms (such as ‘rational’) and models (such as homo economicus) that are used in economics to describe human behaviour and rebuild them with the evidence from psychology, neuroscience, anthropology, sociology, evolutionary biology and other lenses in mind. Dropping the word ‘rational’ (something Daniel Kahneman advocates) and being specific by using ‘logical’ and ‘sensible’ would be a good start.
Rational Choice Theory and how it has shifted human behaviour
Rational choice perspective makes three key assumptions about how humans make decisions. The main principles are that we have coherent and stable preferences, we consider the full range of information relevant to the decision at hand, and we use that information to calculate the option that best fulfils our preferences, in order to maximise our own individual utility (happiness). In order to fit into economic mathematical equations, over the 19th and 20th century the individual human decision agent has been assigned characteristics ever further from real life, leaving us with a picture of a man — homo economicus — who, in the perfected image of rational choice theory, makes consistent decisions over time, has infinite willpower and considers all information when making a decision (Hallsworth and Kirkman, 2020). To any human with lived experience, these assumptions are clearly false and the early years of Behavioural Science give plentiful evidence for this. We have bounded willpower (e.g. hyperbolic discounting, myopia, present bias), our preferences are fluid and not always self-interested (e.g. ultimatum game), and we have an imperfect decision making system with a series of biases and heuristics (e.g. availability bias, take the best heuristic) that lead us to use some information with greater weight than other information when we make a decision (with only a fraction of all information available to us). The upshot is that humans, in real life, are most of the time making illogical, inconsistent and (probably) suboptimal decisions.
However, as rational choice theory — and its specific (subjective) definition of rationality tied up within it — has been the dominant model of human decision making in economics through the 20th century, it has had a profound impact on how people think, act and behave in real life. This is summed up by economist Robert Frank, who argues that ‘our beliefs about human nature help shape human nature itself’. This can be understood through the mechanism of social norms and identities. By using homo economicus in economic models, incentives have been created to encourage people to be like homo economicus. As more people began to act more like homo economicus, and were rewarded for doing so with status and wealth, this (false) descriptive social norm became an injunctive social norm, tipping much of society towards this model of behaviour. This is expressed both through how institutions are shaped and in the stories we tell ourselves about how to lead a good life. This explains why, in western and individualistic countries, many people make decisions through a self-interested lens and how cultural ideals of being logical over emotional have taken precedence.
We can see this phenomenon, where a descriptive model shifts the behaviour of those who believe in it, in the example of the Black-Scholes model, which was adopted in the Chicago Board Options Exchange (CBOE), 1973. The predictions made by the model when it was first introduced were 30–40% wide of the mark. However, over time — while the model was not changed — the behaviour of the people within the exchange shifted towards it so after only a few years, the model predicted within 2% of real behaviour. What started out as a false descriptive model became an injunctive social norm (saying “this is how you should behave”), which in turn shifted the real behaviours of people in the system to be more like the original described model. This gave proponents of the model the opportunity to (falsely) say the model has true descriptive validity (MacKenzie and Millo, 2003). It is very plausible the same has happened at a much wider scale with rational choice theory. At first those who contributed to it recognised it as being far from accurate as a descriptive model. For example, in 1844 John Stuart Mill said political economy ‘does not treat the whole of man’s nature’ and in the 1920s, Frank Knight said this is a model of a creature ‘who treats other humans as if they were slot machines’. But as these models gained greater influence this insight fell away, with Milton Friedman in the 1960s arguing that since in real life people behaved ‘as if’ they were making self-interested choices that rational choice theory is a valid model. The power of these broad descriptive models to shape human behaviour should not be underestimated. As Mary Morgan, economic historian, puts it: rational economic man came to define rationality (what is sensible) and turned into ‘a normative model of behaviour for real economic actors to follow’ (summarised from Raworth 2017).
This is a problem because, as Richard Layard stated at a talk at the RSA in 2020, our society is more stressed than ever before. Economically, we live in a society that is highly competitive and not very cooperative. Politically, society is fractured with polarised views and rising inequality, increasingly so since the 2008 crash in a financial system built off the back of rational economic choice theory. If Behavioural Science is to contribute to people and societies making decisions that lead to their best outcomes (and higher quality of life), it should put forward a new descriptive model of human decision making so that we can drop Rational Choice Theory and ‘rationality’ altogether.
