head-607480_640.jpg
VoxEU Column Frontiers of economic research

Behavioural economics: What we know and how it could be mainstreamed

Behavioural economics has identified phenomena that standard models could not explain. But its critics warn that it is becoming little more than a ‘pile of quirks’. This column argues that the future development of behavioural economics should focus on a streamlining process that will clarify core issues, fill conceptual gaps, and create tractable models. Behavioural models will only become a coherent alternative to homo economicus if this process occurs.

At the turn of the century, behavioural economics was a hot topic. It identified phenomena that standard models could not explain.

For example, the bursting of the dot-com bubble in 2000 came at the end of a period in which standard indicators, such as price-earnings ratios, were completely out of line with the predictions of efficient markets as depicted by what we will refer to as conventional economics.

More generally, in the words of Mullainathan and Thaler (2001): ‘The standard economic model of human behavior includes three unrealistic traits – unbounded rationality, unbounded willpower, and unbounded selfishness – all of which behavioral economics modifies.’ Of course, this statement is itself debatable, as it all depends on the specific definition of rationality that we use. But, if we assume ‘rationality’ as the one predicated by conventional economics (if I prefer A to B and B to C, then I must prefer A to C), the statement is spot on.

That’s why in a 1955 paper on decision-making, Nobel laureate Herbert Simon described the computations implied by the typical conventional economics optimisation process, only to conclude that:

My first empirical proposition is that there is a complete lack of evidence that, in actual human choice situations of any complexity, these computations can be, or are in fact, performed” (Simon 1955).

There are many other tenets of conventional economics that behavioural economics found untrue in real-world decision-making.

Consider the leading character in conventional economics, homo economicus. He is motivated by total selfishness, decides in isolation, has stable preferences, is endowed with advanced computing capabilities, and is equally averse to upside and downside volatility of prices.

Behavioural economics has shown that real world people behave differently.

In their decisions, they often give a role to altruistic considerations and assign a value to reciprocity and fairness. The two-player ultimatum game is an example, in which the first player, who has an amount of money allocated for the game, decides on how the money should be divided. The second player can accept or reject his or her share. If the second player refuses, both players receive nothing. When the ultimatum game was played in social groups in many countries, the results showed that offers less than 20% were often rejected (Camerer 2003).

Figure 1 The ultimatum game

Copyright: Saturday Morning Breakfast Cereal. Reproduced with permission.

As first formally shown by Kahneman and Tversky (1981, 1984), people are also loss-averse: losing €100 provokes negative emotions that are much more acute than the pleasure from gaining an equivalent amount. We change our preferences: if I prefer Italian food, contrary to what conventional economics predicts, I do not always choose an Italian restaurant. We are influenced by framing: we buy the ground beef labelled ‘75% lean’ rather than the one labelled ‘25% fat’. And so on.

Greater realism was certainly needed after decades in which economic models portrayed the individual as an algid homo economicus. That’s why expectations for behavioural economics were so great, as we discuss in a forthcoming book (Chang and Ghisellini 2018).

Unfortunately, behavioural economics hasn’t made significant progress in the last ten years. The most successful school of thought continues to be the one for which the shorthand could be the ‘American school’ and whose leaders are Daniel Kahneman and Richard Thaler.

The American school focuses on the cognitive biases that distort decision-making processes, and on the systems (‘System 1’ and ‘System 2’) to which the mind responds. The two systems are the leading characters in Kahneman's bestseller published in 2013. System 1 is quick, largely based on intuition, and is the first to be called into action in decision-making. System 2 is slower, analytical (and more powerful). It is, however, more costly in terms of elaboration, and so is activated only when it is needed (you can’t fill in a tax form using only System 1).

As an economic policy complement, the advocates of ‘nudges’, inspired by Thaler and Sunstein (2008), contend that policymakers should set up ‘choice architectures’ so that people will avoid falling prey to these biases and will make choices that are in their best self-interest.

There are critics of the American school, notably the ‘German school’, whose leader is Gerd Gigerenzer. Since his first book in 1999, Gigerenzer has argued that many of the biases detected by behavioural economists are, in fact, superior choice tools if one takes into account the complex environment in which people operate (they are ‘ecologically rational’).

Others would stick to evolutionary models that justify almost all apparently ‘irrational’ choices by explaining them as the legacy of behavioural rules adopted by our ancestors when dealing with wild environments. In the literature, there are now countless behavioural factors, each of which has been studied in isolation. Therefore, in the last few years, we can argue that behavioural economics has been growing ‘horizontally’.

No wonder then that the perception of behavioural economics by the general public runs the risk of being distorted. On 24 March 2017, Wikipedia’s list of cognitive biases listed about 200 different biases. Alongside established biases such as the gambler’s fallacy (if in the lottery number 22 hasn’t come up for six months, it will come up next!), the list includes such ‘biases’ as anthropomorphism (the tendency to credit animals and objects with human features) and the ‘Google effect’ (by which it is easier to forget information that is easily retrievable online).

To be fair, the relevance of anthropomorphism to behavioural economics is not entirely clear. The impression is that Wikipedia’s list is a bit of a hodgepodge of everything that comes to mind with a vague psychological flair.

As House (2014) writes:

Today, it seems like behavioural economics has slowed down somewhat. For whatever reason, the flood of behavioural economists we were anticipating 10 years ago never really materialised… Behavioural economics won’t get very far if it ends up being just a pile of quirks.”

Do as overweight people do: Go on a diet

The success of conventional economics in recent decades has been largely due to its generality and tractability, whereas its crisis has been mainly fuelled by its failure in forecasting the behaviour of real-world economic agents.

To become the ‘normal’ economics, behavioural economic theories should outperform conventional economics, and meet the three preconditions – accuracy of predictions; generality; and tractability – identified by Stigler (1965).

But how to get there?

At the moment, behavioural economics suffers from confusing definitions, unanswered questions and conceptual gaps that need to be filled. It is reminiscent of a person who needs a diet both for a detox and a weight loss.

We need a ‘vertical’ streamlining process that will:

  • Clarify core issues: What is the ‘correct’ definition of rationality? When an alleged bias, such as loss aversion, is found to be part of human nature as designed by evolutionary processes, can it still be considered a ‘distortion’?
  • Fill existing conceptual gaps: behavioural economics deals abundantly with biases, but how are expectations formed? Are expectations ‘rational’?
  • And most importantly, generate a tractable reduced form for behavioural economics models. Today we have hundreds of different behavioural factors. They should eventually be translated into a smaller set of primitive factors.

In order to establish behavioural economics' status, we hardly need the ‘discovery’ of yet another behavioural bias. At this stage, we need parsimony and an effort to synthesise what already exists into a general mainstream model as an alternative to conventional models.

As an example of the kind of work that should (and could) be done, consider again the quote by Mullainathan and Thaler (2001):

"The standard economic model of human behavior includes three unrealistic traits – unbounded rationality, unbounded willpower, and unbounded selfishness – all of which behavioral economics modifies."

These three unrealistic traits have typically been dealt with by behavioural economics one at a time, each in isolation from the others, partly because the conceptual reference remains utility as defined in conventional economics (wealth maximisation).

But what if utility is associated instead with emotional wellbeing? Assume that I am an investor, and my utility function is such that it derives more from feeling good socially (in the family, with colleagues, and so on) than from wealth maximisation.

The investment implications of such a utility function are likely to generate what is considered the biased behaviour known as herding, by which I replicate the investment choices of people in my community. Would that be irrational? Not if I subscribed to Herbert Simon’s definition of rationality:

Behaviour is rational in so far as it selects alternatives that are conductive to the achievement of the previously selected goals”.

Following this approach, the statement by Mullainathan and Thaler could be reformulated as: The economic model of human behaviour is based on a definition of rationality, which in turn is based on consistency between actions and goals. This is much more compact, much more tractable, and positive, rather than negative. This is what we need if behavioural economics is to become something more than a set of exceptions to the rule.

References

Camerer, C (2003), Behavioral Game Theory: Experiments in Strategic Interaction, Princeton University Press.

Chang, B, and F Ghisellini (2018), Behavioural Economics: An Unfinished Business, forthcoming.

Gigerenzer, G, P Todd and the ABC Research Group (1999), Simple Heuristics that Make us Smart, Oxford University Press.

House, C (2014), “Is Behavioral Economics the Past or the Future?”, Orderstatistic blog, 28 February.

Kahneman, D (2013), Thinking, Fast and Slow, Farrar, Strauss and Giroux.

Kahneman, D, and A Tversky (1981), “The Framing of Decisions and the Psychology of Choice”, Science 211(4481): 453-458.

Kahneman, D, and A Tversky (1984), “Choices, Values, and Frames”, American Psychologist 39(4): 341-350.

Mullainathan, S, and R Thaler (2001), “Behavioral Economics”, International Encyclopedia of the Social and Behavioral Sciences.

Simon, H S (1945), Administrative Behavior, A Study of Decision-making Processes in Administrative Organization, Free Press.

Simon , H S (1955), “A Behavioural Model of Rational Choice”, Quarterly Journal of Economics 69: 99-118.

Stigler, G (1965), “The Development of Utility Theory”, in G Stigler, Essays in the History of Economics, Chicago University Press.

Thaler, R and C Sunstein (2008), Nudge: Improving Decisions About Health, Wealth, and Happiness, Penguin.

5,249 Reads