Tuesday, 13 November 2012

Behavioural economics: A Short Introduction

Why would a psychologist win the Nobel Prize in Economics? The answer to this question is possibly the most visible sign of an emerging trend in contemporary economics research. Behavioural economics aims to study the emotional, social, psychological factors in economic decision-making, challenging the largely dominant paradigm that every economic decision is made rationally. But how did that paradigm appear? What are the goals of behavioural economics today? This post aims to provide answers to these questions.

Update: Certain elements of our description of the German Historical School were incorrect and have been rewritten. We'd like to apologise for the inconvenience.

Source: uchicago.edu


Contemporary economics is a  chiefly mathematically embedded social science; modern economists use mathematical (eg. game theory) and statistical tools to construct models of the inextricably complex economic environment around us. However, this by far wasn't always the case. To understand today's status quo, we have to travel back in time until 1759, the publishing of Adam Smith's The Theory of Moral Sentiments.

The roots of modern economics

Adam Smith never declared himself an economist - principally, he was an academic philosopher, at first specialising in the field of moral philosophy (something which today would be called as ethics). In The Theory of Moral Sentiments, the part about economics was only a sub-specialisation of moral philosophy itself, and even in the famous Wealth of Nations, Smith mainly addressed the question of how should governments run their economies rather than how they are doing it. Why the methods used in both books were definitely scientific - they relied on earlier sources, posed refutable statements which they tried to prove -, the framework resembled philosophy more than mathematics; there were no equations, graphs or anything often associated with modern economics.

The first paradigm shift

Throughout the 19th century, the approach inherited from Adam Smith remained the same. Classical economists (and even successors like Karl Marx) based their assumptions on political, philosophical and historical grounds, once again more in the roles of specialised politicians, philosophers, historians, etc. rather than economists.
This approach first came under attack from an unexpected direction - Wilhelm Roscher, the German economist proposed to replace the mainly philosophy-based approach with a history-based one. And even though his approach has mainly fell out of favor, it's legacy was important in the later establishment of modern economic history.
The second line of attack was those of the "modern economists"; starting with Vilfredo Pareto, Alfred Marshall and their contemporaries. The approach they slowly began to formulateaimed to transform economics (formerly, and for a long time dubbed as political economy) into a "proper" science, using the methods of natural sciences, and to expunge psychology from the field. In their age, this was understandable; psychology of their time lacked the quantitative experiments which are often used in psychological research today, hence the shift towards a more numerical analysis of the world around us has indeed produced better models.

20th century

It thus happened that by the 1950's, most economists worked with quantifiable, precise models with testable assumptions - which wasn't a bad thing at all. For example, starting in the 1960s, economists like Jack Treynor and William Sharpe began to develop the capital asset pricing model (CAPM), which (by providing an equation to predict expected returns of an investment based on it's volatility)  revolutionised the way equity investors thought about assessing risks.
What this model didn't take into account however was that a) information asymmetry could arise on markets (eg. in the form of arbitrage), and b) that the homo economicus, the ever-rationally thinking concept of human decision-making was not perfect. Just to cite a few everyday examples:
-If the chances of winning are so small, why would people buy lottery tickets?
-If the chances of influencing the outcome are so small, why do people bother with voting?

Rationalising our irrationality

Starting in the 1970s, the psychologists Daniel Kahneman and Amos Tversky began to conduct a series of experiments to work out a more precise model of human risk assessment. One of their most famous experiments was the following one:
Choose between the following options.
A)There is an 80% chance that you win 4000 dollars, and a 20% chance that you win nothing.
B)You surely win 3000 dollars.
Notice that even though the statistical outcome of the first option is bigger (4000*0.8=3200), people still overwhelmingly chose the smaller, but secure win; 80% chose option B.
Now here is another experiment:
Choose between the following options.
A)There is an 80% chance that you lose 4000 dollars, and a 20% chance that you lose nothing.
B)You surely loose 3000 dollars.

Did you notice that the questions are essentially the same, except for the replacement of wins to losses? However, the majority of respondents, 90% now chose option A.

That may be certainly flabbergasting to some readers, and it also seemingly contradicts the lottery and the voting scenarios as well. However, the 80% win scenario (and it's 80% loss counterpart) were only two points of a larger prospect function - what Y percentage of people takes their chances when faced with an X percentage possibility of winning and losing, respectively?
Kahneman and Tversky did a lot of similar experiments, and they compiled a graph representing the essence of prospect theory, the basis of modern behavioural economics.

Source: Wikimedia
 This expanded set of data now explains every phenomena mentioned above. Put in words:

-At a low probability of winning, people overestimate their chances of winning. (lottery)
-At a high probability of winning, people underestimate their chances of winning. (Experiment 1)
-At a low probability of losing, people overestimate their chances of losing. (The fear of your favourite candidate losing the elections.)
-At a high probability of losing, people underestimate their chances of losing. (Experiment 2)


Prospect theory had a profound influence on the way economists assess human risk-taking. While Amos Tversky has passed away in 1996, his colleague and friend Daniel Kahneman went on to win  the Nobel Prize in Economics in 2002. The application of their theory to the CAPM (bluntly put, to replace the linear function of subjective value in the model with theirs), has made the model more precise, while not complicating it's use substantially. 
Since, behavioural economics has established a successful "research recipe" as well:

1)Take a numerical economic model which assumes rational thinking
2)Demonstrate anomalies of the model.
3)Rule out that those anomalies originate from inevitable real-life distortions, eg. transaction costs.
4)Use the anomalies to create new models of behaviour, taking irrationality into account.

Following this framework, behavioural economics has grown from being a controversial fringe area to a still-controversial, but accepted and quickly developing area of economic research. After 100 years, economists are once again questioning the mainstream paradigm of human behaviour. Their methods have changed and their results improved - but one has to wonder, is there anything new under the sun?


Peter L. Bernstein: Capital Ideas: The Improbable Origins of Modern Wall Street. 
László Mérő: The evolution of money. (Original: A pénz evolúciója) Budapest, Tericum Kiadó, 2007.

No comments:

Post a Comment