Should we discard the assumption of rationality?

Ever since I was introduced to economics, I was always troubled by one of its fundamental assumptions – the notion of rationality. Many bestselling economics books, such as Dan Ariely’s Predictably Irrational, and my once-favourite (yes, I finally found a new favourite!) book Freakonomics deal with this assumption. They discuss how restrictive and unrealistic it is, and how behavioural models provide much more accurate representation of human behaviour. To some extent, I agree with their conclusion, but in this post, I am going to argue that the rationality assumption is not as useless as many, especially A-level students, perceive it to be.

Rationality is often misunderstood.

First of all, the idea of humans being rational beings is often misunderstood by students. By default, when one thinks of rationality in economics, one often thinks about the buzzwords of selfishness and utility maximising. Well, the good news is, it’s half right. Rationality is about maximising utility, but it does not necessarily imply selfishness. When we behave rationally, we are expected to maximise our utility based on some sort of objective. For example, in a simple utility model, one can think of two ‘goods’ that we typically consume, leisure and consumption. In this example, we will be attempting to maximise our utility, which depends on the amount of leisure and consumption that we get, and this is constrained by the number of hours we have per day and our wage rate per hour. However, there is nothing intrinsic in the model that prevents us from including, say, altruism into the utility function. Someone who is altruistic and finds joy in helping others could be attempting to maximise an objective that involves personal consumption and donations to a charity. In both of these cases, everyone is behaving rationally.

Rationality is just one of the many assumptions that economists consider.

It may be the most commonly cited assumption, but it certainly is not the only one that economists consider. Behavioural based models and rule based models are often also considered by economists and other social scientists. Through experiments in behavioural economics and neurology, social scientists have documented many cases of systematic deviations from rationality. One easily explicable bias is the status quo bias which finds that people tend to stick to default options. An example is organ donation after death. Countries such as Singapore that has a default of ‘donate’ and an opt-out option see a much larger proportion of the population donating (often more than 80% choosing to donate) when compared to countries such as the UK that have the default option of ‘not donating’ and an opt-in form (only about 20% chooses to donate). This is strong evidence of a systemic bias and irrationality. Some other examples of such biases include base rate bias, hyperbolic discounting and prospect theory.

Although models involving as behavioural and rule based agents are highly useful and their experimental findings are being increasingly incorporated into mainstream economic theory in recent years, they have not come to replace rational actor models completely. This is because rationality can often be a useful assumption that produces illuminating results.

Rationality is a convenient assumption.

  1. Rationality based models are a good benchmark because they represent the ideal case. It is a fact that people can learn from their mistakes, and in many cases they can correct them and become more rational over time.

  2. Rational actor models are often easy to compute as they can be expressed mathematically. Frequently, solving for the optimum just involves basic differentiation. Also, they tend to produce unique solutions.

  3. Finally, if mistakes are random, their effects may actually cancel out and the average behaviour of society could be somewhat largely rational.

In conclusion, while intuitively, rationality may seem like a bad assumption to make, it is not at all the case. We can often get illuminating results from the rational actor model and these results can tell us more about how certain ideal scenarios might look like. For example, under the assumption of rationality, we can often arrive at outcomes like allocative efficiency under a perfectly competitive free market. We can then relax the assumption of rationality and model the different behavioural biases based on what kind of scenario we are studying. This way, we can see how these behavioural biases might influence the outcomes. It might surprise you to find out that actually in many competitive markets, it does not matter even if behavioural deviations exist! Even if actors chose randomly, we might still have outcomes extremely close to the rational-actor model – something that we would not have been able to find out if we did not model rationality. Unfortunately, the details involve a fair bit of math which I will leave you to figure out by reading up more from books if you are interested.

Till next time, dream economics.