Economics as a discipline is a victim of huge misunderstanding. We are seen to be obsessed about utility maximization, profit maximization, efficiency and equilibria (to name a few among the many words in the economist’s vocabulary that appear to degenerate human behavior into some egotistical mathematical equation!). These perceptions are mostly true; however, the misunderstanding lies within the economist’s relationship with the homo economicus. Most people would think that economists are in love with the rational but fictional economic man. However the truth is that economists are in some twisted love-hate relationship with this agent.
Economists love the rational agent, because it gives us an intuitive and powerful basis on which we can describe behavior. Economists hate the rational agent because he falls short of reality and is constantly under a siege of criticism from other economists and/or other stakeholders of economic theory. How can economists alleviate the shortfalls of the rational agent, so he can better represent reality? Behavioral economics.
Behavioral economics attempts to integrate insights from psychology with insights from neo-classical economic theory. Although the term ‘behavioral economics’ is quite unfamiliar to many, one must realize that economics and psychology have never been completely unrelated (a classic example of the relationship between economics and psychology would be Adam Smith’s “The Theory of Moral Sentiments”).
In more detail, the theory modeling behavior in economics has these characteristics: the agent maximizes some utility function using (in some appropriate way) all information that is available to him, he is affected only by his own payoffs and his decisions are temporally consistent.
A quick reality check should raise some eyebrows about these assumptions. For example, don’t we all sign up to the gym, thinking we will exercise every day, eventually offsetting the hundreds of dollars we spend on membership? Don’t we procrastinate on Facebook for way too long, instead of writing that article due tomorrow? (sounds strangely familiar!).
An example would be the economics of altruism, which to some extent ameliorates the misunderstandings the discipline of economics faces. Altruism at a first glance seems somewhat misplaced in the aforementioned economist’s vocabulary, but in theory there is nothing that stops an agent’s utility function to be a function of another agent’s utility. In other words, a perfectly rational agent can also be altruistic, he can be motivated to give and donate, and while doing this he will be maximizing his own utility! This relatively minor adjustment to a rational agent’s utility function helps us (partly) explain the dynamics of charitable giving and altruism that would have been largely ignored in the neo-classical economics literature despite its obvious presence in reality.
Of the many examples from the field and laboratory that provide a case for deviation from the standard modeling assumptions there are issues such as: optimism and over confidence in beliefs, time-inconsistent decision paths and menu effects.
Behavioral economics is like adding the bells and whistles to your ordinary economic model. These bells and whistles help bring economics one step (perhaps quite a small step) closer to reality, and hopefully it will provide the rational agent the emotions and realism it has so often been accused of lacking.
Interesting topic: I’ve always thought that People were rational, just that each person puts a value on different things. If it were possible to find everything that drives a certain someone we could then map out their utility curves/predict their actions, but sadly that is probably too impractical, saying nothing about predicting the actions of the masses.
Behavioural economics is cool because it tries to empirically test for those factors, but from my understanding it’s also too specific/limiting to situations and hypothesis we can experiment on.
I really like the ideas in this article. Mainly because I feel like maximising utility has become a sort of definition for rational behaviour, and that quite often maximising utility is degraded to a matter of simply maximising profits, or other tangible benefits. Eventually it feels like economics teaches people that you aren’t rewarded if you don’t act on perceived benefits, and we often take a conventional meaning of ‘perceived benefits’ to be something like profits/returns.
But everyone places a different value on what gives them happiness, and I think it’s important to know that what makes you happy doesn’t have to be something material. The problem is, this is so often assumed that economics tends to be interpreted as a model of people’s behaviour according to material happiness. Which I suppose is the easiest way to generalise behaviour and incentives, sadly.
I think there is no misunderstanding per se. As much as people might not understand why economists need to start any model with assumptions about rational agents and the like, most people criticise the fact that economics (since early late 19th/early 20th century) tries to model human behaviour mathematically. As you say, behavioural economics brings in a fresh perspective as to why economic agents act they way they do, however as much as it might add “bells and whistles” to an economic model, the underlying model is intrinsically flawed by point of trying to predict human behaviour in relation to economic incentives.
Charlie,
Although all human behaviour can’t be predicted simply in relation to economic incentives, I think it’s a good place to start, because it’s a legitimate generalisation. Of course the model will be flawed, which is where behavioural economics steps in – which might also be flawed in its own right. I guess any attempt at modelling behaviour will have flaws, because no model can capture the uniqueness of every individual’s incentives and motivations.
I think economics does a good job (and increasingly better with time) of modelling behaviour.
but of course, given certain assumptions.
Firstly, I think economic theory (minus econometrics and unless calibrated to real parameters) was never meant to “predict” human behaviour. And also, it is important to realise that “utility” is simply a mapping of preferences on to the real number space, hence being able to map out ones “ultimate” utility function is not only a matter of finding out things that drive this person but these things satisfying the relevant axioms that allow us to build a function to represent these preferences.
I believe the idea is that these fictional agents and situations (who are unreal but we hope to be reasonable) give us some point in a spectrum of people and situations ,for real people to compare themselves to, and of course the positive analysis that we can draw from good theory should be correlated with what we observe in the real world.
and as for behavioural economics, it attempts to fill the gap between the lack of correlation between what economic theory claims should happen and what really does happen. And as far as I can see , using maths to do this, if you can, is absolutely fine.
You explain utility functions and their purpose very nicely, however it does not change the fact that it is still trying to enumerate something that does not lend itself perfectly for that. I agree with you mostly; I was just trying to point out that (in my opinion) the misunderstandings and misconceptions of other people outside economics about rational agents stem from scepticism towards mathematical economics. As far as this goes, I agree that the simplifying language of mathematics is the best tool economists have to come close to explaining economic behaviour.
Anecdotally at least I find laypersons often equate economic rationality with self-interest but I don’t think all the blame should be placed on them, it doesn’t help when economists speaking to a laypersons audience (often implicitly) equating rationality = self-interest.
I agree that behavioural economics adds ‘bells and whistles’ but those bells and whistles aren’t always necessary. I think assuming self-interest is sufficient for a lot of cases. If there is inconsistency between data and the model then should delve deeper into the tool box.