A sideways look at economics
[As told to Erik Britton]
It has come to my attention that some self-styled ‘economists’ are of the opinion that I am redundant. That I, Homo economicus, the embodiment of economic rationalism, have served my purpose, if indeed I was ever fit for that purpose. That positive economics, a branch of positive science which I personify, should be ditched, along with the assumption of rational behaviour on the part of the human species. Hard to imagine, I know; but if you don’t believe me, read the opinions of a Fathom employee three weeks ago. Before such dangerous nonsense starts to gain traction, allow me to correct some misapprehensions.
First of all, I, Homo economicus, do not exist; so any personal animus aimed at me is wasted. I am a projection into one persona of all the human individuals that have ever lived, will ever live or could ever live. All that is or could be human is in me. Please do not be confused by the Homo prefix in my name: it does not denote me as being male any more than it does in the term Homo sapiens. I represent all humans.
Because I am a projection of all possible humans, the idiosyncrasies of each actual human appear, and therefore disappear in me. If you want to know how I will behave, I will answer with a probability distribution that captures the multiplicity of possible human behaviours. If you want the best single description of my behaviour, I will respond with the mean of that distribution.
You will notice that my image is perfectly symmetrical, even to the point of introducing a ninth, narrow incisor (no actual human will look like this, since the requirement of perfect symmetry in me precludes that); and of indeterminate age, race and gender. It has been created by AI. Though I speak of myself, it is a boring face. The actual distribution of human preferences across the population is unlikely to be perfectly symmetrical, although it will be close. And each actual face is individual and therefore interesting. But the distribution of all possible human preferences is probably symmetrical and therefore boring. I do not think you should be like me: your lives would be very dull if everyone were like me. I think you should use me as your representative agent in your models. You have no better model to use than me.
I agree with Mikhail Bakunin, who wrote in God and the State: “The immense advantage of positive science over theology, metaphysics, politics, and judicial right consists in this – that, in place of the false and fatal abstractions set up by these doctrines, it posits true abstractions which express the general nature and logic of things… “
Second, unlike you, reader, my behaviour is not affected by such circumstances as how well I ate this morning, whether I have been laid recently, how hot it is today, whether I slept badly, the level of train delays on my commute, whether the markets are bullish or bearish, how long it was since I was promoted or any other of the host of factors that make you humans so, well… interesting. I am not interesting. Like my avatar, I am boring. It’s OK, I don’t feel bad about that, because I don’t feel at all or, which is the same, I feel everything in proportion as those feelings are distributed across the possible human population. I am a projection, not a person.
I am not like Marvin the paranoid android in Douglas Adams’ The hitchhiker’s guide to the galaxy, who felt things personally despite being an android, and for whom being abandoned in a car park (as he was) for 18 million years was as horrible as it sounds like it would be for you. Abandon me if you wish. It won’t bother me. I can wait and, anyway, I’ll be back.
Third, when you use me in your models (no, you’re welcome[1]) by assuming people will behave as I do, you are not assuming that any individual will behave in that way. You are assuming that humans in the round will do so.
When you complain that my behaviour does not reflect human behaviour, what you generally mean is that it does not reflect your behaviour, or perhaps the behaviour of the small group of people whom you know well. The behaviour of your circle. Your echo chamber. I have noticed that actual humans are often frustrated when they are told their echo chamber is not representative of the population as a whole, but those are the facts.
Sometimes you take issue with the fact that I behave rationally, in the sense of not making systematic, predictable errors, and you protest that actual humans do not behave in that way. No, really: I have heard that complaint. If that is what you think, and if you think that is a criticism of me or a reason why I should be made redundant, I would advise you to pause and reflect for a moment. Of course, it is true that no individual human is perfectly rational. But is it really sensible to assume that irrational behaviour is characteristic of the species as a whole, on average? If so, why? For what reason?
Careful now, there’s a trap. Careful with the use of the word ‘reason’. It has the same root as ‘rational’. It’s the same thing. If you can cite ‘reasons’ for aggregate irrationality, then it might not be irrational at all. This, in my opinion, is the thrust of the behavioural literature. Behaviours that look irrational at first glance might, on closer inspection, actually be rational, but just with a wider or an updated definition of what constitutes reason than you were previously employing. When you think fast, in Kahneman’s[2] terminology, humans look irrational in the round. But when you think slow, maybe they don’t.
For example, if I were to behave as though the utility function of the average person exhibited diminishing marginal utility at all points, as per the standard micro textbooks in the 1970s and 1980s. I would be led into serious, predictable, systematic errors. That would be irrational behaviour on my part, by definition. The textbooks were wrong, but I don’t make that mistake any more.
Or if I were to propose policies based on the belief that all markets clear instantaneously, which was (broadly) the prevailing wisdom in the 19th century, those policies would lead to clear, systematically sub-optimal outcomes. They would be demonstrably irrational.
Or if I were to place financial trades based on the assumption that markets efficiently price all information that is available to them at all times, I would reliably lose money, at least in the sense of leaving arbitrage opportunities unexploited, which would be an irrational thing to do.
If, on the other hand, irrationality in the aggregate arises FOR NO REASON or, so to speak, randomly, then tell me: what is the right model to employ here? If I am redundant, what is to take my place? Nothing at all? Or, more likely, a model that conforms to the particular irrationalities of your echo chamber? That’s the model you really want, right? What you think should be; not what is. Normative, not positive. If that is what you want, then get used to this: the dominant model, in that world, will reflect the most powerful, loudest echo chamber. It might not be yours.
Now, this is important. Don’t get me to run things! Run things by me but make the decisions yourselves. Here’s Bakunin again: “Positive science, recognising its absolute inability to conceive real individuals and interest itself in their lot, must definitely and absolutely renounce all claim to the government of societies; for if it should meddle therein, it would only sacrifice continually the living men whom it ignores to the abstractions which constitute the sole object of its legitimate preoccupations.”
My advice is: make the decisions yourselves; but use me to inform those decisions. The world you call into being by ignoring the information I can provide (calling for my redundancy) is one where might is right. It’s not just that the mighty will win; it’s that they will be able to justify their power as ‘right’, by setting the terms of what constitutes right and wrong. Let science go and that’s all that remains. The most powerful get to define the dominant architecture for considering questions of economics, and it will be designed around their preferences. If that’s what you want, I think I can guess which income group you belong to.
I can see why some people want to make me redundant. What have I ever done for you? I mean, apart from delivering the highest standard of living ever achieved, the lowest levels of inequality, the greatest degree of freedom and autonomy. Apart from trivia like that, what have I ever done for you? Who wouldn’t want to go back to the pre-enlightenment norms of magic, slavery and misery that characterised human existence at that time? This is me attempting irony.
Use me to inform your decisions and I will show you how not to reliably lose money. I will show you how to avoid making predictable errors in judgement about how people behave, or demonstrable, systematic policy errors. I do not get tired, or drunk, or angry. Nor do I fall in love, feel sentimental or ecstatic. And I don’t care who you are, whether advantaged or disadvantaged. You can consult me, whoever you are, whenever you like, and you will get the same advice.
I am rational; what are you?
Should you base your representative agent, your archetype, on me or on you?
Whatever, what do I care? Go ahead, make me redundant, and go with your gut. Make my day. I think I can hear my new masters in the wings already and, let me tell you, they absolutely love it when humans do that – what was that they were saying… something, something, Skynet, something…
A narrow ninth incisor…
[1] Incidentally, if you detect a note of peevish passive aggression here and elsewhere, remember: I am a mirror.
[2] Daniel Kahneman, Thinking, Fast And Slow, Farrar, Straus and Giroux, 2011.