Hesitations on Economics II

One of the reasons I'm not sure I'll become an economist is because of it's confused epistemology, something that makes it unworthy of pursuit "for its own sake," and creates an obligation for the economist to enact change in the world as a result of his or her work. By this, I mean the lack of any clear reason for any of its results to be true, or reflect some deep fundamental truth about the world. This is made clearest by contrasting it with a field that does not have such a confused epistemology: physics. In physics, there is a lot of math used, and physical theories are formulated in the language of mathematics. This is important because this means they can yield falsifiable predictions that we test to see if the theories are correct about the nature of the world. In physics, we started with heuristics models like Newton's laws, that, roughly, just fit an equation to a line: "huh, it appears that gravity is weaker quadratically over distance." The thing is, even this crude heuristic model was scarily accurate: Newton's laws describe gravitation with more precision than any economic theory describes any human phenomenen, ever. And so because of how accurate this theory was, we were led to believe that there is some underlying structure that yields this exact quadratic decay that we're not understanding.

And indeed there was: when Einstein formulated his theory of general relativity, it posited that space-time has a particular shape described by a specific mathematical object. We can't see the mathematical object that we believe underlies space-time (a four-dimensional Lorentzian manifold), but we think it really is there because of how perfectly accurate the theory of general relativity is in predicting gravitational phenomena. Thus, modern physics now asserts the existence of things we cannot see as actually being there in the universe, and is able to test these theories. It rests on the bedrock epistomelogical assumptions that the laws of physics we derive here on Earth apply uniformly across space and time, assumptions that we have never seen violated and have extremely compelling reason to believe true, even if we don't know why.

There are no such equivalent bedrock assumptions in economics. When we derive a set of mathematical equation to predict some economic phenomenon, we might have intuitive and heuristics reasons to think it will or will not hold across space and time. It may just be that a certain economic phenomenon you observe in an experiment is a an artefact of culture at the place and time you observed it. There is absolutely no reason for it to be a universal and timeless law, because economics has never, not once, discovered anything that fits that description. When it comes to the laws of the universe, we have reason to believe it obeys the language of mathematics (ie. A implies B, so you cannot have A without B), but human behavior has no reason to be rational in this way. Now, let me be clear: I am not critizing the "homo economicus" assumptions made by economists -- those are simplifying assumptions without which it would be difficult to model the real world. I am criticizing the fact that they sometimes think their work sheds some light on something fundamental and real in the world, when there's no reason to believe that any of these conclusions generalize to humanity over long time scales. Maybe hyperbolic discounting is just something done by humans today, and our discounting functions will change as a result of changes in cultural values in 100 years! Then all the modern work on intertemporal choice will be irrelevant! Insofar as economists study rational behavior, their conclusions are timeless, but incorrect! And insofar as economists study real world behavior of humans, their results might be "correct" in some myopic and narrow empirical sense of fitting a specific data set, but they are not timeless or universal! This is a confused epistemology!

This brings me to the burden of the modern economist: to do work that will impact millions of lives through policy. One thing that absolutely stunned me was when I spoke to economists and asked them how their work will affect policy, they really couldn't care less. It's one thing to say this when you're working on something theoretical that is more akin to mathematics than economics (like game theory or mechanism design), but if you're working on something empirical like development, behavior, or industrial organization, then you have an obligation to make sure your findings impact policy because you have no reason (circa above) to believe they have any intellectual value as lasting truths about reality in the same way that physical laws or mathematical theorems are lasting truths. If you cannot know that your work is timeless or universal, you damn well better be sure it's going to improve the lives of many real people. Because work that is both irrelevant in 5 years, and never impacts anyone, is work that might as well never have been done, in my opinion. And so it is very worrying when economists don't believe that they can seriously impact policy. The cry of "Oh, politicians are burdened with so many other constraints" scares me because it reminds me of management consultants that have no understanding of the real lived constraints their clients experience, yet make strong and opinionated recommendations from their ivory tower that never see the light of day (and have no intellectual value in and of themselves, either).

TL;DR. I want to do work that is either 1) timeless and universal, or 2) changes people's lives. There are a non-trivial amount of top economists that I don't think are doing either, and this is a reason I am not yet sure I want to be an economist.