Last night in the tinychat, I was talking with libertarians-and-stoya about the role of empiricism in economics. I felt like I made my point, but I had a few appletinis in me so I don’t think I made it as well as I should have made it. So allow me to take this time, when I am sober, to more succinctly state what I was trying to say.
Empirical methods have their role in economics, no doubt. But the question of empiricism isn’t whether tools such as econometrics should have a role within economics, but rather whether they should be the primary basis of all knowledge.
Economics is the study, first and foremost, of acting agents. It’s the study of action undertaken by people, and how those people interact with their environment and also their society. Empiricist methodologies do have their place in helping us understand this behavior, however it’s not exactly appropriate as serving as the foundational cornerstone.
Some would argue that empiricism is the best way to study economics, because economics is a science, and since empiricism is the primary methodology of other sciences, why shouldn’t economics be the same? Well, that’s because the scope of economics is entirely different from the scope of, say, physics or chemistry. In chemistry, the chemist studies a molecule. Molecules exhibit no agency: they react in completely predictable patterns, they follow a set course, they have no free will. Molecules simply are.
Contrast that to humans and their actions, which are the subjects of economics. Unlike a molecule, humans exhibit agency. Humans act of their own volition; in other words, free will exists (a bold assumption, but it’s one I’m going to make). Because the subject whom the economist study is an acting human being, a human being that makes their own choices and exercizes autonomy over themselves, the science of economics is of a fundamentally different nature than the science of chemistry. Empirical methods work great when you know what you’re studying is going to behave predictably and consistently. When you’re not dealing with entities exhibiting agency, it is entirely possible to control the environment of the subject in such a fashion as to isolate and segregate enough traits so that all, except the one or two properties which the scientist wishes to study, can be controlled. When you’re dealing with agents, however, this is not true. That one difference, the difference of free will, changes the entire equation. Free will is ingrained into the fabric of the acting agent. You cannot isolate free will. Therefore empiricism cannot be the primary basis of the science of economics.
This isn’t to say empirical methods don’t have their place within economics. Empiricism is a helpful in the sense that it can demonstrate to the non-believer the validity of the logical deductions of praxeology. It can firmly solidify what you’ve already proven in theory. Empiricism is good for all these purposes. However, at least within the scope of economics, empiricisms role is relegated to a secondary one. The most that an economist can ever hope to gain from epirical methodology is to confirm what they already know to be true. It’s role is tertiary, not primary.
Last paragraph bolded by me. Looking at economics empirically can be extremely useful when you understand the principles, but extremely dangerous if you’re operating without a basic understanding of the sporadic nature of autonomous human action. I would say that it should never be the case that an economic axiom or law relies more heavily on empirical happenstance than it does on a priori praxeology.