The Economist as Scientist, Historian, and Philosopher

The persistence of one thinker in economics over another is something I have been wrestling with, but haven’t come to any sound conclusions. In reading the big names over the last couple hundred years, you can see the threads of who influenced whom and who dropped out of importance. Either their ideas were dead ends to most people and thus remain a historical curiosity like the Classicals, or they have been explained better. We still talk of Keynes but don’t use his models.

It strikes me because what is appealing to me is that Economics is a field still under contention. When I was a kid in undergrad studying Chemistry, part of what was alienating was that most of the big questions had been answered, or so it seemed. To really understand the interactions of particles at these nano-scales you had to get into physics and then into probabilistic math – once you’re thinking of electrons as probability fields and not point particles, a lot of the simple models seem like child’s play even if they’re fairly accurate representations of how the world works.

And it’s this piece of how does the world work is why I keep learning and exploring. I doubt I’ll ever scratch but the surface, but it is a fun journey. So it circles back to economics – we build models to formalize the narratives we were looking at. These models have assumptions built in and should have much fewer graduations of meaning than the slippery human language. The problem I worry about is if this reliance of the mathematical models takes us away from the larger picture. You build the model and look at the data and see if the data fits the model in the past and then see if it fits it going forward. Then if it does or it doesn’t, we can still have argumenta about the model or the data you chose or if it falls prey to Goodhart’s law or the Lucas critique.

Because here’s what I image is the ultimate endpoint in mathematical economics: assume an all-powerful super-computer with access to all the economic data points ever collected. With these powers, it is able to build a model so complex that no living human mathematician can even begin to understand it. But it is perfectly predictive and adaptive to outside shocks and following the recommended policy put forth from this machine maintains low inflation and full employment.

Here’s the question though – would you trust it if no one could explain why it came to those conclusions? I don’t know if I would. And this is why straight mathematical reliance seems to me a dead end. We have to be scientists, yes, but also historians and philosophers to know what the math is really saying. That means not losing contact with the past.