Throughout my academic/research experiences in the social sciences and economic forecasting, it’s become clear that more complex models, whether it’s more variables, dynamics, or nonlinearity, rarely ever perform well. For the vast majority of situations in forecasting, it’s incredibly hard to beat a random-walk or an auto-regression (order 1).
There is no proof or explanation of why in an academic textbook, you just pick it up over time. Notable exceptions define entire subfields. The U.S. Term structure of debt is best modeled by using a set of ODEs to fit the cross-section, and stochastic dynamics to fit the time-series. The complexity there can grow enormously, and leads to lots of dense financial-math research, which actually does improve predictive accuracy in forecasting (still not by much, but it does consistently).
We actually see the same thing in economic analysis using words. While it’s often shakier than economists would like, describing monopolistic dynamics in an essay seems to be a nice approximation of reality in terms of predictive performance. I know this isn’t new to the LW crowd, but I always think of words as simply the painting of reality with non-linear dynamics in the way the human brain evolved to process information. That’s why neural networks, which learn these dynamics, work best for processing language (I think).
It turns out that words, like non-linear equations, are great at fitting data. If you find a subset of reality where you truly can use non-linear models, words, or both, to classify what’s going on, you’re in a great spot for predictive accuracy. Empirically though, in the collective experience of my field, that’s really hard to do. If your model diverges radically from a reduced form, random walk, or basic model, you need to be able to prove it wins.
Unfortunately, our brains do not seem to be good at detecting overfitting. The way I think about it, which like all evolutionary reasoning is questionable, is we evolved to learn nonlinear dynamics as we navigate our world, hunt, form relationships, and live in tribes. The complexity of a self-driving car is only a small subset of how we perceive reality. So, to us, it feels natural to use these words to paint nonlinear stories of reality, of the holy ghost, of Marxist theory, and all these advanced, nonsensical ideas.
Our thoughts suck because we overfit. If someone showed you a regression they fit, where they added a hundred transformations of the series of interest (squared, logged, cubed, etc), and their R2 was equal to 1, you’d tell them they are misguided.What’s the problems pace of Marx? “I fit a series of nonlinear dynamics, using words, to centuries of human interaction, and will use it to forecast human interaction forever.” Well, actually, you can do that. And it could be true. But it also might be garbage—nonsense.
Hey,
I’m gonna give you sort of an unsatisfying answer. I had a similar interest, which resulted in me getting my MSc and working in research at the Fed for a few years, with the goal of sorting it out in my head (ended up going private sector instead of getting a PhD). As far as I have surveyed, there are different models of money, but it’s scientifically an unsolved problem. There seems to be a level of complexity that arises as you increase the number of people on a monetary system, increase industries, increase geographical scale, add new countries and exchanges, and add complex financial systems. As this grows, filtering out what and how, exactly, money interacts with these systems, starts to get very messy.
As an example, during the financial crisis, trillions of dollars ‘disappeared.’ They disappeared because they only ever existed because we were borrowing from our future selves, then collectively lost faith in our future selves having that money, so the money ceased to exist today. Is that how a commodity behaves? Well, now we are trying to build classifications for what is and isn’t a commodity. Of course, you could do the same thing on a gold standard if banks were allowed to issue demand deposits, which combined with fractional reserve banking leads to the same thing.
Monetarism, I firmly believe, isn’t something you can reason through intuitively at a casual level. I decided it wasn’t something I wanted to devote my life to, and even though I spent a couple years working daily in the field, I don’t know that I understand that much (although I do know what I don’t know, which definitely counts as real knowledge).
I think monetary economics is sort of a mind-killer, since trying to intuitively reason through monetarism can take you down many very different paths, all of which seemingly arise from an incredibly reasonable set of axioms and inferences. If you ever listen to really clever Austrians or Keynesians discuss their view, it’s incredibly compelling. That sets off alarms in my favorite heuristic of undeterminism, when multiple models of the world fit the data equally well. It’s super common for blogosphere denizens or naive rationalists to try their hand at monetary economics, convinced they’ve stumbled upon some key insight that means all econ professors are wrong.
I will say, while I didn’t leave the fed enamored or anything, a subset of those economists are brilliant and humble. I notice this flawed reasoning so often, where independent researchers, or researchers in another field, will construct elaborate arguments against the most uncharitable readings of economists arguments. Often they won’t have ever spoken to a notable economist in person. They don’t ever have to present to peers, they never have to formalize their arguments mathematically, and they never bother engaging in the more advanced formulations of economic arguments that wind up in journals. Anyway, I’m getting off track here...
While as a rule I don’t think mathematizing things necessarily makes them clearer, I am convinced it’s the right way to proceed in monetary studies. It forces a strict structure, which prevents us from using words to overfit or get lost. Although the field is so complex, and it’s intertwined with historical narratives that aren’t always easily turned into data sets, so that can sometimes make it harder. The math often gets sorta complicated as well.
Of course, the actual monetary economy has real data. Most of which we can’t collect. So the theoretical models are our way of trying to imagine what the structure would look like, even though they aren’t empirical. Which gets to another problem, which is how confident can one be in theoretical economics? Sometimes the assumptions are incredibly robust, but the systems are often very complex.
One place I will say I think many economists act contrary to LW style rationality, is in choosing a side, rather than taking the rational view that there are many sides with equally valid claims to truth, and they should work together to expose what is correct. It has always struck me as being mind-killed when people state “Oh, I’m a neo-Keynsian so I believe XYZ, you’re a non-Keynsian, so you reject ABC” (or whatever). I mean… maybe the Austrians are all right, and they have this unique perception of reality none of the Neo-Keynsian scholars have, because they have some more profoundly true insight into the mesh of reality that is lost on the other econo-plebs… But that doesn’t seem like the most likely scenario to me.
Or maybe Paul Krugman really is right about everything, but still… I doubt it. He was once a smart young man that had some crucial insights on the theoretical mathematical structure behind international trade, which earned him an econ Nobel. I don’t think he’s in tune with empirical realities though. He’s just a genius at imagining some elegant mathematical structure that characterizes an economy which might or might not map to reality, and then convincing himself it’s actually exactly how reality operates. That’s the big mistake, I think.
If you want to take a look down the rabbit hole, I’d suggest reading Milton Friedman’s books on monetary history. Even his detractors tend to agree his insight and clarity on money is absolutely incredible. He also is great at explaining things without too much math, but still using ratios and dataseries in his books when appropriate.
For shorter term stuff, check out John Cochrane’s stuff, he’s my favorite social scientist, (http://faculty.chicagobooth.edu/john.cochrane/research/papers/cochrane_policy.pdf, http://johnhcochrane.blogspot.com/search/label/Monetary%20Policy). His blog—second link—is really great.