If you’re reading this, you probably already know what econometrics is. Just in case, econometrics is the application of mathematics and statistics to the measurement and prediction of the economy. It’s “all them fancy equations and graphs”. And econometrics is for sale. Anyone who has a political policy position can find and hire an econometrician who will prove their point for them. In what follows, I’ll try to explain why that’s problematic.
Digress: I was a member of the professional marine navigators’ guild. I navigated ships across oceans in the days before GPS. As we practiced marine navigation back then, you made predictions, then took measurements with calibrated instruments, then performed mathematical operations on the values you’d measured, then plotted the results, then adjust course. If you got it right, you arrived safely at your destination. If you got it wrong, you ended up pierced by a rock on some unknown shore.
This, and a multitude of other similar activities, taught us Westerners that numbers are good, because if one is good with numbers, good results occur. Safe voyages, bridges that don’t fall down, electric lights. And numbers are indeed good, if they’re used properly, but not all numbers are used properly.
Econometricians don’t normally use numbers properly. First principle, you can only manage what you can predict. If you can’t predict outcomes, then it’s not management … it’s random stabs in the dark, accompanied by volumes of argumentation and equivocation. The econometricians who claim to have the answers can’t predict outcomes. If they could predict outcomes, they’d all be rich. If they were all rich, at least one of them wouldn’t be able to resist publishing the “how to”. If they published the “how to”, then the markets would begin reacting to every indication that someone was using that method. Before you know it, their “how to” would become a “how was”. There actually was a hedge fund that did this, and crashed, and everyone lost a lot of money.
If an economist could predict outcomes, that would change the outcomes such that they would once again be unable to predict outcomes.
Regression: econometrics might be useful in some very carefully circumscribed situations, for limited purposes over finite spans of time.
Econometrics as it’s practiced today is very similar to astrology bedecked with mathematics, and parading around as astronomy. The economist who practices econometrics started with a desired policy objective. Since most Americans glaze over when presented with a page full of numbers, this econometrician first identifies a graph that will prove their point. He or she then works backwards toward a a system of equations that potentially, conceivably could produce a graph of that type. Finally, the econometricians plugs in the right set of values into the selected set of equations to produce a properly shaped graph to prove the policy position they began from. In other words, it’s a bunch of hocus-pocus, masquerading as science. Or even worse, claiming to be engineering.
Mathematics and statistics are extremely powerful tools for ascertaining the truth. They should never be accepted as part of an argument to prove a position. That’s a wordy way of restating the old saw: “figures don’t lie, but liars figure”.
The study of economics is essentially the study of people. No people, no economy. Obviously, there are physical, real-world constraints that can be enumerated. Except for those limited set of constraints (“exogenous forces”), most of the rest of economics is little else except psychology. What causes people to decide to save for tomorrow vs spending today? How much is ‘enough’? How hard will people work, and how much will they sacrifice in the rest of their lives, for prosperity in the economic realm of their lives? When Jane woke up this morning, was she more inclined to buy a new pair of shoes, or gold? The answers to those fundamental questions do not lie in the realm of mathematics or statistics.
I’ve looked at a fair sample of econometric data and I’ve come to the conclusion that most of it is useless, some of it is as good a three day old dead reckoning position, and a tiny fraction is actually contributes to understanding. In the end, these things will be decided by a majority vote anyway, so logical arguments are largely wasted. I can only hope that more informed citizens will make wiser, or at least less foolish, decisions when they cast their ballot.
G’day all, and may God continue to bless America!