Saturday, August 19, 2017

Two points from "What economists study"

At voxeu, What economists study: A guide for the curious by Christopher Snyder.

Took me a while to get around to reading this, because it didn't strike me as "economics". I mean, if economists can give us things like the Global Financial Crisis, then surely economists are not studying the right things. But when I sat down to read it, I liked it. I liked it even though Snyder writes:

One of the great advances of modern empirical economic research is causal identification—uncovering true causal relationships rather than overinterpreting apparent correlations as causation.

Maybe that's what I do, overinterpret apparent correlations as causation. I look at graphs and try to find things that are similar and things that are dissimilar. But hey, I have my theories, and my theories are based on historical data, and that's good enough for me. I'm no economist, and I can't do the kind of math economists do, "causal identification" and all that, but economists are welcome to use that stuff on my theories any time they want. Meanwhile, there are a couple points in Snyder's article that I want to mention.

1. The Just-Price Theory

In the Middle Ages, philosophers advanced the just-price theory. This argues that value is an inherent property of an object. On this view, diamonds are expensive because of their inherent quality, and water is not.

What a cute idea! Gold isn't valuable because it's rare. Gold is inherently valuable.

Nobody thinks like a Middle Ages philosopher these days, except those who think money should be backed by something that has intrinsic value.

2. Feedback from the Economy

Economists have to devise clever ways to establish causation in nonexperimental data. Pick one ... A macroeconomist might pick Romer’s (1992) study of whether fiscal or monetary policy deserved more credit for the US recovery from the Great Depression. The problem: a bad economy can feed back to make beneficial policies look damaging.

A bad economy can feed back to make beneficial policies look damaging. Yeah, exactly, and that's very important. There is another version of the same idea: A good economy can make any policy (good or bad) look good.

There are techniques that rely on such ideas. A common one is to separate the good economy from the bad and study only the bad years. Any study that starts in 1980 or after, ignoring the "golden age" that followed WWII, is one of these. There must be millions of 'em.

No comments: