Megan McArdle Knows Galileo Was Waaay Over-rated
As many long-time readers know, I’m working (with a co-author) on a book that looks at a bunch of different series – from abortion to crime to the unwed motherhood – and how they evolved over the administrations from Ike to GW. Since both I and my co-author have a special interest in economics, about half the book covers economic topics – from economic growth to the national debt. FWIW, I hope to have an announcement about the book soon as I just signed a contract and dropped it off at the mail.
One thing we’ve found – and this also should be no surprise to long-time readers – is that Democrats generally outperform Republicans on economic issues. So I was interested in this quote by Megan McArdle a couple weeks ago:
But I expect that four years from now, we’ll still be having the same conversations with proponents of “cancer clusters” and Democrats convinced that they can scientifically prove that Democrats are better for GDP by doing ham-fisted regressions of Democratic presidencies with a few tightly correlated economic variables. What’s the mechanism? What makes electric power lines cause cancer, but not the earth’s vastly more powerful magnetic field? What policies did Harry Truman and Bill Clinton have in common (but not with Richard Nixon) that caused this marvelous confluence? Well, maybe we don’t know the mechanism exactly, but never you mind: just look at that bee-yoo-ti-ful correlation!
I’ve been swamped, so I kind of forgot about it, until until I saw that quote referenced approvingly elsewhere. Anyway, I don’t know the first thing about cancer clusters, so I won’t comment on that at all, but I do know something about that other issue.
FWIW, in our book, we do discuss the mechanism, but only in passing. Our book’s purpose is to provide an over-view of what actually happened so explanations of “why” – while there – are necessarily limited. There’s only so much that can be covered without making a book that is four volumes long that nobody will read. Still, we do try to explain why economic performance skews the way it does.
For what its worth, and I don’t want to think too far ahead because I’m swamped, but I have in a mind a second book that is about nothing but the mechanism.
(cactus says turn the page)
But in my opinion, macroeconomics is not a science, and won’t be for a long time, precisely because many practitioners spend too much time on the mechanism and not on observing what happened. For example, the “supply side” story, mechanism and all, makes perfect sense. Its simple. Its elegant. In fact, its beautiful. I like it. But it doesn’t match the data in the US. And most of the supply side success stories to which we were treated in the past few years – the small eastern European countries, the Celtic Tiger, and Iceland – all rank up there as this decade’s Argentina now. And remember, there was a time that Argentina looked good too, and the same cast of characters was talking them up too. All of which is to say… the supply side story is sensible, simple, elegant and useless. Absolutely and completely useless.
For macroeconomics to one day aspire to be science, it has to begin to function like sicence. And science doesn’t start with a story, however pretty. It starts with observation. Tycho Brahe spent decades observing and cataloging movements of the “heavenly bodies.” Eventually, he proposed a model of the solar system. It was wrong, but it was an attempt to explain what he had observed. But it all started with observation. And though his observations form one of the building blocks of modern astronomy, Brahe had no mechanism. None.
Kepler, Brahe’s assistant, stared at these decades of observation and somehow boiled it all down to three simple laws of planetary motion that are still taught today in high school physics. To call that an amazing feat is to understate the accomplishment. But Kepler couldn’t give Megan McArdle a mechanism because he never figured one out.
And then there was Galileo. Many physicists consider him to the first modern scientist; he experimented, he observed, and he laid down, among other things, laws of motion that help form the basis for modern physics. But he didn’t have a mechanism. He understood how bodies moved, but not exactly why they moved the way they do. And while he didn’t have a mechanism, the prevailing, er, theory at the time had one. And proponents of that mechanism felt that Galileo’s theory, based on observation, contradicted their mechanism, and boy did that cause Galileo no end of trouble. But as far as advancing science is concerned, Galileo’s observation are far more important than his persecutors’ mechanism.
If my knowledge of physics was better, I could go on to the present day. And while my knowledge of biology is as superficial as my knowledge of physics, I know its the same story there. Darwin started with observation, the Theory of Evolution, the mechanism, came later. And its the same story in most areas of human endeavor. And not all the great work was done by giants. Galileo may have been a genius, but Brahe’s contribution came from having the patience to observe and observe and observe some more, taking precise and careful notes all the while.
So the focus on a mechanism, quite frankly, is misguided, when there is not even agreement about what the facts actually are. Mechanisms matter, but facts are paramount. And I say that as someone who has in mind a second book that is all about the mechanism for why some Presidents outperformed others. I’ll give you an obvious example of lack of grounding in the facts. McArdle asks:
What policies did Harry Truman and Bill Clinton have in common (but not with Richard Nixon) that caused this marvelous confluence?
But that question is simply wrong. McArdle may as well be asking why a fish is like four wheel drive but not a liquor store. Truman is the one Democratic President who Republicans seem to adore. GW compared himself to Truman. Sarah Palin compared herself to Truman. Right wing blogs write favorable things about Truman. Could I possibly have replaced Truman’s name in the previous sentence with Clinton’s, or Carter’s, or FDR’s?
Additionally, data on real GDP per capita is available since 1929. There were fourteen Presidents between 1929 and 2008, the last year for which we have data, six Democrats and eight Republicans. All the Democrats except one, Harry Truman, are in the top half of the sample. Truman is actually second from the bottom, right above Hoover. (Rankings here.)
So… McArdle’s question should be a different one. It should be:
What policies did Harry Truman and George Bush (either one) have in common (but not with Bill Clinton) that caused this not so marvelous confluence?
There’s also the question of whether Truman was a special case, what with presiding over the transition at the end of World War 2. But then again, that transition, that might also have been what precipitated actions that more closely resembled those that would have taken by George Bush (either one) than Bill Clinton.
Anyway, I think I have the answer to these more relevant questions. And I think I know the mechanism by which the answer works. But the first step is always knowing what the data says.
In McArdle’s post, she points out that often one can find correlations between unrelated things. As my econometrics professor used to say, there are artifacts in the data. But here’s a tip for McArdle: don’t stop when you see a pattern in data. Look at a lot of data, try to find similar data, perhaps from other places or other times. Because sometimes what looks like patterns disappear with a bit more data, or another way of looking at things. But sometimes, they don’t. And if the pattern doesn’t go away, but it still doesn’t match your mechanism, then the mechanism is wrong.