Jumping To Conclusions

In discussing the recent book Moneyball, in which author Michael Lewis profiles Oakland As General Manager Billy Beane, Tyler Cowen of Marginal Revolution observed:

An obvious question: if it is so hard to measure the performance of first basemen, when there is a slew of publicly available statistics, how about the rest of the economy?

It's an important insight -- and one about which I'll have more to say in the near future -- but it's not the most troubling one which one can take away from the book.

In fact, it isn't that hard to measure the performance of first basemen. Oh, I don't mean to imply that sabermetrics (the study of baseball) is as well understood as physics or as well documented as chemistry, but the field is no longer in its primitive stages, either, and there are well-established theories and methods. The real problem with baseball isn't the difficulty of measurement, but the difficulty of overcoming entrenched interests who are threatened by new ideas.

It's the equivalent of the difference between understanding the benefits of free trade and convincing politicians that they should support free trade. You're not going to find many mainstream economists who thought that George Bush's steel tariffs were a good idea, but that didn't seem to have much of an impact on the people imposing them -- perhaps because they had an interest in doing so. Similarly with baseball: it's not that you can't measure a first baseman's performance with all the statistics that are collected, but that if you can, people currently in positions of power within the industry, who are comfortable with existing techniques but not the new ones, will face losing their jobs to those who are skilled with both. So they have an incentive to denigrate the idea that these new methods work. (Perhaps a better analogy than steel tariffs is school vouchers. Certainly there are unresolved issues about measuring how well they work, but the real problem is that the people who would be directly affected -- the school bureaucrats and teachers -- have no interest in even trying.)

The primary question one has after reading this book is not "How can we possibly hope to measure the economy, given the statistics we have?" but "How can we possibly hope to get people to try new ideas when they got where they were using old ideas?"

By the way, Moneyball is wonderful; and I recommend it to anybody who has even a tiny bit of interest in baseball. Lewis is an excellent writer (I also recommend his fascinating, though somewhat dated, Liar's Poker), and he makes a topic which could be dry into a page-turner. Those with a more intimate knowledge of the subject matter will see that Lewis's understanding of the topic is somewhat superficial, but the book is still a great read, and it provides a window into the behind-the-scenes operations of a baseball team that you can't get from the sports pages.