Economists' models perfectly predict behavior... in their mathematical fantasy world

I was very disappointed by academic economics while earning my Ph.D., in part because the field seemed to care only about trees, never the forest. And it only cared about certain trees, the biggest and most easily accessible. So I always enjoy reading another article showing how narrow-minded economists have been.

The book, and Ms. Reinhart’s and Mr. Rogoff’s own professional journeys as economists, zero in on some of the broader shortcomings of their trade — thrown into harsh relief by economists’ widespread failure to anticipate or address the financial crisis that began in 2007.

“The mainstream of academic research in macroeconomics puts theoretical coherence and elegance first, and investigating the data second,” says Mr. Rogoff. For that reason, he says, much of the profession’s celebrated work “was not terribly useful in either predicting the financial crisis, or in assessing how it would it play out once it happened.”

“People almost pride themselves on not paying attention to current events,” he says.

In the past, other economists often took the same empirical approach as the Reinhart-Rogoff team. But this approach fell into disfavor over the last few decades as economists glorified financial papers that were theory-rich and data-poor.

Much of that theory-driven work, critics say, is built on the same disassembled and reassembled sets of data points — generally from just the last 25 years or so and from the same handful of rich countries — that quants have whisked into ever more dazzling and complicated mathematical formations.

But in the wake of the recent crisis, a few economists — like Professors Reinhart and Rogoff, and other like-minded colleagues like Barry Eichengreen and Alan Taylor — have been encouraging others in their field to look beyond hermetically sealed theoretical models and into the historical record.

“There is so much inbredness in this profession,” says Ms. Reinhart. “They all read the same sources. They all use the same data sets. They all talk to the same people. There is endless extrapolation on extrapolation on extrapolation, and for years that is what has been rewarded.”

This rings so true. For several decades, I read Business Week religiously, but nothing I read ever seemed relevant to any of my classes or any of the research anyone I knew was conducting.

Similarly, I wanted (in the mid 1990s) to study China’s emerging economy, but professors generally seemed to think studying anything other than the U.S. or Western Europe was a waste of time. And I had been fascinated since college by Kahneman and Tversky’s “prospect theory” about how people behave systematically irrationally, but the entire economics field was built atop two bedrock assumptions: 1) that people behave rationally or, at least, “as if” they are rational; and, 2) that people are perfectly informed about all relevant information or, at least, behave “as if” they are perfectly informed. Relaxing these ideological constraints was too difficult — both mathematically and econometrically — except in very simple models. Viewing themselves as social physicists rather than historians or sociologists, most academic economists stuck with the absurd but mathematically tractable assumptions. They’d rather correctly predict and explain behavior in their imaginary fantasy world than get things mostly correct in the real, messy world of human beings.

I found the experience frustrating, so I admire anyone who persevered against the powerful tide pushing the field toward mediocrity and managed to achieve real insight into the messy, complex world we people actually inhabit.

Posted by James on Sunday, July 04, 2010