Civil engineers are rarely able to predict the collapse of US highway bridges, at least to within a period of 12 months. How should we feel about that fact? Maybe I’m biased, as my grandfather was a highway engineer in Michigan, but I feel much better knowing that civil engineers are unable to accurately predict highway bridge collapses.

Screen Shot 2016-02-03 at 9.41.28 AM.png

One reason they are so unsuccessful is that they are responsible for both predicting and preventing bridge collapses. So while their conditional forecasts of bridge collapses may be excellent, their unconditional forecasts are lousy. Thus civil engineers may be able to say, “If you don’t fix the crumbling 3rd avenue bridge overpass, it’s likely to collapse within 12 months”. But then other civil engineers in the state highway department go out and prop up the bridge. So it doesn’t end up collapsing. Conditional and unconditional forecasting ability—they are very different concepts.

Economists have lots of good models of the economy, and would probably be able to make pretty good conditional forecasts of recessions. “If you let the price of NGDP futures contracts fall by 8%, a recession is likely.”

But economists don’t just predict, they control monetary policy. As soon as a recession is predicted, the Fed tries to prevent it. They raise NGDP growth expectations. Thus while conditional forecasts of recessions might well be excellent, unconditional forecasts of recessions are lousy. And that makes me feel much better about the economics profession, just as the inability of civil engineers to predict bridge collapses makes me feel much better about civil engineering.

Engineers and economists, creators of the modern world. Without either you are in the Stone Age. Add engineers and you get up to North Korea, with its 300 meter high skyscrapers and its atomic bombs (and starving people.) Add engineers and economists and you get to South Korea.

Screen Shot 2016-02-03 at 9.34.37 AM.png