About Economic Forecasting

In the the previous post I disparaged the ability of economists to estimate the employment effects of the minimum wage. I'm skeptical because economists are notoriously bad at constructing models that adequately predict near-term changes in GDP. That task should be easier than sorting out the microeconomic complexities of the labor market.

Take Professor Ray Fair, for example. Prof. Fair teaches macroeconomic theory, econometrics, and macroeconometric models at Yale University. He has been plying his trade since 1968, first at Princeton, then at M.I.T., and (since 1974) at Yale. Those are big-name schools, so I assume that Prof. Fair is a big name in his field.

Well, since 1983, Prof. Fair has been forecasting changes in real GDP over the next four quarters. He has made 80 such forecasts based on a model that he has undoubtedly tweaked over the years. The current model is here. His forecasting track record is here. How has he done? Here's how:

1. The median absolute error of his forecasts is 30 percent.

2. The mean absolute error of his forecasts is 70 percent.

3. His forecasts are rather systematically biased: too high when real, four-quarter GDP growth is less than 4 percent; too low when real, four-quarter GDP growth is greater than 4 percent.

4. His forecasts have grown generally worse -- not better -- with time.

How hard can it be to forecast the direction and magnitude of macroeconomic activity given the plethora of relevant data at hand? It can't be as hard as estimating the employment effects of changes in the minimum wage, where such effects are subtle and perhaps impossible to measure (e.g., employers shave unmandated employee benefits or allow working conditions to deteriorate; businesses are not continued or started that might otherwise continue or start).

By the way, Prof. Fair also has a model of presidential elections that he first published in 1978 and has tweaked six times. (Links are here.) According to the current model, Pres. Bush will receive 51.7 percent of the popular vote even under three extreme conditions: zero percent real GDP growth in the first three quarters of 2004, 5 percent growth in the GDP deflator over the first 15 quarters of the Bush administration, and zero quarters (out of the first 15) with real GDP growth at an annual rate of more than 3.2 percent. In other words, a Republican incumbent is a shoo-in for re-election. Ha!

I rest my case.