Economic Forecasters and Okun's Law

Abstract

Okun’s Law is an empirical relationship between unemployment rates and real growth rates discussed in most macroeconomic textbooks, with a onepercentage-point increase in unemployment associated with about a 2-percentage-point reduction in output growth. We examine whether the Okun relationship is also evident in the predictions professional economists make about unemployment and growth reported in the Wall Street Journal. We estimate versions of Okun’s Law on actual real-time data to see what the economists could have learned before making their forecasts. We then estimate Okun’s Law on an unbalanced panel of forecasts by economists who made at least 5 sets of predictions. We find support for the hypothesis that the forecasts exhibit Okun’s Law but find the Okun coefficient estimated from the forecasts is somewhat smaller than the coefficient estimated from actual data. Additionally, the Okun coefficient from forecasts is larger later in the sample period despite being smaller in the actual data. We find no significant differences in estimates of Okun’s Law across different types of firms employing forecasters. We do, however, find evidence that the forecasters believe the Okun coefficient is larger for unemployment-rate increases than decreases, contrary to the real data. We then analyze a balanced panel of forecasts by economists who responded to each of twenty-two surveys. When coefficients are constrained to be identical across economists this smaller sample produces results similar to the larger sample except for showing no indication of separate Okun coefficients for increasing and decreasing unemployment. We find, however, significant differences in estimates of Okun’s Law across economists, indicating differences in their forecasts reflect differences in both outlook and their Okun coefficients.