In a Wall Street Journal op-ed entitled “Government Forecasters Might as Well Use a Ouija Board” (appearing today, Oct. 16, 2014), Professor Edward Lazear of Harvard University cites several historical examples of gross inaccuracies in government macroeconomic forecasting that nonetheless were critical to policy formation. In light of this history, he properly urges government officials and political leaders to exercise far more humility in making claims based on such forecasts. 

Nonetheless, Professor Lazear leaves out of his discussion the principal reason why econometric forecasts are so notoriously bad.  Unlike models of the physical world where the data are insentient and relationships among variables are fixed in nature, computer models of the economy depend on data grounded in motivated human action and relationships among variables that are anything but fixed over time.  Human beings have preferences, consumption patterns, and levels of risk acceptance that regularly change, making mathematical relationships derived from historical data prone to being grossly inaccurate representations of the future.  Moreover, there is little hope for improved accuracy over time so long as human beings remain sentient actors.  It is no wonder that macroeconomic forecasting is largely an exercise in futility.