The Romer/Romer letter to Professor Gerald Friedman marks a turning point. It concedes that there are indeed important issues at stake when evaluating the proposed economic policies of Presidential Candidate Bernie Sanders. These issues go beyond the political debate and should be discussed seriously between and among professional economists.
All forecasting models embody theoretical views. All involve making assumptions about the shape of the world, and about those features, which can, and cannot, safely be neglected. This is true of the models the Romers favor, as well as of Professor Friedman’s, as it would be true of mine. So each model deserves to be scrutinized.
In the case of the models favored by the Romers, we have the experience of forecasting from the outset of the Great Financial Crisis, which was marked by a famous exercise in early 2009 known as the Romer-Bernstein forecast. According to this forecast (a) the economy would have recovered on its own, in full and with no assistance from government, by 2014, (b) the only effect of the entire stimulus package would be to accelerate the date of full recovery by about six months, and (c) by 2016, the economy would actually be performing worse than if there had been no stimulus at all, since the greater “burden” of the government debt would push up interest rates and depress business investment relative to the full employment level.
It’s fair to say that this forecast was not borne out: the economy did not fully recover even with the ARRA, and there is no sign of “crowding out,” even now. The idea that the economy is now worse off than it would have been without any Obama program is, to most people, I imagine, quite strange. These facts should prompt a careful look at the modeling strategy that the Romers espouse.…Dare I say smackdown.
Economic Forecasting Models and Sanders Program Controversy
James K. Galbraith