Tyler Cowen claims:
What might be Robert Barro’s argument?: We are once again witnessing the renaissance of old Keynesian economics as a theory of the long run not just the short run...
I think Tyler is wrong here. Old Keynesian arguments are not turning into a long-run theory. Rather, empirical reality is telling us that when inflation is very low--like now--the long run is likely to be a lot longer to arrive than we economists had previously thought it was.
Those whom Tyler calls "Old Keynesians"--me, Justin Wolfers, and Paul Krugman--used to think that the economy would not stay away from its long-run full-employment growth path for even as long as five years, and thus that when you looked out at the future five or more years in the future you should always use your long-run model rather than your short-run model.
But that is because up until the past decade cyclical variables like the employment-to-population ratio nearly always did return to their pre-recession levels within five years of the previous business-cycle peak. In the post-WWII U.S., only the 1957-8 recession was not followed by a rapid and full recovery.
Since 2000, things have been very different.
You can claim that supply-side rather than demand-side factors are responsible for our failure to reattain the full year-2000 peak in the employment-to-population ratio--that because we are richer now than we are then we wish to spend more time in school and more time in retirement than we did in the late 1990s. (You have a hard time sustaining that argument, however: only America's upper class looks to be significantly richer than it was back in the late 1990s.) But no matter what you strike as the full-employment level of the employment-to-population ratio today, we are not there and nobody is forecasting that we are going to get back to even the 2006 employment-to-population ratio in the next four years.
Empirical reality has told us that--at least when inflation is very low, as it is at present--the short-run is not less than five years but (shudder) can be as long as fifteen.
Justin, Paul, and I still have our long-run theories. It is just that empirical reality has told us that they do not apply until the end of the next presidential term.
We very much wish that it were otherwise.
But the principal intellectual task facing a macroeconomist today is to figure out how to change what he or she believes in response to the fact that empirical reality is speaking very loudly.
I get the impression that all of my right-wing colleagues are putting their fingers in their ears and telling empirical reality, in the words of Smeagol the Stoor: "I'm not listening. Not listening".