J. Bradford DeLong:
Sluggish Future, Finance & Development, March 2017, Vol. 54, No. 1, IMF: We should adopt appropriate fiscal policies that provide for expansionary investment.
You are reading this because of the long, steady decline in nominal and real interest rates on all kinds of safe investments, such as US Treasury securities. The decline has created a world in which, as economist Alvin Hansen put it when he saw a similar situation in 1938, we see “sick recoveries… die in their infancy and depressions… feed on themselves and leave a hard and seemingly immovable core of unemployment…” In other words, a world of secular stagnation. Harvard Professor Kenneth Rogoff thinks this is a passing phase—that nobody will talk about secular stagnation in nine years. Perhaps. But the balance of probabilities is the other way. Financial markets do not expect this problem to go away for at least a generation.
Eight reinforcing factors have driven and continue to drive this long-term reduction in safe interest rates:...
The natural response to this secular stagnation is for governments to adopt much more expansionary tax and spending (fiscal) policies. When interest rates are low and expected to remain low, all kinds of government investments—from bridges to basic research—become extraordinarily attractive in benefit-cost terms, and government debt levels should rise to take advantage of low borrowing costs and provide investors the safe saving vehicles (government bonds) they value. ..
Critics of Summers’s secular stagnation thesis miss the point. Each seems to focus on one of the eight factors driving the decline in interest rates and then say that factor either will end soon or is healthy for some contrarian reason.
Since the turn of the century, the North Atlantic economies have lost a decade of what we used to think of as normal economic growth, with secular stagnation the major contributor. Only if we do something about it is it likely that in nine years we will no longer be talking about secular stagnation.
John Taylor provides a couterargument (I chose to highlight one over the other based upon my agreement with the arguments):