When the housing bubble burst, all that AAA-rated paper turned into sludge. So investors scurried back into the haven provided by the debt of the United States and a few other major economies. In the process they drove interest rates on that debt way down.I could also add that sovereign debt problems in the world contribute to shortages of safe assets, as do new banking regulations, for example the liquidity coverage ratio requirements in Basel III.
I've written extensively about safe asset shortages, particularly their consequences for monetary policy. You can find that stuff in my blog archive, working papers, and forthcoming and published work, for example:
Scarcity of Safe Assets, Inflation, and the Policy Trap, with David Andolfatto
Scarce Collateral, the Term Premium, and Quantitative Easing
Central Bank Purchases of Private Assets
A safe asset shortage is actually a property of Krugman's own model, though I'm not sure he ever picked up on this. Krugman used his work with Gauti Eggertsson to back up with formal analysis the kinds of things he was saying in blog posts. So, the Eggertsson/Krugman paper focused on a situation in which the nominal interest rate is at the zero lower bound because of tight borrowing constraints, in which case you can get some welfare benefits from increasing government spending. What fixes the problem in his model, in a clean way, is actually an increase in government debt, which serves to relax borrowing constraints and to raise the real interest rate. That is, because of the safe asset shortage the real interest rate is inefficiently low, and fixing the problem raises the real rate.
Recognizing the existence of a safe asset shortage can help explain a lot of things. For example, you might think that inflation is puzzlingly high. In most models we work with, a long period of zero nominal interest rates would produce a deflation. But, at the zero lower bound on the nominal interest rate, the tighter are the liquidity constraints that bind because of the safe asset shortage, the higher will be the inflation rate. To see how that works, see the papers I listed above.
Safe asset shortages also have a bearing on how we might think about "secular stagnation," which Larry Summers has been talking about. Eggertsson and Mehrotra's "A Model of Secular Stagnation," is, like Eggertsson/Krugman, actually a model of a safe asset shortage. E-M write down a fairly conventional overlapping generations model with some borrowing and lending and a binding borrowing constraint. This economy wants an outside asset badly - both to permit intergenerational trade and to bypass the borrowing constraint. But that's not something the authors consider.
In any case, there's hope for Paul Krugman. Maybe soon he'll be a committed neo-Fisherite.