Sunday, September 2, 2018

Secular Stagnation, Myth and Reality

I ran into an article by the usually excellent Joseph Stiglitz on secular stagnation, entitled "The Myth of Secular Stagnation," see here. The gist of the article appears to be that the idea of secular stagnation is some sort of a ploy to absolve policy makers from responsibility of the slow recovery from the Great Recession. I think this view is so fundamentally wrongheaded that it seems worthwhile waking this little blog up from the dead to offer a brief comment on this notion.

I suppose I should disclose some relevant biases before going further. When Larry Summers first mused about the secular stagnation idea in a speech at the IMF in the fall of 2013, Neil Mehrotra and I wrote a little paper a couple of months later, see here, which as far as I know was the first attempt to formalize it in a modern DSGE model. Later, with the help of our graduate student Jake Robbins, we moved beyond a simple theoretical illustration to explore a quantitative version of the hypothesis, see here, in a paper that is coming out in American Economic Journal: Macroeconomics. In the meantime, I have written a series of paper with said Larry Summers, and various co-authors, that have explored several aspects of this idea, see here, here and here. Admittedly, it comes as a surprise, if I had unwittingly been participating in some enterprise that supposedly absolved policymakers of any responsibility for the slow recovery!

The biggest, and perhaps most obvious, problem with Stiglitz argument, is the following: If the secular stagnation hypothesis is correct, this does not in any way absolve policy makers from responsibility for a slow recovery. Instead, it does exactly the opposite. If correct, the secular stagation hypothesis implies that policymakers should have done much more in 2008 than existing theories suggest.

What is the secular stagnation idea anyway? In prior work, the way most people, myself included, had thought about the crisis of 2008 when the US and most of the rest of the world hit the zero lower bound, was that it was due to some temporary forces, such as being generated by a debt deleveraging cycle (see e.g. my work with Paul Krugman here) or being driven by problems in the banking sector (see e.g. joint work with Del Negro, Ferrero and Kiyotaki, see here). But in any case, most of these theories where ones in which the forces leading to ZLB were temporary, and thus one strategy for policy (for example if the cost of policy intervention was considered very high) was one of just waiting it out, as "soon all would be well" using Stiglitz words.

What separated Larry's secular stagnation hypothesis from much of the earlier work was that he suggested that the forces that might be driving the fall in the natural rate of interest, triggering the ZLB, might not be temporary after all but instead ones that would not necessarily revert themselves. The literature has identified several plausible candidates, such as demographic change, fall in productivity, global savings glut, rise in inequality and so on, essentially any force that might trigger the relative supply of savings and investment to be such that the natural rate of interest is permanently (or very persistently) negative. What was sort of interesting about modeling the secular stagnation hypothesis was that one needed to do both an open heart surgery on the aggregate demand side of traditional DSGE models (to allow for permanently negative interest rates) and also do a radical change on the supply side to allow for the possibility of a permanent demand recession (a big no-no in traditional macro which typically assumes long run neautrality). In any case, the bottom-line of this research, contrary to what Stiglitz appears to think, is that the secular stagnation hypothesis gives an even stronger case for aggressive intervention, e.g. in 2008. Thus far from being "just an excuse for flawed policies" the hypothesis gives a compelling reason to believe that more should have been done in 2008.

It is hard to end this little note, without responding briefly to Stiglitz's notion that events of last year have "put a lie to this idea" but Stiglitz suggests that the the fiscal expansion under Trump is responsible for some of the current recovery (a suggestion I will take for granted for the purpose of this argument, but one that could be contested). It is odd to suggest that current events put a "lie to the idea" of secular stagnation, for a recovery based on fiscal expansion is precisely the prediction of the secular stagnation theory: With low interest rates there is more room than usual for a fiscal stimulus, i.e. it is less likely to call for rapid offsetting increase in interest rates by the Federal Reserve than usual due to the absence of inflationary pressures created by the fiscal expansion (which so far, seems pretty much on the mark, as inflation remains subdued in spite of the large fiscal expansion). So here, Stiglitz, seems to get things exactly upside-down. At the end of the day, I suspect that the implication of a secular stagnation diagnostic of 2008 is -- after all -- very much in line with the view Stiglitz expresses here and often elsewhere, that the "fallout from the financial crisis was more severe, and massive redistribution of income and wealth toward the top had weakened aggregate demand" and that "the downturn was likely to be deep and long" and that what was needed was "stronger and different from what Obama proposed". Indeed, the secular stagnation hypothesis puts structure on those very arguments which I am quite sympathetic to. 



Finally a little figure and some reflections on the future. The figure above shows the cut in the Feds Fund rate in response to past three recessions as identified by NBER (in gray bars in the figure). In the early 1990's the Fed had room to cut rates from 10 to 3 (700 basis points), in early 2000 from 6.5 to 1 (550 basis points) and in 2008 from 5.25 to 0 (525 basis points). The secular stagnation hypothesis entertains the possibility that the observed fall in real interest rate (evident in the figure by the downward trend in long-term rates shown in red) over the past decades is "secular" which implies that come next recession, the Fed may have much less room to cut rates than it has had before, smaller than on the last 3 occasions. As much as I would like to hope that this will turn out NOT to be the case, it seems to me to be exceedingly likely the Fed may run out of room yet again when the next shoe drops, especially if the current recovery ends in tears in the next year or two as many now appear to be suggesting. With somewhat limited options to do monetary expansion at that point, this suggests we should be thinking hard about what fiscal policy can do the next time around. I suspect Stiglitz would agree on this point with those of us that have been entertaining the secular stagnation hypothesis.