Progress and Confusion

3. Rethinking Secular Stagnation after Seventeen Months

Olivier Blanchard, Kenneth Rogoff, and Raghuram Rajan
Published Date:
April 2016
  • ShareShare
Show Summary Details
Lawrence H. Summers

I salute Olivier Blanchard and the IMF for so open a dialogue at the conference on so wide a range of macroeconomic hypotheses. What I want to do in this chapter is talk about three things: I want to express why I think that the risk of secular stagnation is an important problem throughout the developed world. I want to contrast the secular stagnation viewpoint with two views that I regard as heavily overlapping—the debt supercycle view that Ken Rogoff has put forward and the savings glut view that Ben Bernanke has put forward—and explain why I think they’re very similar but, as far their nuances of difference are concerned, I prefer the secular stagnation view. And then I want to reflect on the policy implications of this general view of the global economy over the next decade.

What has happened in the seventeen months since I first broached the possibility of secular stagnation at another conference that Olivier organized?1 Interest rates have sharply declined in the United States, where quantitative easing was ended; in Japan, where it was continued; and in Europe, where it was initiated. And index bond yields remain remarkably low over quite long periods of time. Growth has remained sluggish, and growth forecasts have been revised downward to a substantial extent (table 3.1, figure 3.1).

Table 3.1Ten-Year Interest Rates, April 2015 and November 2013 (Initial IMF Speech on Secular Stagnation)

Figure 3.1Changes to Medium-Term GDP Forecast since October 2013, Five-Year Effects.

Source: IMF, World Economic Outlook, October 2013 and April 2015.

Real forward rates over the 2020–2025 period, which should clearly be after the hangover of the financial crisis, are negative in Japan and the euro area and barely positive for the United States (table 3.2). The so-called five-year indexed bond yields are negative or barely positive. And nowhere in the G-7 is market-expected inflation over a ten-year period likely to reach the 2 percent target.

Table 3.2Real Rates, 2020–2025 (%)

So even as we have moved a year and a half further away from the crisis, market pricing has actually moved in all the directions you would expect if you thought that there was a chronic excess of savings over investment (table 3.3).

Table 3.3Breakeven Ten-Year Inflation (%)
Note: *CPI inflation.Source: Bloomberg.

Now, Ken Rogoff argues the debt supercycle view, that the current weakness is the temporary result of overindebtedness.2 It seems to me that there is a pretty good way of distinguishing between one aspect of the debt supercycle theory and that of secular stagnation in what the markets are telling us. The debt supercycle view does not have a ready explanation for the low level of real interest rates, nor does it have a ready explanation for the fact that real interest rates have fallen steadily. In fact, real interest rates have trended downward over fifteen years regardless of whether measured for the United States through TIPs (figure 3.2), or measured globally by the IMF, or measured in almost any way you can measure (e.g., figure 3.3), and had already reached quite low levels in the aftermath of the last recession in 2003.

Figure 3.2TIP Real Wages Have Fallen Steadily. US TIPS Ten-Year Real Yield.

Source: Bloomberg.

Figure 3.3Along with Rates in the Rest of World. Rest of World Real Rate.

Note: 10 Year Yield – 5 Year Prior Inflation, weighted by PPP GDP.

Source: Bloomberg, IMF.

For instance, figures 3.2 and 3.3 show US TIPS rates and the average level of world real interest rates. If you looked at the trend from before the crisis in 2007 and extrapolated the trend to now, you would not be wildly off. And yet the whole debt overhang idea is that what is really defining life right now is that we were dealing with some overhang of the crisis, so if that were true, there should be no way to dictate the current level of real interest rates from stuff that happened prior to the crisis. And I would argue that it is closer to right to say that real rates are spot on the trend.

Ken Rogoff suggests an alternative hypothesis for explaining the low level of real interest rates, which is a generalized increase in the level of risk in the world. I agree with him that if there was a substantial and generalized increase in perceived risk, you might expect that to lead to a decline in real interest rates. You would also, however, expect it to lead to a decline, rather than an increase, in asset values, given that it was those assets that had become more risky. You would expect it to manifest itself in a measurable and clear increase in implied volatilities, as reflected in options markets. You would expect it to reflect itself in a dramatic increase in the pricing of out-of-the-money puts. But the opposite has occurred. It is far cheaper today to buy insurance against the Dow falling below 12,000 than it was a year ago, or two years ago, or four years ago. If one thinks about the market pricing of risk, it seems that risk has declined rather than increased. And it seems to me that the length of time that markets are forecasting low real interest rates makes the stagnation fairly secular or the debt supercycle very long, at which point the distinction blurs.

And what is the temporary debt overhang–induced headwind that is thought to be present in a major way today but that will be gone in three years? Corporate balance sheets are flush. The spread between LIBOR and other yields are low. Debt service ratios are at abnormally low levels. Whatever your indicator of repair from the financial crisis, it has mostly happened. And yet with interest rates of zero, the US economy is still likely to grow by only 2 percent this year. I do not see a good reason to be confident that the situation will be significantly better three years from now.

So, first, it is not like real interest rates are in some kind of extraordinary place relative to what you would have thought based on pre-2007 trends. Second, there is no obvious theory as to what the removable headwinds will be from this point on. And third, any debt overhang would itself be endogenous. Why did we have a vast erosion of credit standards by 2005? Why were interest rates in a place that enabled such bubbles? Because that was what was necessary to keep the economy going with adequate aggregate demand through that period. So even if a debt overhang were occurring, it would in a sense be a mechanism through which secular stagnation or oversaving produced damage. It is not an alternative to the idea of secular stagnation.

What about Ben Bernanke’s idea that there this a savings glut?3 The idea that there is a chronic excess of saving over investment (i.e., secular stagnation) and the idea that there is a savings glut kind of sound the same. That is because in many respects they are the same. As best I can tell, there are two questions on which Ben Bernanke and I have somewhat different judgments. One difference—where I would amend my views in his direction—has to do with the importance of taking into account the open economy aspects in thinking about this question of secular stagnation. I very much agree that a structural increase in savings that took place internationally would lead to the low real rates that I described. If I had my remarks at the IMF last year and in 2013 to do again, I would put more emphasis than I did at the time on the importance of the international aspects of increased saving relative to investment. But in either event, the fundamental reality is the same: an increase in saving leads to lower interest rates, which leads to a shortfall of demand.

In a forthcoming paper with Gauti Eggertsson, Neil Mehrotra, and Sanjay Singh, I further explore the open economy aspects of secular stagnation. We show that secular stagnation can persist on a global basis so long as the world real interest rate is below zero.4 When that is the case, countries with relatively attractive investment opportunities (but barriers to external investment) will be wary of increased foreign investment because capital inflows could cause their domestic interest rates to fall below zero, thereby risking domestic secular stagnation. Alternatively, a country mired in secular stagnation will be relatively better off if it spreads that stagnation to its trading partners. Put simply, secular stagnation is a contagious malady.

I think it is important for understanding the global economy and a crucial issue for US economic policy that Europe and to some extent Japan are indeed exporting their secular stagnations. They are exporting their secular stagnation by having a very low equilibrium interest rate, leading to capital outflows, leading to currency depreciation, leading to the shift in demand toward Europe and Japan and leading toward a shift in the shipping demand away from the United States.

Ten years ago, when Ben first started talking about a savings glut,5 there was a substantial and legitimate concern that countries were pursuing mercantile policies either because of the economic development advantages of running substantial trade surpluses or because of a desire to build up reserves so as never to have to deal with the IMF. This led to substantial capital exports. Today there is much less of that going on, with the notable exception of Germany. Rather, today many emerging market economies are each in their own way having substantial troubles, and the environment is much less attractive for capital than would have been expected five years ago in Brazil, in Russia, in China, and in South Africa. It is not so much more attractive in India as to offset all of that, and so for a different reason you are seeing a shift toward capital exports from emerging markets.

So, yes, the rest of the world is providing savings to the industrial world, and the rest of the industrial world is providing savings to the United States. That is an important part of the savings glut, and an important driver of secular stagnation.

The other difference is Ben’s suggestion that the savings glut is a relatively transitory phenomenon that will be repaired. Perhaps in the fullness of time it will. I think it is very difficult to read market judgments about real interest rates as suggesting that that scenario is likely. It is very difficult to read the IMF forecasts—with continuing downward revisions in medium-term growth—as suggesting that we are in a process of repair and healing, rather than suffering from a more serious problem. So, it’s my judgment that for the relevant medium-term policy horizon (as I have no useful views about 2040 or 2050) the challenge of absorbing savings in productive investment will be the overriding challenge for macroeconomic policy.

What are the implications of that view? First, there is a compelling case for pro-productive investment policy. Pro-productive investment policy raises supply and raises demand. When countries initiate it domestically it leads their exchange rates to appreciate and, therefore, has favorable international spillovers. It is a question of values and a question of specifics to what extent the right pro-investment policies involve public investment or involve incentives or the removal of barriers to private investment. But if one accepts this diagnosis, the most straightforward response is a focus on pro-investment policies. For the United States there is a compelling case, at a time when net federal infrastructure investment is zero, for an increase in infrastructure investment. And there is also a compelling case for a set of measures that would free up and stimulate energy investment in the private sector, and also a compelling case for corporate tax reform. The details will vary from country to country, but the clearest response and the one that avoids the most difficulties is a pro-investment response.

Second, monetary policies. I think it is a mistake to overdo the exogeneity of monetary policy. Here Ben Bernanke was very clear, and I thought exactly right, in his blog, in explaining that if monetary policy is seeking to regulate economies for full employment, the real interest rates that are set will respond to the broader structural factors that determine equilibrium real interest rates.6 I am worried about the view that if equilibrium real rates have to be very low, then monetary policy just needs to figure out how to make them very low with some combination of pre-commitments, revised inflation targets, and quantitative easing. I worry about that view not because I think doing nothing is better in the face of fiscal inaction but because I think the catalogue of risks associated with pursuing negative real interest rates is significant. The risks are to financial instability and for the creation of bubbles in an environment where interest rates are significantly less than growth rates, and for the international spillovers that may result from such policies. I relate to the concerns that the public sector will build infrastructure badly, and that that will be inefficient, and we should not be advocating that. But I do not understand why there is not a concern that the incremental investment projects that the private sector does not undertake now with a zero percent interest rate, but would choose to undertake with a negative 1 percent interest rate, will be projects that are of particular value. And I worry about the substantial impetus to financial engineering that hyper-low interest rates generate.

So my judgment is that whether we call it a savings glut, a debt supercycle, secular stagnation, or a quacking duck, we need to recognize the reality that the defining challenge is going to be absorbing all the savings in a satisfactory way in the global economy for the next decade. The first priority for policy should not be financial engineering in either the private sector or the public sector but a concerted effort to identify and find the means of financing the most productive investment opportunities globally.


Lawrence H. Summers, “IMF Fourteenth Annual Conference in Honor of Stan Fischer,” May 2015,

Kenneth Rogoff, “Debt Supercycle, Not Secular Stagnation,” VoxEU, Centre for Economic Policy Research, April 22, 2015,

See Ben S. Bernanke, “Why Are Interest Rates So Low, Part 2: Secular Stagnation” (blog post, Ben Bernanke’s Blog, March 31, 2015,, for Bernanke’s recent post on secular stagnation, and “On Secular Stagnation: A Response to Bernanke” ( for my response.

Gauti B. Eggertsson, Neil Mehrotra, Sanjay Singh, and Lawrence H. Summers, “A Contagious Malady? Open Economy Dimensions of Secular Stagnation,” forthcoming, 2015. For a prepublication pdf, see

Ben S. Bernanke, “The Global Savings Glut and the U.S. Current Account Deficit,” remarks by Governor Ben S. Bernanke before the Virginia Association of Economists, the Sandridge Lecture, March 10, 2005,

Ben S. Bernanke, “Why Are Interest Rates So Low?,” blog post, Ben Bernanke’s Blog, March 30, 2015,


    Gürkaynak, Refet S., BrianSack, and Jonathan H.Wright. 2006. “The U.S. Treasury Yield Curve: 1961 to the Present,Federal Reserve Finance and Economics Discussion Series 2006–28, 2006.

    Gürkaynak, Refet S., BrianSack, and Jonathan H.Wright. 2008. “The TIPS Yield Curve and Inflation Compensation,Federal Reserve Finance and Economics Discussion Series 2008–05, 2008.

    Other Resources Citing This Publication