Skip to main content

18 October 2021

Former US Treasury Secretary and Director of the National Economic Council for President Obama, Lawrence H. Summers gives keynote discussion

Qatar Centre for Global Banking & Finance

Professor David Aikman was joined by former US Treasury Secretary and Director of the National Economic Council for President Obama, Lawrence H. Summers (President Emeritus and Charles W. Eliot Professor at Harvard University) for a keynote discussion at the International Finance and Banking Society 2021 Oxford Conference titled 'Financial System(s) of tomorrow'.

qcgbf lawrence summers
David Aikman, Director of the Qatar Centre for Global Banking and Finance with Lawrence H. Summers, President Emeritus and Charles W. Eliot Professor at Harvard University.

Lawrence H. Summers discussed the inflation outlook in the US and how overheating and inflation pose the greatest risk to the US economy. Watch the recording of the discussion below or continue reading to find out his insights into the currents macroeconomic outlook and its policy implications.


The topic of the day has been the inflation outlook both in us and around the world, where we've seen large increases in inflation in many countries this year. Just this morning, in the UK, the CPI inflation read is the biggest on record, going back to when the Bank of England was granted independence for monetary policy in 1997. You've been very vocal about your view that overheating and inflation are the predominant risks to the US economy. Central banks on the other hand, are pretty sanguine about this risk, as they view it as a transitory affair. Could you start by giving us your views and why you think this could be a more persistent phenomenon than central banks expect?


I think all of us, particularly when we're in policymaking positions, tend to want to make different mistakes rather than mistakes we've made before. Over the last 40 years, the dominant mistakes have been overly worrying about inflation and being too restrictive for too long. Those were certainly the mistakes that were made in the wake of the financial crisis. The inflation alarmists (of whom before now I’ve never been one) have pretty consistently been wrong. And inflation has been low and constant for nearly 40 years, which means almost any econometric approach will conclude that there's nothing that reliably predicts increases in inflation. Because if you regress a constant term on any set of variables, you're going to get zero coefficients. So, I think the intellectual environment, both in terms of the past experience of policymakers and the nature of the era that's considered relevant for statistical study, imparts a very substantial bias towards serenity. I think that there is a tendency, and you know a great deal of error in statistics, comes from excessive emphasis on a particular null hypothesis.


Something similar is true for policymaking. We tend to go with a null hypothesis which is what we thought before until it's been proven to be wrong, at which point it's often too late to respond optimally. It's like all of these tendencies are running through central bank thinking. I think they're reinforced by the fact that there was an immediate imperative of providing liquidity, providing aggregate demand, and providing economic support in the wake of Covid. People who have an imperative tend to lose it somewhat slowly. All of those things, along with the natural human desire not to inflict pain and not be seen as a pain inflictor, all of that contributes to it being natural, that if inflation were a problem, there might well be a tendency at this moment to miss it and to respond belatedly.


If I look at the US economy, it seemed clear to me that when the gap in terms of lost incomes was running at about $30 billion a month below the previous trend, it seemed to me obvious last January, that when we had a program that represented about $200 billion a month in stimulus, that was coming on top of a $2 trillion savings overhang. It was coming on top of record expansionary monetary policy. It seemed to me that we were headed for the bathtub overflowing. Demand very substantially exceeding trend or potential output, to the extent that there was supply impairment because businesses closed, there were early retirements, people couldn't get to work and certain sectors had to close. That only reinforced the bathtub overflowing by shrinking the capacity of the bathtub. So, it seemed to me that there was a presumption that the bathtub was going to overflow. Nothing that has happened since then has changed that presumption in my mind at all.


What has mostly happened is that we've had more evidence of more inflation more promptly than I would have expected, and so core CPI inflation in the US ran at a double-digit rate in the first and second quarter of the year. Of course, lots of that is transitory, so no one is suggesting that there aren't any transitory factors generating inflation. But the question is whether a norm is being established that is significantly above the 2% inflation target? That's what happened in the face of a much smaller fiscal expansion in the 1960s, setting the stage for the great inflation of the 60s and the 70s. Then, the swing in fiscal policy was a quarter of what it has been in the last years, so I think we have seen has been pretty rapid inflation.


Now, of course, it’s true that price increases across sectors are not even. But if you imagine that demand was being over expanded in a way where it collided with supply, what would you expect? You would expect that there would be some sectors where there would be major bottlenecks, and they would have particularly large increases in inflation. And that's kind of what we have seen. Is all this going to end? Is it going to end quickly? Is inflation going to return to normal consistent with a rapidly growing economy? Maybe, but it's interesting that the median inflation, the thing that strips out all the outliers that people used to emphasise, had its worst month in 14 years last month, so that's not so encouraging. The businesses I talk to and the surveys of businesses report that people now think bottlenecks will be part of their economic environment until mid-2022. But the two biggest markets in the economy, the labour market, the housing market, both it seems to me are cause for alarm. Nobody disputes that the labour market's getting tighter and tighter, and yet already we've got the highest vacancy rate we've ever seen in the United States and the lowest rate at which employers can fill vacancies that we've ever seen in the United States. That seems to meet presage wage acceleration.


Now we look at the housing market. House prices are up almost 20% year on year. If you look at new tenants moving into rented housing, they're paying 17% more than the previous tenant did. That tells you that we've got some rapid housing inflation, and that's the largest component of the CPI. It's not in the statistics yet, but either it's coming in the statistics, or it says something about the quality of the statistics. Nobody can really argue that housing costs aren't going up very substantially.


So I look, and I see more tightness in labour markets, more pressure coming from housing markets, continued presence of bottlenecks, more fiscal policy coming online, and no plans to raise interest rates until 2023 and a change in the Fed's doctrine from “we're going to remove the punch bowl before the people get plastered” to “we're going to remove the punch bowl only when we're absolutely sure that many people are getting plastered”. It seems to me that people are going to be revising their view about inflation expectations upwards. I could be wrong of course; demand could create its own supply. We could have some substantial interruption in economic activity from some other source. I don't think Covid is going to be enough to make me wrong, because at this point, more Covid is acting as much as a supply shock as it is a demand shock and therefore it could be a potential contributor to inflation.


So, I'm pretty worried about the inflation picture right now. It's remarkably reminiscent to me of the 1960s. First, there's denial that there's inflation, then there's the explanation that inflation's due to a set of special factors. Then there's an increasing view that maybe some inflation's okay, or at least inflation's okay relative to the consequences of trying to curtail inflation with policy, so I'm anxious. I have to say this is a financial conference. I have to say I'm quite surprised and puzzled that my concerns are not more manifest in the bond market. Break-even inflation rates have indeed picked up, but they haven't picked up that much, and I'm not with a completely clear explanation as to why that is.


You mentioned policy framework changes. The Fed changed its framework, and it's now committed to allowing inflation to overshoot two per cent, to make up for past undershoots. The motivation for this was all looking back to when we had deflationary pressures as the primary concern, the proximity of the effective lower bound and so on. Are you concerned, though, that this change risks baking in the Fed being behind the curve? Is it well equipped to deal with the risks that you see?


I think it's in general a mistake to respond as a policymaker to demands for clarity about future actions. It is easy to say that transparency is a good thing, it is easy to say that people want to understand, and it's good to give people more information, but the reality is that we live in a world of unknowns. We don't know what's coming, so it's very difficult to specify our reaction functions. So, setting specific targets, you have to balance the extra information and clarity that you may be providing against the risks to your future credibility or your future ability to optimise. When you make a statement of intention, and something unexpected happens, then you either have to carry through on what you said, which may not be what is best to do under those circumstances, or you may have to be in the position of sacrificing your credibility and explaining that something new has happened.


So, I think that a past generation of central bankers typified by Alan Greenspan understood what the Delphic oracles understood, which is that if people think you're omniscient and omnipotent, but you're really not, it's a good idea to speak with a certain oracular vagueness, rather than with a certain numerical precision. I think we have an era in central banking that is confused on that kind of point. I think the average inflation targeting was a device for providing credibility for the idea that policy would be looser longer, and maybe it contributed a bit to that. I'm not sure that rates at any point have been lots lower than they would have been in the absence of that policy. I think it doesn't make much sense when you have an instrument that operates with a year, or a year and a half lag, to say that you're only going to respond in response to events and not any anticipation of events. So, I'm not very enthusiastic about the new framework, though I understand the pressures that led to its adoption.


It's no surprise that with the increase in public debts we've seen that debt service ratios will be very sensitive to interest rates. QE has heightened that effect by removing maturity from private sector balance sheets and, you know, shortening in effect the maturity of the government's liabilities. What are your thoughts on the risks of an inflation spike spilling over into a public debt crisis? And do you think central banks have enough independence to take the necessary actions to head off inflation given what that would mean for public debt sustainability?


I think you are raising the set of issues that I think is most important for macro-financial researchers to be thinking about right now to illuminate policy debates. I don't pretend to have all the answers, and I don't pretend to have views that I won't have changed over the next year. I'd make these observations.


First, at the time the Maastricht criteria were propounded, real interest rates in Germany were in the 4-5% range, and nominal interest rates on 10-year bonds were in the 7-8% range. Almost any theory of borrowing would tell you that the reasonable and appropriate level of borrowing is attentive to the interest rate environment in some combination of nominal and real terms. So whatever one thought about fiscal policy some years ago, it seems to me that it's appropriate to have more expansionary fiscal policies today. I think that point may be insufficiently appreciated.


Second, I think it is reasonably clear that there are a set of structural factors that mean that, apart from extraordinary near-wartime events like the response to Covid, we have had a set of structural changes that have raised the propensity to save/raise perhaps the propensity to avoid risk and reduce the propensity to invest and to bear risk and that has led to substantially lower neutral interest rates.

I think all of that means that we, probably in this period, need to think about deficits somewhat differently than we have in the past, both because there's likely to be more sustainability than there once would have been and because there is more need to absorb saving. The issue of savings absorption, I think, was the central macroeconomic issue of the pre-pandemic decade. I think we mostly didn't recognise it, and therefore, we had an excessively protracted recovery relative to what we needed to have. I don't think it's clear where we're going to be in the aftermath of wartime fiscal and monetary support levels during Covid.


I think how much room we have to expand fiscal policy without pushing interest rates up is an absolutely crucial question. I don't pretend to have a settled view on it. I think it is striking how deficits have risen, and we have not seen corresponding increases in real interest rates.


Gita Gopinath has a recent column in the Financial Times that expresses concern that another “Taper Tantrum”-style episode would be disastrous for emerging markets and developing economies. I'd be interested in your views on how the Fed can begin to taper its asset purchases while mitigating that risk of huge spillovers for the rest of the world?


I read Gita's piece closely. I read everything she writes, and I have enormous respect for her and the IMF. I must say I did have the perspective slightly that, in their work, they want to make sure we don't have inflation because that would be dangerous. They want to make sure we don't have tapering because that could be dangerous. Having warned about all the risks, it seemed to me that the essence of policymaking is deciding which risks to emphasise, and which risks to place less emphasis on, not to simply point up all the various risks in this situation.

My own view is that whenever the Fed starts tapering, it will be the most heralded change in monetary policy in the history of the planet. They discussed planning the process of planning to plan, and then they discussed planning to plan. Now they're discussing planning. By the time it happens, it's hard to believe that it's going to set off any shock waves in markets. I'm not sure just how large the effects of QE have been on interest rates. So I think the largest risks to the developing world come from what has always been the source of risks to the developing world, which is the industrial world losing control over expanding suffering overheating inflation and having to hit the brakes hard. So it seems to me that delayed action that is much the greater risk for the developing world and emerging markets rather than overly accelerated action.


We have a lot of young economists in the audience, and I think it will be very interesting for those people working in macro to hear your views on what topics you would like to see more focus on going forwards…


When I was a policymaker, I read I believe every NBER working paper on a macro topic, except for any paper that used the term “DSGE”. My experience with papers that used the term “DSGE” was they derived very specific conclusions from premises that were no more than allegories. And so I found much more interesting papers that put forth a different interpretation of events or that pointed up a hypothesis or offered an explanation for a correlation rather than papers that were put in a methodological straight jacket of some sort. So that's one type of suggestion for researchers.


I think one of the very large phenomena of our moment is the extraordinarily low level of real interest rates both right now and in prospect, and it seems to me that that is pregnant with implications. What does it say about the social marginal productivity of investment? If you think that a safe real interest rate has something to do with the certainty equivalent marginal product of capital, that number is remarkably low. What does it say about asset pricing? How do those very low interest rates affect patterns of asset pricing? What are the causes of this extraordinarily low level of real interest rates? What does it mean for traditional thinking about corporate finance and debt-equity behaviour? How should we think about a world of high debt ratios and low debt service ratios in terms of our concerns about financial stability? What does it say about the implications for the capital markets functioning and efficiently allocating resources if interest rates are so low that debt service payments don't represent a significant barrier and so people in business can stay in business for a long time?


Am I right in my broad vision that the macroeconomic problem of the 1980 to 2010 period was the maintenance of discipline and the avoidance of the temptation to inflate and that the macroeconomic problem of our era is the absorption of all saving, in a world where there is much less demand for physical capital because so much IT can be produced so cheaply? Because so much of what is valuable is not tangible or is of a low mass? These seem to me to be very fundamental kinds of macroeconomic/macro-finance questions.


I think the other set of issues that I would point out is that the paradigm in which we have operated for the last almost two generations, has been a paradigm in which stabilisation policy is all about the second moment of output. What policy is about is about stabilising it because, of course, it's all about price rigidities, and nominal things don't affect real things in the long run and so you can go above, or you can go below. But what you're trying to do is to control the second moment. That's not why I went into macroeconomics. I went into macroeconomics because I thought a better stabilisation policy could avert busts without shaving off booms to an equal extent. I believe that the Great Depression didn't have to happen. I believe that the 2008 financial crisis didn't have to happen. And so, a set of theoretical and empirical constructs in which it is admitted as a logical possibility that better macroeconomic policy can raise the mean level of output and reduce the mean level of unemployment over time rather than merely affect the amplitude of the fluctuations seems to me extremely important.

In this story

David Aikman

Director of the Qatar Centre for Global Banking & Finance