Interview: Larry Summers, economist
The veteran policy advisor talks about his controversial past advice, the threat of inflation, and more!
Larry Summers is a sort of famous-economist-at large. Officially he’s a professor of economics at Harvard, and in fact he still does quite a bit of interesting academic research — on the returns to innovation, on declining worker power, and much else. But he’s most popularly known as a Democratic policy advisor. He’s been Chief Economist at the World Bank, Treasury Secretary under Bill Clinton, director of the National Economic Council under Barack Obama, and even president of Harvard University (where he famously was mean to the Winklevii and was involved in various controversies).
In recent years, Summers has increasingly come under fire for his judgement calls before and during the financial crisis of 2008-9, in which he downplayed the risks of financial innovation and suggested limiting the size of Obama’s stimulus. Earlier this year, Summers’ warnings to Biden to limit the size of the Covid relief bill went unheeded. Now Summers is warning about higher inflation and calling for tighter monetary policy, and the Biden Administration is trying to decide how much to heed his concerns.
In the interview below, we discuss those calls that Summers made in the past, as well as his current concerns. We also talk about his economic policy priorities and the culture of the econ profession.
N.S.: Usually, my interviews are about giving people space to lay out their ideas. But few have had as much space to lay out their ideas as you have! So let's start off with a challenging question here. In 2005, even as bad loans piled up in the U.S. financial system, you called Raghuram Rajan a "Luddite" for sounding the alarm about financial innovation. After the crash, a memo you wrote to the Obama administration is widely credited with having limited the ambition and scope of the 2009 stimulus package. Given these big mistakes in the past, what's your case for why leaders in the Democratic Party should heed your warnings on inflation and spending now?
L.S.: I don’t think my arguments, or those of others, should be judged ad personam. They should be judged on their logic and supporting evidence which we can get into.
This is not the place for a lengthy rehearsal of controversies over 2009 and I’m admittedly not unbiased. I am confident that any reasonable observer who reads the memoirs of or talks with President Obama, Rahm Emanuel, David Axelrod or other major participants will conclude that the constraints on stimulus proposed and enacted were political rather than based on economics. This view is supported by the fact that Congress cutback the Administration's proposals before passing them by a razor thin margin.
I made clear to the President, and my colleagues in 2009, that “the risk of our overdoing fiscal stimulus is conceivable like the risk of my diet making me anorexic. It is not the right thing to be worried about.”
I’ve been largely on the Keynesian side of economic policy debates for a long time—pushing hysteresis theories, arguing with Delong in 2012 BPEA that fiscal policies can pay for themselves, propounding the secular stagnation hypothesis, resisting Fed rate increases until they see the whites of the eyes of inflation and supporting new monetary policy frameworks and strongly supporting the CARES act. So my track record is as anything but an inflation alarmist.
Since the financial crisis I have quoted the line attributed to Keynes “when I see things I didn’t expect, I change my views—and you?” I wish we had been able to legislate something like Dodd Frank in the 1990s. It didn’t happen because of failures of prescience and the political context.
The line about Rajan’s view being Luddite that you quote was not in context a general assertion about financial regulation and stability. Rather I was asserting that a large part of financial instability derives from very traditional forms of finance, like mortgages, rather than primarily from financial innovations. Looking at the decades long agony of Japan, the experiences across Europe in the last decade and the centrality of mortgages in the US crisis, I think this caution was appropriate.
N.S.: OK, but now you're warning about sustained higher inflation. Is there a quantitative model that predicts that we're in danger of sustained inflation now, or is that a judgement based largely on your experience doing macroeconomic policy advising and observing financial markets? And to the extent it's the latter, doesn't your instinct become an important question?
Also, let's talk about the secular stagnation hypothesis. For years, you warned that the U.S. had entered a low-demand, low-growth equilibrium -- a trap from which the only escape was substantial fiscal stimulus. So...did we do it? Did the Covid relief bills bust us out of that stagnation, so that now we're back in a region of normal demand, where we have some sort of tradeoff between inflation and unemployment again?
L.S.: Fair questions. My arguments regarding inflation do not derive from a specific econometric model. Indeed I believe the problem with much of the conventional wisdom is that models fit to the last 40 years of data, where inflation has never accelerated, will almost definitionally yield complacent conclusions.
My judgement has been that there is a one third chance of significant sustained inflation, a one third chance that inflation would be contained by some kind of significant slowdown and a one third chance of a downturn. This rests on several pillars.
First, I looked in conventional ways at aggregate demand and supply. As of the beginning of the year, it looked to me like worker incomes were running $30 billion a month below trend and this was rapidly declining. Against this gap, we were adding a stimulus program that would spend out 150 billion a month or more, an especially easy monetary policy with the loosest financial conditions in history, and a savings build-up from rationed spending of 2 trillion mostly in liquid assets. The question to me was why, given standard views about inflation arising from demand and supply and reasonable multiplier estimates, this wouldn’t risk inflation.
I have not seen any economic analysis recommending a 2.8 trillion fiscal program from the perspective of last winter.
Second, I am influenced by historical analogies. The fact that the 2009 program was too small did not for me justify a program that measured relative to the prospective GDP gap was 5 to 10x as large. I was struck also by the similarity between all the arguments being made this spring and those made by the Vietnam era as inflation was baked in before the supply shocks.
Third, the data flow as I read it corroborates my concerns. Job opening and quit rates suggest record levels of labor tightness, as do surveys of employers about hiring. Unemployment and employment measures suggest some slack relative to pre pandemic NAIRU estimates but these are imprecise and record levels of structural change are likely to raise NAIRU estimates. We know wage acceleration is very sticky but it is coming.
The same can be said about housing, the largest part of household budgets. Whether the problem is lags or flaws in the indices I don’t know but housing has been pulling down measured inflation. In reality whether you look at ownership prices or rental prices, they have risen at a double digit rate for an item that comprises a quarter of household budgets.
Core CPI rose at a double digit rate during the second quarter. Of course most of this is transitory. But that was never the basis of my argument, it was entirely unpredicted by those seeking to explain it away, and leaves aside the question of whether inflation expectations are in danger of becoming unhinged.
By the way, transitory goes both ways. When in the next few months used car prices mean revert the CPI will be distorted down. How clearly this is acknowledged by the inflation optimists will be a test of integrity.
Expectations are I think in danger of being unanchored especially when so many are so intent on promoting demand as the central objective of policy.
Noah, I’ve gone on at length here. With respect to secular stagnation: with deficits running at double digit rates, continuing massive Fed bond buying and a record level of job vacancies we don’t have it. Much as Alvin Hansen was right about the economy of the late 1930s and then World War II changed things, everything macroeconomic has changed in the last 18 months.
I saw secular stagnation as emerging from a gap of perhaps 3 percent between Intended saving and investment, with normal deficit levels. No surprise that it’s blown out by double digit deficits.
For the medium term I’m not sure. The fact that consensus forecasts look to negative real rates, still very large-- though reduced-- deficits and an economy that is not overheating, suggests a return to secular stagnation or at least a major saving absorption problem.
I’m not sure how to think about the medium term or more broadly about the implications of very low real rates.
N.S.: OK, so suppose this inflation isn't a blip, and does persist. How high of an inflation rate should we be worried about, given the dangers Delta Covid poses to the economy right now? We do flexible inflation targeting, not absolute inflation targeting, right? Is 4% core PCE dangerous?
Also, what's the best tool for tamping down inflation? I've suggested that forward guidance -- basically the Fed reassuring people that it hasn't abandoned the Volcker stance toward inflation -- would be effective at keeping interest rates down so the government doesn't have to cut spending much. Do you think that's reasonable, or do you think spending needs to be cut? And if you cut spending, what would you cut spending on? The relief bill already passed, and most of what's on the table now is the kind of public investment you've been advocating.
L.S.: If it appeared that inflation expectations were coming to exceed 4 percent, I would be very worried that we were repeating the mistakes of the 1960s and 1970s. We would likely see a severe recession certainly with a real risk of further increases that would be a cause for serious concern. We might also see a stagflation period with relatively high unemployment. And likely there would be consequences for the Fed and the government’s credibility more broadly. I am struck that inflation contributed heavily to Richard Nixon and Ronald Reagan’s election.
I strongly doubt, at this point, that the delta virus economic dislocations can be effectively addressed with easy money. Many of the issues are on the supply side as, for example, with
parents not working. Others are narrowly sectoral, as with travel. In neither case, will general demand stimulus be efficacious.Likely it will generate inflation.
In general, monetary rather than fiscal policy should address inflation. I doubt though that rhetoric about inflation, while massive QE is ongoing, will be very credible. But if it raised real rates, that would be effective.
I think fiscal policy should be secondary as you suggest. To the extent it is to be used, I’d advocate raising taxes not scaling back public investments.
N.S.: So what do you think monetary policy ought to be doing right now? Is Powell being too slow to taper off QE, or to announce a taper?
Also, suppose monetary policy does bring inflation back down to target by the end of this year. Would that mean your worry earlier this year about the size of Biden's Covid relief bill were unfounded, because monetary policy always had the situation under control?
L.S.: As my recent column argues, I think QE is very hard to justify given the level of output and inflation currently and prospectively and its toxic impacts on inequality and financial stability. I think we should already be tapering and should be moving to taper as rapidly as we can, without panicking markets. I’d be surprised if the Fed could responsibly hold off on rate increases for more than a year though this depends on the data flow. In general, I find the Fed’s failure to acknowledge issues like housing price inflation troubling.
I don’t follow your second question. It’s premise of monetary policy controlling inflation seems almost impossible. Monetary policy in the conventional view and mine operates with a substantial lag of a year or more. There has been no tightening of any kind and indeed real rates have fallen and financial conditions have eased.
Perhaps you mean would my warnings have been unfounded if inflation gets back to target by the end of the year.? Yes, I would be surprised if that happened. But, no, my warnings might not have been unfounded. My argument back to February was about inflation arising next year as the economy overheated. If say used car prices fall towards the end of the year, pulling down measured inflation, that will hardly be grounds for serenity, if as I expect, there is demand pressure pushing up inflation as aggregate demand growth exceeds supply growth.
Your general question is a fair one. Anyone with a serious view should be prepared to say what events would call their view into question. If the economy enjoys rapid growth over 2021 and 2022, totaling say 10, and it is not generally agreed by the end of 2022 that we have an inflation problem, I will have misjudged the inflation process. If over the next year the economy is importantly constrained on the demand side without a spike in interest rates I will have misjudged the determinants of aggregate demand.
Perhaps the same discipline of specifying when you would doubt your understanding of events should also be applied to those serene about inflation. The Fed and the Administration have done much more forecast revising than they have acknowledging their errors in judging inflation.
N.S.: OK, so I think we've covered the inflation topic pretty thoroughly here; I'd like to touch on more general economic topics. Other than managing aggregate demand, what are your top economic policy priorities right now? If you were an economic adviser to the Biden administration, what would you be pushing the administration to do?
L.S.: I think the single highest return investment opportunity for the US right now is a major Marshall plan for covid containment. Beyond the moral case with millions dying around the world and the global economy losing trillions, there is the reality that no one is safe until everyone is safe because of mutations. There are the further considerations that there is no better way for us to garner international goodwill, contrasting ourselves favorably with China or assure the world is ready for the next pandemics that will surely come.
I think it’s reasonable to estimate that the global go forward cost of covid is 5 to 10 trillion. The expected costs of other pandemics over this decade are likely another 5 to 10 trillion. (The coronavirus risk is rising 5 to 10 percent a year according to the reinsurance industry.) Call the cost 15 trillion. If the US spent 100 billion over the next decade and it had only a 10 percent chance of reducing these costs by one third, the payoff would be 5 to 1.
In terms of ongoing investment areas, I’d highlight investment in R&D and in infrastructure. Work I have done with Ben Jones suggests an extremely high social rate of return to science. The inventors of quantum mechanics or PCs, monoclonal antibodies or far improved batteries capture only a small part of the benefits. It follows that without support investment will be way insufficient. I continue to think we underinvest in physical infrastructure and that it has a set of benefits like reducing trade barriers that comes from widening the scope for exchange
I think the agreements driven by Secretary Yellen on corporate taxation that move us towards stopping a race to the bottom rather than trying to win it. There is much more to do to increase tax progressivity by stopping abuses in areas ranging from estate taxes to Roth IRAs to charitable deductions.
Finally if I were in government I would be pushing for a trade strategy that recognized the importance of low priced inputs for the competitiveness of US exports. I’d also think,that forming economic alliances has to be part of a strategy for dealing with China.
N.S.: Gotcha. One more question. In the past few years there has been quite a lot of discussion about the culture of the economics profession. I've argued that the culture is changing, so as to be less aggressive, more based on data, less based on who can shout people down. Do you think that's true? And if so, what do you think of the change? Finally, what do you think needs to be done in order to bring more women and underrepresented minorities into the profession?
L.S.: I think there is no question that economics has become more empirical over time. And it’s range has increased into more social policy oriented topics. There has also been a big upwards drift in the extent of collaborative work.
All of this has reduced the cult of veneration of raw smartness that was more part of the profession when I entered it than it is today. That makes economics more welcoming and congenial for more people. That is progress.
Also I think economics has been affected by broader trends in academe. Brutal Socratic Paper Chase type dialogue is no longer acceptable in law and business schools. Medical residents can no longer be forced to work 100 hour weeks. Norms regarding the civil treatment of students have changed pervasively.
A significant factor may also be changes in computing technology. For many years empirical workers worked at a computer center of some kind. Usually the turnaround was much faster late at night. That kind of a locker room environment was surely off putting to many.
So, I think you are right that the culture has changed and for the better. I do think we need to keep changing the culture while at the same time recognizing that not all analyses are equally good and not all opinions are equally valid. Forceful argument contributes to scientific progress and the challenge is to maintain that strength of the profession without being off putting to any group. We have a long way to go.
I have tried hard over the years to give opportunities to women and minorities. It’s always been important to me when I hired RAs or chose collaborators that no one could doubt that they were chosen because of my admiration for their capacity not to meet a target. Without naming names I think I’ve been fairly successful.
I think we need to bend over backwards to not discriminate in any kind of recruiting. A student of mine was a driver of the “no interviews in hotel bedrooms policy” and I was very proud of her efforts. I’m sure there are more habitual practices like that that should be changed and we should seek them out.
We also need to focus on promoting generational turnover in a world where the fraction of women and minorities in current cohorts is much greater than in cohorts that graduated decades ago. That is why as Harvard president I emphasized promoting assistant professors rather than hiring people who had already done their most important work.
There is a lot that can, should and I hope will be done but I would question the idea that demographic groups have perspectives on the multiplier, wage premiums or the shadow price of carbon and I would oppose the explicit or implicit quotas that are pushed in some environments.