Notes from an Inflation Event from May...
A not-quite-debate with Charles Goodhart
Opening Statement
Will there be inflation?
There will be inflation. Right now we are attempting to rejoin traffic at speed. Here in the United States we are convincing 10 million workers to rejoin the labor force. Here in the United States we are convincing 500,000 employers to employ them. Furthermore, we need to move 4% of the labor force across sectors because we are going to have a much more online and more last-mile distribution-heavy economy in the future than we had in the past. Significant relative wage differentials in favor of expanding sectors will be needed to pull workers in. Because of the psychological impossibility of cutting nominal wages in modern economies—this has been the case for a century or so—higher real wages in growing sectors means rising wages overall. Moreover, there are lots of reopening supply bottlenecks. Most of those bottlenecks will kind of show up as supply shortages and much longer delivery times: the world auto companies are not paying other chip customers for their places in line for the chips coming out of TSMC, Samsung, and company, but are waiting their turn, yelling for political help, and smiling because the chip supply shortage with available chips parceled out in proportion to demand has accidentally allowed them to collude and raise prices. Not all of the reopening bottlenecks are showing up as delays and shortages. Some will show up in higher prices.
And that will produce inflation.
How much inflation? We do not know. We have never tried to do anything like this before—so much, so fast, perhaps five times the scale of the largest previous structural adjustment. Assuming that relationships are linear is an extremely bad thing to assume when we are so far out of previous experience. There will be inflation for the next year and a half.
How persistent will the inflation be? Will it continue into 2023 and beyond? My guess, and it is just a guess, is that it will not. It took many repeated substantial shocks to de-anchor inflation expectations over 1965-75 to create the inflation we saw in the late 1970s. We are unlikely to have such a series of shocks. And in the absence of such a series that de-anchors expectations, the Phillips curve is and has been relatively flat.
Economists have been going back over the experience of the 1960s and the 1970s. The most impressive piece I have seen is by Emi Nakamura and John Steinsson, both of whom Berkeley stole from Columbia three years ago. They took a look at whether the Phillips curve was flat or steep in the United States in the 1970s and early 1980s. They came up with the somewhat paradoxical conclusion that while the Phillips curve was quite steep at the macro level, with relatively small changes in employment and unemployment leading to substantial changes in inflation, it was very flat at the micro state and the city level, with an excess supply or an excess demand for labor in a particular city or a particular state having very little impact on the prices of non-tradables and the wages of people in those states and cities.
This points strongly to the nationwide formation of expectations and the impact of expectations of inflation on inflation outcomes as the dominant factors. People mark up their demands for wages and their plans for prices according to what they think the average rate of inflation is going to be. And actual supply-and-demand pressure cause only small deviations from those plans. At the moment it looks like inflation expectations have for a generation been very stable—highly anchored. People see year-to-year movements in inflation, but they do not incorporate those movements into their expectations. Rather, they expect inflation to revert to its past mean value. Thus people have strong confidence in central banks. As long as that confidence in central banks lasts, the inflation we will see over the next year and a half will not lead to higher inflation in 2023 and beyond.
We will be burning rubber, leaving skid marks on the road as we enter traffic at speed. We will not boil over and melt our economic engine.
Can the Fed stop inflation? It definitely can. And I think it will: it has ever since 1979. Before then, it required a very strange concatenation of events for the Fed not to stop inflation from 1969 to 1979. I have been powerfully influenced by Arthur Burns's late-1950s presidential address to the American Economic Association. He contrasted the old business cycle in which governments did not take strong action to boost the economy out of a recession with the new one. In the old business cycle, labor union demands for wage increases were always greatly curbed by the fact that having high wages would make its members' jobs an obvious target when the inevitable big recession came. But, Burns said, we live in this new business cycle in which people expect the government to prevent large downturns, Thus there are no curbs on labor union wage demands. Inflation is then not a problem that can be solved by economic means. It has to be solved by political means—congress has to do it.
Moreover Arthur Burns thought, the independence of the Fed is a very important thing. It will not survive if Congress believes that the Fed deliberately triggered a deep recession. These all are the reasons that led Arthur Burns to make his 180-degree turn at the start of the 1970s, to oppose his long-time patron, the guy who had appointed him, President Richard Nixon, and begin calling for wage and price controls. That was a very surprising thing for a central banker to do.
But you needed more than Burns to get the inflation of the 1970s. You also needed Richard Nixon and John Connally, who broke the Bretton Woods system because, as their then-underling Paul Volcker said, they were both strongly in favor of bold action but had no preferences as to what or in which direction the action should be bold. You had to add a strong union movement both in the United States and Britain—less so in the rest of Europe—that did really not understand the stakes and did not realize that the continuation of a labor-relations system quite congenial to them depended on their willingness to exercise wage restraint and so assist governments keep inflation from rising.
Add in the 2.5%-point/year slowdown in productivity growth in the early 1970s.
Add in King Faisal of Saudi Arabia and Ayatollah Khomeini of Iran.
All of those things had to be working against the normal culture of central banks and inflation control for 1969-1979 to happen. I do not see the very same, a similar, very strange concatenation of events coming in the next decade. I think in the medium run of three to four years, Fed culture—which Jay Powell is pushing the limits of—will reassert itself inside the Eccles building. The Federal Reserve will do what it takes to keep inflation from continually overshooting its target—2%/year for the PCE deflator, which means something like 2.5%/year for the core CPI.
Slides:
Rough Reconstruction of Other Things I Said:
Let me offer a very brief precis of what I understood Professor Goodhart to be saying. He focused on the structural factors driving inflationary pressures. Times of hyperglobalization, in which the world's international division of labor changes and deepens, are times in which there are deflationary pressures—so many opportunities to lower costs are being grabbed. Times in which the world economy's global division of labor is more or less static tend to be times of inflationary pressures. Times in which demands on governments are relatively stable are times of neutrality. But times in which demands on governments undergo structural increases—inflation and seigniorage are ways of raising money, and inevitably as governments try to meet the commitments that their electorates demand they resort to them at some point at some time to some degree.
For example, during the Second Hundred Years, from the coup of William of Orange in 1688 on up through the defeat of Napoleon at Waterloo in 1815, the British government faced the extraordinary demands from the financing required for the British Imperial fiscal-military state. And in the years after World War II you had the expansion of the social-insurance state. Those were both sources of structural inflationary pressure. From 1980 on up to very recently, Goodhart is saying, hyperglobalization has been putting deflationary pressure on the world economy. That is now coming to an end. And the aging of society is going to put enormous structural-inflationary pressures on global-north governments. I agree with all of this.
But—even though it is always very intellectually hazardous to disagree with Charles Goodhart: people who do so usually wind up embarrassed and apologizing three years later—I find myself thinking that while these structural pressures on inflation are there, they are second-order. The first-order factors for the next generation are much more the monetary and the macro and the expectational. Why? Because in our age the inflation rate is an explicit target of central banks and governments in a way that it was not up until, say, 1980 or so. That makes a difference: central banks focus on neutralizing the structural factors, and are able to do so.
Even earlier, structural factors could be overwhelmed by substantial changes in monetary affairs. The discovery of massive amounts of extremely low-quality gold in the Witwatersrand made a huge difference, once that had triggered the construction of both a modern railroad from the Natal coast to transport it and a world-class frontier organic chemistry industry in Johannesberg to process it. The deflationary pressures of the late 1800s came to a sudden end, and from 1896 on the world economy shifted into an inflationary boom
We economists are stuck now where we have been since the late 1930s, when Ragnar Nurkse looked back over the preceding 50 years, and the concluded that much depended on whether expectations were mean-reverting or extrapolative. Countries in which people thought that exchange rates and inflation rates and price levels would return to normal had enormous freedom of action in their policies, and could conduct extraordinary successful economic stabilization campaigns. Countries in which expectations became extrapolative—ones in which a decline in the value of the currency now led people to think it is likely to decline further and we need to sell, in which people think that inflation now means not just inflation but higher inflation next year—had no good options. I think Nurkse's insight is true. Our problem is that we have nothing good coming from either psychology or economics as to when such expectational shift of régime takes place. We can point to individual historical examples: Franklin Roosevelt's inauguration, the Volcker disinflation, Francois Mitterand’s declarations that he was going to reflate the French economy which led to an outburst of inflation even before he was elected, and left him with no good policy options in the late 1970s.
One thought: we look at our financial markets, and we see that right now participation and engagement appears to be, to put it politely, highly elastic. The comments of Elon Musk on Saturday Night Live and of participants in Reddit forums can cause massive and immediate divergences between prices and fundamental values—divergences that shock investors and traders who I would have thought were very well-informed and cautious in their hedging. A market as "frothy" as the financial markets we seem to be building today on top of social media—those are ones in which extrapolative expectations, and thus loss of the ability to successfully stabilize, are more likely.
It took a decade, starting with Lyndon Johnson's calling William McChesney Martin into the Oval Office in 1966 to ask him not to raise interest rates just then, and ending with the very overoptimistic decisions of the Ford Treasury and of the Burns Federal Reserve after the 1975 recession that expectations were normal and that reflation was appropriate. I feel as though things are about as well anchored now as they were in 1965. That means that if we get through this wave by 2024, things are likely to go back to normal—except for all the structural factors that Professor Goodhart emphasized.
I would point out that there is an awful lot of inertia in nominal long rates. If there is a single regularity in the long bond market over the United States over the past 70 years, it is that it is very difficult to get bond traders as a group to very quickly revise their anchor points as to what these things should be selling for. During the run-up of inflation from 1965 to 1981, bond traders appear to have been at least half-oblivious to what was happening. After the Volcker disinflation, anchor points in bond markets kept long-term real interest rates at extraordinarily high levels for a full generation. I think Professor Goodhart is right: tapering is likely to be the first thing that centra. But even so, the nominal long bond rate is likely to remain substantially anchored for a decade—which would then lead one to expect a decade or so of even lower real long rates than we have now, hard as that may be to believe.
Yes, being a central banker is going to be very difficult over the next 10 years. But when has it not been very difficult? In the 1990s, when Alan Greenspan and company had only either good options or better options, because the world economy was stable and they had considerable help from the U.S. administration headed by Bill Clinton which was aggressively interested in long-run fiscal stabilization and solvency. Lloyd Bentsen and Robert Rubin were Secretaries of the Treasury, and were eager to make the Fed's job much easier, and worked hard and spend political capital to accomplish that. That was about the only decade in which being a central banker was a fun and a happy experience.
Yes, debt levels are—I won't say unprecedented, certainly Britain after the Napoleonic Wars had a much larger debt overhang in proportion to GDP—. But counterbalancing that is the extraordinary willingness of and in fact eager demand by investors around the world for safe assets, which they define as the liabilities of governments that possess exorbitant privilege on a large scale—issuing their own debt to add to the world stock of reserves, and then rolling over that stock at interest rates less than the growth rate of the economy as a whole. It is very close to being a free lunch for Britain, for the European Union, for Japan, for the United States—and possibly soon for China—to issue debt.
Normally, when you borrow or when you accept deposits, you have to pay people interest for the use of their money. But Britain, the EU, Japan and the US are instead, effectively charging their lenders fees for keeping their money and wealth safe and liquid.That is and has been the situation for the past 15 years. How long will this last? No one can tell. But as long as it lasts, the market economy is telling us that, worldwide, investors regard the debts of the countries with exorbitant privilege as extraordinarily valuable assets. In a market economy you have very strong incentives if you can—you become healthy and wealthy, and people think you wise if you do—create incredibly valuable assets.
What worries me is the extremely short duration of all of these governments' debts. I really wish that the treasuries were testing the market by attempting substantial increases in average debt maturity. I still do not understand why Janet Yellen isn't issuing both a nominal and a inflation-indexed consol right now, and seeing what the market for those things are. That would bring us a great deal of valuable information about what the available space for fiscal management is that we do not now have.
(Remember: You can subscribe to this… weblog-like newsletter… here:
There’s a free email list. There’s a paid-subscription list with (at the moment, only a few) extras too.)
Why in the world would anyone but a consol which was not indexed to inflation? Passively watching the value of your "perpetual" security eroded too nothing by even moderate inflation is a mug's game.