The big picture on inflation

One last dispatch in 2021

Next year will mark 15 years since the first stirrings of the global financial crisis and also, as it happens, since I began writing for The Economist. I started on a freelance basis, writing for the economics blog, at a time when Freakonomics-style papers were all the rage and the stakes of the arguments seemed incredibly small. But soon enough all of what was then called the blogosphere was trying to figure out what a CDO was and whether nationalizing America’s biggest banks was a good idea or a bad one. 

The decade that followed was a trying one, but for those of us fortunate enough to be writing about economics for a living, it was also a fascinating experience, not just watching but participating in the intellectual ferment, as policymakers and practicing economists and lowly keyboard jockeys tried to figure out what the hell had happened to the world economy and what we ought to do about it. Both conventional wisdom and policy practice changed, as a consequence of events but also because of that long-running conversation, and it was just possible to feel that in some small way one’s writing—one’s blogging, for christ’s sake—made a difference.

Things feel different now, for reasons I can’t quite identify. Maybe it’s the compressed timeline along which we’re operating, but contributions to the conversation from active academics seem to me to be fewer and farther between, and what we have gotten shows up in the form of twitter threads and op-eds rather than working papers. And I suppose in some way it feels as though we all of us are directing our comments at our respective audiences rather than at each other, which may again be a product of twitter and the substackification of the landscape.

But hey, maybe I’ve been part of the problem, trying harder to make forceful rhetorical points than to explain my thinking. So, let’s try a different tack.

My position is that inflation at current levels is a problem, and also that keeping inflation under control is the responsibility of the central bank, but that central banks should be slow and cautious in tightening monetary policy. I’m not convinced that the decision to taper asset purchases faster was a good one, and I feel strongly that rushing to raise interest rates by a full percentage point or more in 2022 would likely be a mistake. I take that view because it seems to me that even in the absence of rapid tightening inflation is likely to fall; because we don’t have a very clear idea what a neutral policy stance is likely to be a year from now and tightening too much would be much more costly than tightening too little; and because the balance of risks seems to me to be tilted to the downside rather than the upside. One thing informing this view is my belief that a sustained wage-price inflation spiral, of the sort the world experienced in the 1970s and which would likely take a nasty recession to halt, is unlikely to develop.

How do I get there? There are two main elements to my view on this. The first concerns how it is that a central bank does its job. The Fed has many policy tools available to it which can directly affect the supply and cost of credit, and it absolutely matters, to some degree, what the federal funds rate happens to be at a particular moment. But the game the Fed is really playing, the deep magic, is a matter of coordinating expectations: of participants in financial markets, first and foremost, but secondarily (and often via financial markets) of everybody else. That makes it sound like the Fed is pulling a confidence trick, and in a way it is, albeit it an important and (often) socially beneficial one. If we all save at once the economy will tank and if we all spend our savings at once inflation will skyrocket, and so the central bank’s role is to herd people’s outlook to a place where neither of those things happen, and instead people and firms spend enough to keep the economy growing about as fast as it can. It achieves this through a combination of what economists often call cheap talk and policy actions which reinforce the talk.

Now, one thing I think we all learned during the 2010s is that the Fed’s ability to Jedi mind trick everyone into spending an optimal amount is constrained. One significant constraint on this ability is the Fed’s own commitment to using its Jedi powers, but that limit interacts with the other constraint—structural factors influencing the behavior of savers and borrowers—in an important way. What I mean is, a credible central bank could probably jawbone a depressed economy into inflation and an inflationary economy into depression with sufficient ingenuity and dramatic flair, but at the end of the day these are buttoned up central bankers and they’re going to keep their behavior within certain social parameters no matter what the condition of the economy.

What matters for our purposes is whether the Fed has lost its capacity to coordinate expectations—to get everyone on the same page concerning where the economy is headed—without resorting to extreme policy measures. If you wanted to tell a story in which the outlook for inflation is worse than I think it is, one way you could do so would be to argue that the inflation rates we’ve experienced so far have been enough to shake market confidence in the Fed. If that were the case, then the interest-rate hikes the Fed has led us to expect would likely be insufficient to get the inflation problem under control, which would mean that inflation will accelerate by more than markets or the Fed expect and rates will eventually need to go higher still to get everything back on track.

Looking at bond markets, it strikes me as difficult to conclude that the Fed has lost credibility. Yields toward the short-end of the curve have gone up over the past year, but they haven’t soared, which means that markets don’t anticipate a ton of rate hikes to come—certainly not many more than the Fed has suggested it will deliver. And yet breakeven rates of inflation aren’t flying upward; on the contrary, five-year breakevens have dropped substantially from November levels. Now, markets may be forecasting the economy wrong; perhaps inflation will come in higher than we expect. In recent months, though, when inflation has surprised to the upside, market adjustments have been modest. In November, the figures on consumer price inflation came in very hot, well above consensus. In response, the five-year breakeven rate ticked up, but did not soar, while fed fund futures increased the outlook for interest-rate hikes in 2022, but not dramatically. Indeed, it seems to me that movements in breakevens in recent months overwhelmingly reflect shifts in realized inflation, and not some growing skepticism of the Fed.

Indeed, one could argue that an overreaction to inflation now, under unusual circumstances and while the economy remains well short of pre-pandemic employment levels, could undermine the Fed’s long-run effort to persuade markets that it can boost a depressed economy in a pinch, even when interest rates are at zero. Over the long 2010s, when millions of people suffered because unemployment remained well above sustainably low levels for a decade, it seemed clear that in the absence of sufficiently stimulative fiscal policy what the Fed needed to do to boost growth was to make a credible commitment to behaving “irresponsibly”: that is, to promise credibly today to tolerate inflation tomorrow. At current inflation levels you could reasonably say job done, but a panicked round of rate hikes which choke off the recovery before employment has fully recovered would put that achievement at risk. A lot of people seem keen to forget that a lack of credibility on the employment side of the mandate can be more damaging than a lack of credibility on the inflation side.  

But why should anyone think, given where we are now, that inflation will fall on its own or that one percentage-point increase in the fed funds rate should put the recovery at risk? The other element that we need to consider concerns structural factors influencing saving and borrowing. 

We can think of the global economy as consisting of a market for goods and services and a parallel financial market which matches savers and borrowers. Suppose that within this global economy there is a sudden increase in saving. In that case, both the goods market and the financial market are suddenly knocked out of whack: the former because there is too little spending to keep production capacity occupied and the latter because there is too little borrowing to provide savers with the assets they would like to accumulate. Something has to adjust to bring things back into balance, and this typically occurs through a shift in the return on saving (which is also the cost of borrowing), facilitated by central banks. So, the yield on assets declines, which means that some people want to save less while others want to borrow more. We end up with more assets to sell to savers, and more money in the hands of borrowers, whose spending in the goods market makes up for the shortfall in spending associated with others’ increased saving.

Now, since the early 1970s we have seen a secular decline in asset yields. This seems to have been a reflection, in part, of reduced appetite for investment—because governments spent less on public investment while private investment also declined in some sectors and got cheaper (because of falling tech prices) in others—which in turn meant that there was less need to borrow. But it also reflected an increased appetite for saving. For a variety of reasons, the distribution of global income shifted such that more flowed toward people and institutions which were keen to buy assets and less flowed to people and institutions which were keen to buy goods and services.

So there have been occasional energy shocks, for example, which shift money from households and firms that would have preferred to use their income to buy goods and services into the hands of oil exporters, which use quite a lot of their income to purchase financial assets. At the same time, populations have gotten older, which means that more income is being earned by people who are eager to buy financial assets to prepare for retirement, and relatively less by people who aren’t yet at that point in the life cycle. And, inequality has risen, which means that more money is flowing to people who can afford to buy lots of financial assets and less to people who would really like to be able to buy more goods and services. The result of this has been a long surge in asset prices and corresponding decline in yields, alongside a chronic problem of weak demand, as policymakers have continually struggled to make sure there is sufficient borrowing to clear both goods and financial markets. 

We were very much in this world in 2019, but the pandemic seemingly upended everything. Government borrowing surged. While some of the money raised by that surge was transferred to people who just turned around and spent the money on financial assets, some of it was also redistributed to people who spent the money on goods and services. Actually, though, they mostly spent the money on goods, such that goods demand skyrocketed, even as growth in the supply of goods slowed for a variety of pandemic-related reasons. 

What does all this mean for the future? So, one very important thing to note is that despite all of these dramatic changes, we do not appear to be in a world in which there is an excess supply of assets relative to desired saving. That is, the surge in inflation has not been accompanied by a surge in asset yields. Yields on government and corporate bonds are low, the earnings yield on the S&P 500 is low, house prices have grown faster than rents, and so on. This strikes me as important. Another story you might tell about why I’m wrong about inflation is that the saving behavior we observed prior to the pandemic has changed in some fundamental way, such that governments run enormous deficits forever even as more money finds its way into the hands of people and firms which want to buy goods rather than assets. 

This doesn’t appear to be happening, though. Governments don’t seem to have undergone a regime shift in their attitudes toward deficits; borrowing is poised to fall by a lot over the next year. Clearly, people who found themselves holding more money than they’d expected to hold over the past two years were keen to use a lot of it to buy assets. Maybe low unemployment will lead to a radical unwinding of inequality, which will change matters. I’m doubtful, but that’s a possibility. It won’t lead to a reversal of demographic trends, however. And I think that overall one should conclude that we have not after all left the old macroeconomic world behind.

That doesn’t mean that American stimulus didn’t matter, I should note. Domestic demand within the US economy clearly rose as a consequence of government borrowing and redistribution. But increased borrowing within the US was largely offset by increased saving elsewhere in the world. As a necessary consequence of this, America’s current-account deficit rose. Which is to say: global savings remain abundant, but patterns of saving have shifted in order to accommodate surging American demand.  

If global savings are abundant, though, then why is global inflation soaring? I think the answer is that inflation hasn’t materialized as a consequence of developments in the balance between saving and borrowing so much as it has materialized as a consequence of changes in the composition of spending. Over the past two years, people really seem to have developed an appetite for more house and more stuff to put in the house. That has come as an enormous shock to an economy in which services account for an overwhelming share of GDP. 

In the pre-pandemic world, people spent a lot of money paying others to do stuff for them: serve them drinks, do their nails, show them around the tourist sites, and so on. Since then, a vast portion of that money has instead been spent acquiring additional residential square footage and then obtaining things to place in that square footage. And that shift in spending placed enormous stress on production networks which simply weren’t prepared for the crush, and which have also faced period interruptions associated with the pandemic.

Based on the above, how might we expect matters to develop over the coming year or two? One possibility is that things will go back to (something like) the way they were before, and that much more money will begin flowing toward services. The benign way in which this happens is that the threat posed by the pandemic recedes, and we get a simultaneous shift in spending (which eases pressure on goods production) and increase in labor-force participation. That strikes me as a reasonable expectation, but if the labor force did not rebound very much then service-sector inflation could become a problem. Spending on services is largely spending on labor, so a surge in services costs (if that materialized) could quickly become a surge in labor income, which would push the world closer toward a wage-price spiral. 

It seems increasingly likely, though, that in some important respects we do not go back to the old normal. That may mean that we don’t see much of a rebound in labor-force participation, which would be disappointing. On the other hand, production of most goods can and will increase over time, and the economy is in general better at boosting productivity in goods-producing industries relative to service industries. Barring unforeseen shocks, we are very likely to find ourselves drowning in chips and cars by 2023, with disinflationary consequences.

Now one important exception to the supply-will-quickly-respond story is housing. Housing is also an area in which continued healthy demand for assets exacerbates rather than mitigates the demand problem (in that people with money to save may choose to invest that money in a home). If you want to tell a low-drama story in which inflation remains high, housing is where you’d look. Now it does happen to be the case that the Fed has tools which can be used to narrowly target housing. It could tighten mortgage-lending standards, for instance, and then meanwhile the federal government could stop putting tariffs on lumber and local governments could approve more construction. (Permitting, though, is looking relatively healthy.)

But ideally you don’t raise interest rates by all that much. You don’t do this because there isn’t a massive excess in borrowing relative to saving, because the economy is undergoing some significant structural shifts which are likely to occur faster and more smoothly in the context of healthy demand growth, and because questions of credibility don’t require it. You also don’t do this because the economy is global, because much of the world took on massive amounts of debt over the past two years, and because a little bit of monetary tightening in America could result in a massive rise in saving elsewhere as foreign governments and firms are forced into crash deleveraging. The Fed, very responsibly, acted with a concern for global conditions back in the spring of 2020, not least because ignoring those conditions would have resulted in nasty collateral damage to the American economy. If it acts now as though the US is a closed economy, things could get ugly very quickly.

On that note, there is another way that things could go, which would be far more worrying than any scenario described above. As I mentioned, the global economy has certain imbalances which show up both within financial markets and markets for goods and services. America is a net importer of both a large amount of savings and a large amount of goods (and a net exporter of assets). This means that if there were to be an episode of acute deglobalization, then America would suddenly find itself with lots of assets relative to saving and far too few goods relative to spending. In other words, if the global economy were to suddenly look a lot more like it did in the 1970s, then macroeconomic conditions in America would also look a lot more like they did in the 1970s: when both the 10-year bond yield and consumer-price inflation rose into double digits.

Now, that’s holding consumer demand in America constant, when in fact that kind of acute shift would probably be associated with an enormous surge in desired saving as everyone panicked about whatever was going on in the world: China invading Taiwan or a massive emerging-markets financial crisis or a new and worse pandemic or what have you. But it is a possibility of which we ought to be aware. Weighing up that possibility also makes clear that to get a regime shift in inflation, such that it remains high in the face of measured rate hikes, is likely to require a major shift in the way the world economy works. It was that sort of shift that put the world economy on its low-rate trajectory way back in the 1950s. And I suppose that what I’d like to see from people who have a different assessment of inflation is some sort of detailed story about how things have changed and why we shouldn’t expect the structural forces which have caused saving to pile up to continue working as they have over the past half century.

Ok! That’s going to be it from me until the new year. Here’s hoping you all have a safe and merry holiday season.