Kartik B Athreya

Good afternoon, it's a pleasure to join you today at MBA First Friday. First, the disclaimer: The views expressed today are mine alone and not necessarily those of the Federal Reserve Bank of Richmond or the Federal Reserve System.¹

I'll stress three themes today. First, the past few years have seen unprecedented real economy shocks, which are particularly hard to deal with for any central bank, and so drove a lot of the inflation in the wake of the pandemic's arrival. My second theme is that in the longer run, this is reversed: Real forces don't create average inflation at all. Instead, central bank policy (almost) completely does. Third, the surge in broad inflation from late 2021 onward has driven a rapid and decisive normalization of policy by the Federal Open Market Committee (FOMC). And against the general health of the economy, especially in the labor market, this in my view offers a plausible path to a return to at-target inflation without sizable slowdowns in the U.S. economy.

Let's start in late 2021. Despite higher case counts, severe COVID-19 cases were not quite as terrible as the winter before — daily deaths were about half their 2020 level. The consumer was snapping up a whole lot of stuff — real personal consumption expenditure (PCE) was up 7 percent, and retail sales were up nearly 14 percent year-over-year. Indeed, the consumer was shopping so much so that supply chains were at a breaking point not just because of disruptions to shipping, port handling capacity and semiconductor manufacturing. Rather, people wanted stuff now, not later — a point I'll return to.

It was around then that we started recording sizable changes in prices that were, very importantly, widespread. For comparison, note that just before the pandemic in February 2020, only about a third of prices in the PCE basket were growing faster than 3 percent, and overall inflation was low. By March and April 2021, annual PCE inflation was very high at around 7 percent, but I'd note that less than half (40 percent) of consumer prices were growing faster than 3 percent. Later that year, though, from October to December 2021, inflation was still high, at around 7 percent, yet now the share of fast-growing prices had grown very large: to around 70 percent.

Transitory(!) Shocks and Short-Run Inflation

In an economy with generally well-managed inflation, in the short run (think month-to-month), inflation prints would be tightly linked to real demand and supply forces facing businesses and households that we could all point to (think supply chain and worker fears of COVID-19). These price movements help relative prices (think airfares to patio furniture) reflect changes in the true scarcity and value of some things relative to others, and hence are not to be prevented.

Now think about early to mid-2020: Inflation had been very stable for a couple of decades, and it made all the sense in the world that the prices of various consumer durables rose relative to the plane tickets.²

But when a whole lot of prices start to move at the same time, in the same direction, and in similar amounts, that's not changing relative prices. Instead, it reflects a monetary policy in need of normalizing. In other words, these are times where we can talk about "overall demand in the economy being high," and hence how rates need adjustment. Moreover, if left alone, such dynamics risk the formation of inflation expectations that drift in ways that can then lead to prices and wages moving up in even more sustained ways.

Yet, throughout this period, and even today, longer-run expectations for inflation have remained anchored. This is a major win, for its direct benefits of course, but heavily because it allows price-setters in our economy to react in a targeted way to the real "demand and supply" conditions in front of them and not worry about how to adjust against possibly economywide changes in pricing that would come with any broad inflation. In other words, when inflation expectations can stay anchored, the price system in our economy can do what it's supposed to.

To see this, imagine that future inflation became hard to predict. Now airlines and restaurants have to think about much more than demand and supply in their markets. They have to wonder about how (the many) others in the economy are setting prices and wages. This hinders them and the rest of us from locating prices that help equate demand and supply and makes life messy.

Bottom line: When real stuff happens, relative prices have to change to reflect scarcity unless we want a lot of disruption in daily life. When inflation expectations get unmoored, prices in the economy, especially in the wake of real shocks, get away from reflecting scarcity, and we all suffer. But expectations are not unmoored, so I view us as having a promising path forward.

The Humble Bureau of Weights and Measures?

If medium- and longer-term inflation expectations have remained stable despite enormous and unprecedented shocks, how did that come to pass? Here I'd say monetary policy since the mid-1990s has set us up well. Most of all, it reflects acceptance of the view that in the longer run, the central bank is a (and I'd say in most instances, the) key to longer-run inflation control.³ Through sustained effort, central banks had perhaps finally become the effective "Humble Bureaus of Weights and Measures for Prices" we might have all hoped for. They prioritized the stability of general price level, and the rest of us stopped thinking about it, and instead focused on the more fundamental forces operating in our lives as workers, consumers and businesses. Good outcomes for inflation and employment followed. And on the employment front, our Chair has noted that without price level stability, it is a priority that simply cannot be executed on.⁴

I view two specific things as especially important in getting us here: one big and strategic, and the other smaller and more tactical. The big-and-strategic thing is that we now tell the American public our inflation target, and strive to follow through in ways that convince them we mean it. The latter then leads to the smaller and tactical thing we do. Once inflation expectations are anchored at our target, we work to set short-term rates at a level that tracks the real economy — in a nutshell: higher in good times and lower in bad ones, and with a promise to react to broad inflation, should it occur.

In a perfect world, we'd have everyone understand that our plan was inflation at 2 percent, and then to follow up, we'd set rates consistently two percentage points (which is our inflation target) above the real interest rate that we think would prevail in a well-functioning market system. That latter interest rate has a name: the "natural" or "neutral" rate. It can't be directly observed though; it needs to be inferred, something we work at, but can do only with substantial uncertainty.⁵ Yet, in the quarter-century window from the mid-1990s to 2020, the Fed managed, evidently, to do this.

In fact, it managed this in a time where we saw systematic inflation in some sectors (services, growing at more than 3 percent per year), and outright deflation in others (durable goods, e.g., shrinking at more than 1 percent annually).⁶ That is, even as the relative prices of services compared to goods increased rapidly — around 4 percent annually — for two decades, this proved no barrier to delivering headline inflation at target. I stress this because it drives home the point that sustained inflation cannot come from steadily increasing relative prices, or real constraints, or really (almost) anywhere else. It can come only from central bank policy that needs adjusting.

High Inflation Today, and What I Missed

Alright, let me shift gears away from longer-run inflation control, which has worked well here and in most advanced economies, to the pressing question of how to deal with the problem in front of us today: high inflation that has been much broader, and that has now lasted much longer than plausibly attributable to one-off, or sector-specific, real shocks to the economy.

I'll first acknowledge that as an advisor to an FOMC member, I was slow to see this broadening of inflation. So, what could I have looked for in real time that I missed? Here I'm coming to the view that the September 2020 FOMC statement that policymakers would not raise rates so easily as the economy strengthened was, in hindsight, a more significant shift than I thought it was in real time.⁷ Additionally, I've revised my view of the power of the fiscal support that in the end was broader and longer lasting than I expected. Of course, as a central bank we do not judge, or opine on, fiscal policy. We instead adapt monetary policy given whatever those decisions turn out to be. So, while I was arguably late to the call, once evidence began to accumulate that inflation was no longer being driven just by one-off sectoral real supply and demand factors, it became unavoidable that normalization in monetary policy was needed.

Now recall that by mid-2021 and beyond, consumers were really spending and employers were really hiring. Pent-up demand aided by low rates and fiscal tailwinds were operating in full force. But supply constraints had simply not abated sufficiently — supply chains, and labor force participation, was simply not as elastic as it may once have been. The only way these facts could all be squared was if the real interest rate faced by U.S. consumers rose and convinced people not to do so much today, and instead put some spending off for tomorrow. In other words, somewhere in 2021 especially, the short-term neutral real interest rate likely went up.

But we didn't hike rates then, and were not really talking about a serious normalization either, so what gives? My read is simple: First, central banks, here and really anywhere, weren't faced with a cookie cutter shock. And neither we nor anyone else knew precisely how high the neutral real rate has jumped.⁸ So hiking rates sharply in the midst of a pandemic with all manner of cross currents wasn't so obvious.

Moreover, since longer-term inflation expectations didn't drift upward, I personally saw a good case to "look through" at least the initial phase of inflation — say until mid-2021. I also worried that when shocks that drive up the real costs of production hit us — i.e., of getting supplies, finding workers willing to risk COVID-19, and so on, monetary policy might face a real dilemma. An aggressive squashing of inflation would necessarily require tolerating a drop in real economic activity — and with it employment — all at a time when uncertainty about the future was rather high.

Still, something does have to give. If we don't hike, then real interest rates will stay low — and a flush consumer won't be incentivized to respect pandemic-driven scarcity; and businesses will start to hike prices to stay ahead of costs; and workers will want raises.⁹ All this is amplified if businesses and consumers start to worry about our commitment to inflation — which only further reduces the real rate of interest we all face now, making current spending incentives even greater, and so on!

In sum, severe real shocks hit our economy; monetary and fiscal policies were potent and stayed so for a long period; the consumer was ready to spend' and the real headwinds on the supply side (labor availability, supply chain, etc.) turned out to not be so transitory. So altogether, these changes lasted long enough to get businesses to expect high demand and high real costs of inputs and labor, and workers to seek higher wages. Pricing and wage changes then followed and delivered — at least in the proximate sense of the term — the high and broad inflation we've experienced for about a year now.

The TL;DR here is that supply shocks are just nasty to face as a central bank, and novel ones are even worse. And even if you could force inflation into a box, the harshness of such efforts might well be worse than the inflation itself, especially if you suspected — like I did initially — that the shock would be, ahem, transitory.

I'm also not sure that the need for normalization of monetary policy was obvious in, say, mid-2021. FOMC members have in fact conveyed this point. In a speech earlier this year, Gov. Chris Waller noted as challenges to kicking off an earlier normalization the considerable difficulty of forecasting during the pandemic — where public and private sector forecasters alike underforecasted inflation, and where preliminary labor market data turned out to be weaker than in subsequent revisions.¹⁰

What Next?

Ok, if inflation is the central bank's problem, and the events of the last three years eventually required policy normalization, what about the path ahead for policy? Well, here I want to say: A lot of work has already been done. By the same measure I've been using: the gap between the natural rate and the policy rate, we are not in accommodative territory by any measure. Given target inflation of 2 percent, a good estimate for the neutral nominal interest rate is about 3.3 percent (i.e., 1.3 percent plus 2 percent) versus today's effective funds rate of 4.3 percent. All this has been done promptly — 400 basis points of hikes in a year is nothing to sneeze at, as the bankers and real-estate-connected folks of Maryland will surely agree.¹¹

You will surely now be thinking: What about recession risk? I'll first note that we're always thinking about recession risks. Richmond Fed President Tom Barkin spoke at this event two years ago and tackled just that topic.¹² While I predictably agree with the thrust of those remarks, I want to note a couple of features of the current environment that make me quite optimistic that we will not experience a recession, but rather a slowdown in growth without a spike in unemployment — which is really the reason recessions matter to all of us: That's where the suffering is.

Here's why. First, if you want to win a race, pick a starting place near the finish line. That helpful starting place is the booming labor market we have, even a year into policy normalization. The unemployment rate is at 3.5 percent as of this morning, and there were over 10 million job openings, meaning nearly two job openings for every unemployed person. The quit rate remains at an elevated 2.7 percent as of November. Data from the Atlanta Fed shows workers still have plenty of reason to quit, with job switchers seeing 8.1 percent year-over-year wage growth, compared to 5.5 percent for job stayers. That 2.6 percentage point gap in November is the second highest since the pandemic started. Unemployment insurance claims remain near their very low pre-pandemic levels.

Second, FOMC members' communications show an essentially unanimous desire for policy tightening. In the latest Survey of Economic Projections from December's FOMC meeting, the median FOMC member projected the fed funds rate to rise further, peaking between 5 and 5 1/4 percent, with the lowest projection at 4.9 percent next year, compared to the 4.33 percent we're at today.

For both reasons, my current base case is not that of a recession, but that of slow path to inflation normalization. I don't expect inflation to jump down to target, as inflation inertia is a real thing, but rather expect a slow return, first to around 3 percent on a 12-month basis as 2023 comes to a close, and then further toward the 2 percent target. On the real side, I expect unemployment likely will stay below 5 percent, while GDP growth will slow but also stay away from negatives in 2023 and beyond. All this even as I expect the most interest-sensitive sectors soften further this year.

Must the Beatings Continue Until Morale Improves?

I want to close with the question of to what extent is restrictive — and not merely neutral — policy needed now? Well, first of all, our mandate requires us not to ignore the present high inflation in lieu of a future where it returns to target through a neutral-for-long policy.

But it's also possible to at least imagine a quick return to target inflation that doesn't bring harsh real costs with it. What if a central bank that understands perfectly what the neutral rate is announced that it would hike policy rates to neutral, but not beyond, and then manage them in that vicinity? If fully understood and believed, price setting by businesses would "jump" to the new path of lower inflation and stay there. No need for "restrictive" policy, and no pain! Just the Humble Bureau of Weights and Measures reasserting itself after an extremely strange one-time shock that temporarily dislodged inflation from our 2 percent target. But that's a whole of lot perfection and omniscience for any bureau.

Yet, we have done impressive things on this front in the past. I want to note here a recent essay by King and Lu in a volume we just released in honor of the eminent monetary economist Marvin Goodfriend.¹³ In it, the authors summarize a view of Marvin's: "…he indicated that the "successful [read: no recession] preemptive policy action in 1994 brought the economy to virtual price stability. Inflation and inflation expectations were anchored more firmly than ever before."

But maybe now, more than a year into high inflation, I'm being a bit wishful? After all, it's too late for preemption, and in the interaction of a central bank and the private economy, we must acknowledge that something — unprecedented shocks notwithstanding — that was not supposed to happen, well, did. We've got a 2 percent inflation target, and we've had around 5 percent inflation over the past year.¹⁴ So, we also need to consider the communication problem that creates — the credibility implications, if you will. And for this reason too, my view is that the moderately restrictive policy now in place is helpful.

To finish at the beginning, the past few years have seen unprecedented real economy shocks, and those are particularly hard to deal with for any central bank, and hence are primarily responsible for the initial inflation we saw in 2020 and much of 2021. But what we saw by mid-2021 crystallized the policy imperative to normalize, and the FOMC has since moved decisively and substantially. Fortunately, the labor market's robust state makes the path ahead less daunting than usual for a return to at-target inflation without sizable slowdowns in the U.S. economy. And fully granting that forecasting — especially about the future — is hard. This is my base case.

Thank you, and I am happy to take any questions.