By Alex Williams
2021 was a challenging year for the Federal Reserve. Between a complex labor market recovery and the first major inflationary episode in decades, the Fed’s new framework and forward guidance were both put to the test. Throughout, Employ America has advocated for the Fed to take positions that prioritize the labor market recovery, in terms of both employment and incomes, while embracing a credible, broad and inclusive definition of “maximum employment.”
To understand how the Fed is interpreting its “maximum employment” mandate in the current context, it’s worth going through Chair Powell’s remarks and Q&A for the December FOMC meeting. Chair Powell took a number of positions that Employ America has taken as to how Fed policy should be conducted and understood, while also falling into some of the traps we have flagged for how labor market data and monetary policy mechanisms should not be understood. Some of Chair Powell’s quotes have been edited here for clarity and brevity from the FOMC press conference transcript.
Wage growth - including that at the bottom end of the distribution - is not causing inflation:
“Imbalances related to the pandemic and the reopening of the economy have continued to contribute to elevated levels of inflation. In particular, bottlenecks and supply constraints are limiting how quickly production can respond to higher demand in the near term…as a result, overall inflation is running well above our 2 percent longer-run goal and will likely continue to do so well into next year…the drivers of higher inflation have been predominantly connected to the dislocations caused by the pandemic.”
“Wages are rising at their fastest pace in many years…but thus far, wage growth has not been a major contributor to the elevated levels of inflation.”
“Wages are not a big part of the high inflation story that we're seeing. As you look forward, let's assume that the goods economy does sort itself out and supply chains get working again, and maybe there's a rebalancing back to services and all that. But what that leaves behind is the other things that can lead to persistent inflation. In particular, we don't see this yet. But if you had a situation where real wages were persistently above productivity growth, that puts upward pressure on firms and they raise prices. It would take something that was persistent and material for that to happen and we don't see that yet.”
Since most calls for the Fed to tighten policy point to the elevated inflation we see today, it is imperative to understand where and why inflation is happening. The demand-side effect of reopening has caused inflation in services, while the bulk of realized inflation overshoot over the past 12 months have come from goods-oriented dynamics. This matters, because the proximate cause of goods inflation is not the domestic labor market. Instead, we are seeing prices rise in response to a lack of intermediate physical inputs - say, semiconductors for automobiles or container chassis for port throughput - or problems in the overseas portions of the global supply chain, including environmental controls on steel and aluminum production ahead of the Winter Olympics (“Olympic Blue”) and ambitious public health responses to COVID-19 outbreaks even at key points of the supply chain. These disruptions are problems of physical capacity, logistics and time, not domestic labor.
It is also important to keep three sets of inflation dichotomies separate here: supply-driven vs. demand-driven, narrow vs. broad, and transitory vs. persistent. Everyone has their preferred explanations for inflation, whether centered around improved risk preferences for services consumption (e.g. leisure and hospitality services / demand-side), around bottlenecks (durable goods, supply-side), fiscal transfers (all consumption, supply-side), or the pace of the labor market recovery (rent in particular, demand-side). The breadth of each dynamic varies, as does the degree to which each affects the supply side or demand side. Crucially though, these dynamics are unlikely to extend beyond the next twelve months, and become increasingly less relevant without the introduction of additional assumptions and assertions.
While price pressures may start out narrow, if those narrow price pressures hit goods that serve as intermediate inputs for a wide range of finished products, inflation can become broad-based. The semiconductor shortage tells a version of this story: first it was hard to get a new PlayStation, then it was hard to get a new car. However, with production scaling up aggressively, we are likely to see an end to this dynamic as well.
Even the more reliably persistent components of inflation - housing especially - are really a function of the growth in aggregate incomes. Rent inflation is more functionally driven by the growth rate of aggregate demand (as proxied by aggregate labor income) than it is by construction wages.
As job growth likely slows, aggregate labor income growth, no matter how you measure it, is likely to slow down in 2022 vs 2021.
Inflation alone is not a sign that the economy has exceeded “maximum employment”:
“The inflation that we got was not at all the inflation we were looking for or talking about in the framework. It really was a completely different thing: it had to do with strong monetary policy and fiscal stimulus in an economy that was recovering rapidly. And that was happening in an economy in which there were these supply-side barriers, which effectively led to, in certain parts of the economy, what you might call a vertical supply curve. Automobile purchases are very interest-rate sensitive and you would think demand would drive up the quantity of cars, but it can't, because they don't have semiconductors. So that was a very different kind of inflation. This is not the inflation we were looking for under our framework at all.”
Many mainstream models of monetary policy rely on a “divine coincidence” - the assumption that the point at which inflation begins is the point beyond which employment gains cannot be made - which Chair Powell implicitly rejects here. We push back on this idea at length in multiple parts of this piece, most especially the section on commodity-driven inflation in the mid-2000s. We are pleased to see Chair Powell agree that the presence of inflation does not indicate that the recovery has finished. Inflation can spike or fall for a variety of reasons, and only a few of those reasons directly involve labor market dynamics or financial conditions. Inflation is nearly 7% year-over-year in Spain right now, but explaining that inflation primarily through monetary or labor market factors would be extremely difficult.
Unobservable “neutral rates” are a poor guide for contemporaneous policymaking:
“We'll make policy based on what we're seeing in the economy rather than based on what a model might say the neutral rate is. We all have the experience of the last cycle, where we all, through that cycle, were trying to estimate what the neutral rate was, and it turned out -- you know, I think we learned a lot from seeing what happens. We ended up cutting rates three times after raising them to two and a quarter to two and a half percent.”
We applaud Chair Powell’s empirical data-oriented approach in contrast to reliance on models of the neutral interest rate. While raising interest rates broadly slows the economy and lowering interest rates allows it to move faster, the precisely optimal “neutral” rate is unobservable and changes over time. Worse, the impact of interest rates themselves change over time, as different economic structures react differently to changes in interest rates. Setting the course for monetary policy by an unobservable r-star, rather than observable facts about labor markets and financial conditions, risks unnecessary policy mistakes. The Fed seems set to tighten policy and that will marginally decelerate the pace of spending and income growth. The precise elasticities and net effects are far less clear and are always dependent on context.
“Maximum Employment” is a complex quantity that can’t be easily described by a single indicator:
“If you look at our statement of longer-run goals in monetary policy strategy, maximum employment is something where we look at a broad range of indicators. And those would include, of course, things like the unemployment rate, the labor force participation rate, job openings, wages, flows in and out of the labor force in various parts of the labor force. We'd also tend to look broadly and inclusively at different demographic groups and not just at the headline and aggregate numbers.”
“If you want to look at maximum employment, you look at prices and quantities, and the main price you look at is wages. And, look, you could get to 20 different indicators, if you wanted to, easily, but labor force participation, the unemployment rate, different age groups, prime age labor force participation in particular gets a lot of focus, the JOLTS data get a lot of focus. Wages are really one of the great signals, the quits rate is another one.”
Throughout 2021, we at Employ America often returned to the idea that “maximum employment” is a complex concept that cannot be expressed as a single value for a single indicator. While it is heartening to see Chair Powell acknowledge that the Fed considers a range of indicators, we would greatly prefer for the levels of each indicator associated with “maximum employment” be made public. Chair Powell has repeatedly indicated the value of the Prime-Age Employment-Population Ratio (EPOP) and Employment Cost Index to the Fed in understanding labor markets. Publishing near-term and long-term goals for each of these indicators would be a step in the right direction.
Labor market variables - most especially the Labor Force Participation Rate - are pro-cyclical, rather than fixed:
“The pattern in past cycles has been that labor force participation tends to recover in the wake of a strong recovery in unemployment, which is what we're getting right now…The reality is, we don't have a strong labor force participation recovery yet and we may not have it for some time.”
In 2020, we warned about the possibility of a slow recovery following the initial shock of job loss from the pandemic. Thankfully, the scale of fiscal stimulus and the monetary policy response were largely sufficient to prevent the slog from being as bad as it could have been. More recently, we have also emphasized how important it is to recognize that the Labor Force Participation Rate, especially after adjusting for age, can be cyclical. Prime-age labor force participation and employment rates can trough at far lower values for disenfranchised groups, which Chair Powell acknowledged elsewhere in his testimony.
Where we would disagree with Chair Powell is in his repeated use of arguments that implicitly link inflation to pandemic-driven issues with “labor supply.” Chair Powell uses the Labor Force Participation Rate - a survey-based measure prone to misclassification pitfalls even in the best of times - as a proxy for labor supply, while suggesting in multiple places that insufficient labor supply is hampering the supply-side from recovering. Making this suggestion undercuts the overarching theme that inflation is not primarily a labor market phenomenon.
Chair Powell’s insistence on the Labor Force Participation Rate as a proxy for labor supply also strikes a sour note. In talking about this cyclicality, it would improve the Fed’s communication strategy to replace LFPR with less ambiguous measures, like the prime-age employment-population ratio. Despite weirdness in LFPR, prime-age EPOP shows a clear run of strong job growth, especially when compared to other recent recoveries.
“Maximum Employment” is liable to be at one level in the short term, but achieving that level does not preclude higher levels of employment in the long term:
“The thing is, we're not going back to the same economy we had in February of 2020. And I think early on, the sense was that that's where we were headed. The post-pandemic labor market and the economy, in general, will be different. The maximum level of employment that's consistent with price stability evolves over time within a business cycle and over a longer period, in part reflecting evolution of the factors that affect labor supply, including those related to the pandemic.”
This was a central point to our report on the stance that Fed policy should take in quantifying and explaining the level of maximum employment. The pace of labor utilization gains consistent with the Fed’s near-term inflation goals may be slower than the current pace of improvement, but that does not mean we should abandon the recovery. As the supply challenges outside the domestic labor market are given a chance to abate, we are likely to see an opportunity to push for faster labor utilization gains that can still be made consistent with a 2% longer run inflation target. Crucially, there is no fixed level of labor utilization that defines “maximum employment” once and for all.
The disparate impacts of monetary policy on disadvantaged groups must be taken into account when setting policy:
“The recent improvements in labor market conditions have narrowed the differences in employment across groups, especially for workers at the lower end of the wage distribution, as well as for African Americans and Hispanics.”
“[When judging the distance to maximum employment] we would also tend to look broadly and inclusively at different demographic groups and not just at the headline and aggregate numbers.”
In many places we have emphasized the fact that macroeconomic and monetary policy can have disparate impacts on disadvantaged groups, along with the fact that recoveries have to go on long enough to extend their benefits to these groups. It is heartening to see that Chair Powell is paying attention to these metrics, however, we would like to see them play a larger role in the Fed’s communication and justification of future hikes.
However, it has been equally disappointing to see Chair Powell respond to repeated questioning about the set of indicators the Fed relies on with a different grab-bag of measures each time. If the Fed is indeed relying on a fixed set of indicators, these should be communicated consciously and deliberately. Throwing together examples on the spot suggests that the Fed is not as committed to understanding and communicating the level of maximum employment as Chair Powell’s February speech about “broad and inclusive goal[s]” would suggest. Instead, the Fed should clearly communicate which indicators matter the most to its policy stance, and provide critical values for those indicators.
Monetary policy works on financial conditions almost contemporaneously through expectations, rather than with long and variable lags:
“I do think that in this world where everything is -- or the global financial markets -- are connected together, financial conditions can change very quickly. My own sense is that financial conditions affect the economy fairly rapidly, more quickly than the traditional thought of, you know, a year or 18 months. When we communicate about what we're going to do, the markets move immediately to that. Financial conditions don't wait to change until things actually happen. They change based on the expectation of things happening.”
We applaud Chair Powell for emphasizing the fact that Fed policies work nearly instantaneously because they work primarily by altering financial conditions. Asset prices move quickly when markets absorb news of potential changes in Fed policy, as that news changes the expectations market participants hold about risk sentiment and the discounted value of future cashflows.
With that being said, Chair Powell could stand to be clearer about the labor market tradeoffs involved in adjusting policy. The goal of tighter Fed policy is to decrease businesses’ willingness to spend, including on expenditures for both capital and labor. Changes in Fed policy work quickly, but meaningful tightening slows down inflation primarily through first slowing down business spending and household income growth; inflation generally tends to lag cyclical production and spending dynamics. By the time tighter policy influences demand sufficiently to offset price pressures caused by supply bottlenecks, that tighter policy may prematurely stall the labor market recovery. Or it may be that supply chain issues work themselves out along the same timeline as Fed policy, and we may see more aggressive disinflation than was intended by policymakers. In any event, cooling inflation by tightening Fed policy is both a bank-shot and one that necessarily involves slowing down the labor market expansion to be effective.