The U.S. Federal Reserve’s (Fed’s) job to tame inflation is far from done. That, at least, is what the labor market report for January quite impressively suggested. While benchmark revisions most certainly distorted the numbers and make it hard to compare them to previous reports, a drop of the unemployment rate to a post-pandemic low of 3.4% will not comfort central bankers in their mission to get inflation back on track to their 2% target.

Looking at modern U.S. economic history, there has rarely been such a high degree of tightness in labor markets. What we are seeing in the data right now translates into demand for workers outstripping supply by a factor of almost 1.9. That’s what you get when comparing the number of available jobs to the number of people unemployed. By that measure, here have only been a few occurrences in modern history that show remotely comparable levels of labor market tightness. Interestingly, they all happened when the U.S. economy was on a war time footing of some sort: World War II, the Korean War, the Vietnam War and the trade wars with China towards the end of Donald Trump’s presidency.^[1] Chart 1 shows that elevated labor demand always has played a crucial role in all those instances as the vacancy rate (as a percentage of the labor force) exceeded the unemployment rate (as a percentage of the labor force).

Chart 1: Times when labor demand outpaces labor supply are rare

Sources: Reproduction of Michaillat, Pascal, and Emmanuel Saez. 2022 (https://doi.org/10.48550/arXiv.2206.13012) by the use of data from Barnichon (2010), Petrosky-Nadeau & Zhang (2020) and Bureau of Labor Statistics (2022), DWS Investment GmbH as of 3/2/2023

 

In a classical war time economy, it might be straight forward to argue that parts of the available labor force are drafted to serve military needs directly or indirectly.^[2] At the same time, the production of war necessities picks up, thereby generating a demand overhang for labor. The first part of the explanation does not hold right now as we did not experience any major mobilization for action overseas. As in the early stages of WWII, however, there could be a deglobalization effect at work. Interestingly, the current trend started during the escalation of the trade war with China and was intensified after pandemic lockdowns ended while global trade was still muted. We believe that global matters, like peak globalization, ongoing trade disputes and corporation’s risk assessments to produce critical supplies at home rather than abroad played a major role leading up to the current labor tightness. But we are also interested in structural, domestic forces that may have contributed to imbalances in labor markets, not least as they relate to monetary policies.

Recent research indicates that one such domestic factor could be the massive financial support to U.S. households during the pandemic as this might have enabled low-income earners with few educational qualifications to switch to better jobs.^[3] Perhaps for the first time in their working life, many younger high-school graduates or drop outs, may have had a financial cushion that enabled them to wait for the right job with the right pay or to move to a different location, instead of accepting what was immediately within local labor markets. The consequence would be that especially firms that offer lower-paying jobs face a shortage of labor supply and are forced to pay higher wages to attract workers. This argument gains some support when looking at wage development since the onset of the pandemic^[4]: Jobs that earned below the average pay in the beginning of 2020 now pay round about 16.5% on average more while jobs above average pay gained “just” 13.4% on average. This number decreases further for the top 20% of income jobs to a gain of “only” 12.3% while the bottom 20% paying jobs gained 17.1% on average. On top of this domestic upgrading effect, mainly driven by younger workers switching jobs, the pandemic has accelerated demographic effects. Baby boomers are now retiring in droves. Together with less migration from outside the U.S., further intensified the supply situation.

On the demand side, however, it remains a bit puzzling what exactly caused such a drastic increase in jobs and job vacancies. One hypothesis on top of the peak globalization and job repatriation argument is that firms extrapolated their needs for workers out of the post-pandemic boom. While some of them currently revise those needs, they might still try to hold on to parts of their expanded workforce in face of a slowdown - it might be more expensive to re-hire them later. Another factor that most likely contributed to the elevated demand are options to work remotely, potentially lowering the costs that employees impose on firms and in general increasing the dynamic of the labor market. In specific locations and labor market segments, it is also quite plausible that during the slow but steady period before the pandemic, one or a few employers with significant market power may have been hiring less than they would have in a competitive market, to minimize their wage bill. Such monopsony effects, as economists call them, appear to have grown more significant at the lower end of the U.S. labor market in recent decades, but become harder to sustain in today’s more fluid situation. While it is quite hard to track those factors individually, indicators on labor demand do not yet indicate a fast easing as chart 2 shows.

Chart 2: High-frequency indicators for labor demand remain elevated

Source: Institute for Supply Management (ISM), National Federation of Independent Business (NFIB), Bureau of Labor Statistics, DWS Investment GmbH as of 3/2/2023

 

There are many implications of such a high degree of labor market tightness, and the various explanations outlined above. But they all tend to imply that central bankers must do more to re-balance the economy. So, when will the Fed know that it has done enough and switch to a “wait-and-see” mode markets had been hoping for?

The open jobs to unemployed ratio seemed to be one favored metric central bankers like to watch when attesting an “extremely” tightness of the labor market.^[5] While the concept is straight forward in interpretation, it lacks the ability to derive a definitive guidance for monetary decision making. This is usually done by comparing the current unemployment rate with its natural level, the so-called non-accelerating inflation rate of unemployment (NAIRU). Prevailing theories suggest, that if the actual unemployment rate is below the NAIRU, inflationary pressures are likely to build. The resulting difference is a standard input to most central bank reaction function models. A major drawback of this concept is that it does not incorporate short-term demand trends for labor and that NAIRU itself can change over time, along with structural changes to an economy. Estimates for NAIRU are derived out of mathematical economic models and therefore typically reflect deviations from a long-term theoretical output gap, which itself cannot be observed or measured directly.

Another, just recently established, metric by Michaillant and Saez seems to overcome this caveat: the so-called efficient unemployment rate.^[1] In contrast to the NAIRU, it incorporates the actual demand for labor as it is calculated as the geometric mean between the unemployment rate and the vacancy rate. For comparison: recent estimates for the NAIRU are around 4.25% while the current efficient unemployment rate ranges between 4.8% and 5.2%, depending on how narrowly unemployment is defined.^[6] Narrow unemployment in this context is defined as the unemployment rate that results if job losers on temporary layoff are excluded from the calculation. Wide unemployment in this framework is simply the representation of the usual U-3 unemployment rate that everyone in politics and in the media tends to talk about. And this marks another favorable feature of the efficient unemployment rate – it is flexible to adopt other interpretations of unemployment. For example, you can either look at wide unemployment or strip out temporary layoffs as in the narrower measure outlined above.

Taking the concept a step further, the resulting unemployment gap (that is the efficient unemployment rate minus the actual one) now can form the basis for a monetary policy recommendation. This is usually done by the use of multipliers. Based on estimates in the recent academic literature, one might suggest that the relationship between the efficient unemployment gap and rate hikes is around 0.5, meaning that the interest rates should be twice as much as the gap itself.^[7] Applied to the current situation this would imply a policy rate between 2.8% and 3.6%. We, however, believe this multiplier is too high. Current policy rates by far exceed those levels and are not sufficiently high to bring back the labor market into an equilibrium, e.g. closing the gap. Also, recent history suggests that a multiplier around 0.25 seems to be more appropriate (see Chart 3).

Chart 3: Federal funds rate as suggested by the efficient unemployment gap

Sources: Michaillat, Pascal, and Emmanuel Saez. 2022 (https://doi.org/10.48550/arXiv.2206.13012), Own Calculations, Haver Analytics, DWS Investment GmbH as of 3/2/2023

 

One explanation why the multiplier appears lower may be the high degree of liquidity that is still in the system. More liquidity has a positive effect on financial conditions, which in turn might dilute the effect of policy rates on the economy.^[8] The ultimate consequence out of this finding, if it turns out to be true, is that the current tightness of the labor market rather calls for policy rates in the range of 5% – 6%. For us, this assessment seems to be a realistic one and implies that the Fed indeed must further increase policy rates.

As of now we hold on to our terminal rate call of 5.0 - 5.25% but highlight that such a forecast most likely reflects rather the lower end of plausible outcomes. Inflation remains sticky as labor markets remain out of balance. At the same time, we still hold on to our narrative of a very shallow recession in 2023, although the current strength of economic momentum might delay this by a quarter or so. This gives the Fed room to increase rates further. Overall, the risk for inflation, and so the risk for policy rates, remains to the upside as labor markets remain extremely tight by historical standards.

Overview: key economic indicators***

	2023				2024
	Q1	Q2F	Q3F	Q4F	Q1F	Q2F**	Q3F	Q4F
GDP (% qoq, annualized)	0.0	-0.6	-0.2	0.4	1.6	1.4	2.0	2.0
Core inflation (% yoy)*	4.3	3.4	3.1	2.9	2.5	2.4	2.3	2.2
Headline inflation (% yoy)*	4.3	3.1	2.9	2.7	2.4	2.3	2.2	2.1
Unemployment rate (%)	3.7	4.2	4.4	4.6	4.8	4.8	4.7	4.5
Fiscal balance (% of GDP)				-4.3				-4.5
Federal funds rate (%)	4.75 - 5	5.25 - 5.5	5.25 - 5.5	5.25 - 5.5	5.25 - 5.5	5 - 5.25	4.75 - 5	4.5 - 4.75

*PCE Price Index
** Forecast
*** As of March 2023

Source: DWS Investment GmbH as of March 2023

Forecasts are not a reliable indicator of future returns. Forecasts are based on assumptions, estimates, views and hypothetical models or analyses, which might prove inaccurate or incorrect.

Past performance is not indicative of future returns.

Sources: Bloomberg Finance L.P. and DWS Investment Management Americas Inc. as of 3/1/23