The US labour market has been tight since the Covid-19 crisis. The labour participation rate (62.3% in December 2022 compared to 63.3% in December 2019) and the employment rate (59.5% in December 2022 compared to 61% in December 2019) are still not back to their pre-Covid-19 levels. In particular, the rise in early retirements, for health-related reasons or driven by a lack of desire to continue working1 has reduced the number of active participants in the labour market, despite the strong job gains bolstered by the buoyant economic recovery in 2021. If companies initially hired back unemployed workers to meet their needs, which helped bring the unemployment rate down to its lowest level in 50 years (3.5% in December 2022), they soon faced hiring difficulties and labour shortages. As the Beveridge Curve shows (see chart 1), job openings remained dynamic while the unemployment rate was already close to (or even below) the natural unemployment rate2. In March 2022, there were up to 2 vacancies for every one jobseeker.
However, the 2022 inflation shock and the deteriorating US growth prospects from summer 2022 onwards started to dampen the labour market momentum, even if tensions have persisted. In December 2022, nonfarm payrolls slowed sharply but they remained relatively strong (see chart 2), particularly in business services (+212k over one month) and in the goods-producing sector (+39k).
The momentum in employment has supported wages and inflation
In order to meet their labour needs companies have increased wages to encourage certain inactive people (“unemployment halo”) to return to the labour market, as well as to encourage certain jobseekers to move into new sectors (even to train them), particularly in the hotel and catering sectors. Labour shortages have also prompted companies to offer higher wages to attract people already in employment (poach workers). This practice has resulted in an increase in employee resignations for new positions (“Great resignation”), which may have accentuated the shortages. Wages (average hourly earnings, AHE) in the private sector therefore increased by 4.6% year-on-year in December 2022, with a relatively homogeneous distribution between the goods (4.4%) and services (4.6%) sectors.
This wage dynamic was a significant supporting factor for household purchasing power, against the background of strong inflation, which is growth-supportive too. However, these wage increases have also contributed to higher inflation, as a number of companies have passed on the wage increases to their selling prices3. Although a wage-price loop, i.e. a reciprocal spillover effect between wage increases and price level increases, has not materialised in the United States (see chart 3), the fall in inflation will depend on the deceleration in wages, and therefore on the easing of the labour market.
Is a slowdown in the labour market possible without a recession?
The consequences of a slowdown in the labour market are being debated, in particular on the extent of the future fall in economic growth.
Further to the work of William Beveridge, the economists Olivier Blanchard, Alex Domash and Lawrence Summers4 show that the drop in the US job vacancy rate, as a result of the economic slowdown, should lead, as is usual, to an increase in the unemployment rate, which is likely to have a negative impact on consumption. The current position of the US labour market along the Beveridge Curve (top left on the chart) is the result of a strong recovery in economic activity and less efficient matching, which is explained by greater (job-to-job) reallocation as well as by recruitment difficulties. This situation implies that the natural unemployment rate has probably increased in comparison to its pre-Covid-19 level, which means that the US labour market is even tighter.
More broadly, economic history tells us that a drop in the job vacancy rate is highly likely to lead to an increase in the unemployment rate. Theoretically, for a decrease in the job vacancy rate not to involve such an increase in the unemployment rate, a decrease in reallocations and/or a decrease in hiring difficulties would be required. However, the Fed has no leverage on these parameters. If history is destined to repeat itself, the contraction in the economy would therefore be inevitable as the labour market deteriorates, whose turnaround could be rapid and substantial.
“Is this time different?”
The economists from the Federal Reserve (Fed) and the regional central banks are however putting forward a more optimistic analysis, emphasising the uniqueness of the current situation. According to Andrew Figura and Chris Waller5, the Beveridge Curve is more convex than before, which implies that a sharp drop in the job vacancy rate would result in only a small rise in the unemployment rate. According to their calculations, a drop in the job vacancy rate from 7% to 4.6% would only increase the unemployment rate by one percentage point. Moreover, this drop would leave a relatively high demand for labour, which would limit redundancies. Historically, redundancies are only significant when the demand for labour is low.
The authors also show that job openings mainly concern workers already in employment6 (see chart 4): so their disappearance would have only a limited impact on the unemployment rate. Let’s assume that there are two types of vacancies: those for unemployed workers and those for workers already in employment. This distribution of vacancies has in fact changed over the past decade. During the 2000s, job openings were evenly divided between the two categories, whereas since 2020 positions for employed workers have tended to represent a greater proportion than for others. In this context, monetary normalisation, while reducing the demand for labour, would mainly lead to the disappearance of positions aimed at workers already in employment and the reduction in these job openings would not have any significant impact on the unemployment rate.
While both arguments provide convincing ideas in relation to the consequences for economic activity of a slowdown in the labour market, nonetheless they remain hypothetical. What does the labour market tell us today regarding the possibility of the United States going into recession? Two indicators, one coincident and the other advanced, allow us to assess the risk of recession.
According to Claudia Sahm7 , a recession emerges when the three-month moving average of the unemployment rate rises by at least half a percentage point above its minimum over the past 12 months. According to the latest data available (see chart 5), the United States is not in recession8. The main downside to approaching recessions with this line of thinking is that it does not predict them and only indicates that a recession is occurring when an economy goes into one.
Based on the evolution of the smoothed unemployment rate, Thomas Mertens9 shows that a recession occurs when the series has bottomed out but its rate of change is rising. In his recent publication for the Federal Reserve Bank of San Francisco, he concludes that the US is also not facing an imminent recession; however, this could change in the coming months.
In short, if the labour market continues to slow down gradually as is currently the case, recession could be avoided but disinflation would then take longer, which the Fed would probably not want to see. On the other hand, a sharper deterioration in the labour market would more likely cause the American economy to fall into recession, the slowdown in inflation would be more abrupt and the Fed could ease the pressure more quickly. More than ever, the labour market will be key to growth and inflation in 2023.