What is driving the cooling of the US labour market?
The US labour market is cooling down. The Fed acknowledges it and the statistics confirm it: the unemployment rate has increased 0.7 pps so far this year and in June job creation hit its lowest levels since 2021. Should we be concerned?
There is no doubt that the US labour market is cooling. The Fed acknowledges it and the statistics confirm it. The unemployment rate has increased 0.7 pps so far this year, reaching 4.1% in September, while job creation has slowed and in July hit its lowest level since 2021. In this article, we will analyse the causes of these changes by seeking to answer a key question: should the cooling we are beginning to observe in the US labour market concern us or does it respond to a normalisation following the post-pandemic turbulence? The answer is not simple, but it is of crucial importance.
The Bureau of Labor Statistics (BLS) produces two employment reports that offer different labour market metrics: the Establishment Survey, which is based on surveys of non-farm businesses, and the Household Survey. The former collects payroll data, providing information on hours worked, wages and job creation. This is where the well-known non-farm payrolls, the most widely used set of statistics for measuring job creation, come from. On the other hand, the Household Survey, which is similar to the Labour Force Survey published in Spain, classifies the people surveyed as employed, unemployed or inactive, thus generating the unemployment rate and the figure for the labour force. Each report has its limitations: the Household Survey includes farm workers, self-employed workers and domestic workers, but is based on a smaller sample. The Establishment Survey counts the number of non-agricultural jobs, rather than employed individuals, so it is possible to include several jobs that are held by a single person. The Household Survey, in contrast, counts each person only once, regardless of how many jobs they have. While the two reports have differed in the level of employment they have reported in 2024, the trend both suggest is similar. Here we will use the figures from the Household Survey, as they offer us a broader view beyond the data on payrolls and non-agricultural jobs.
The second chart clearly illustrates the imbalances in the US labour market since 2020.1 The gap between the demand and supply series serves as an indicator of labour market tightness: when demand outstrips supply, the market is said to be tight; when the opposite happens, it is cooling.
At the peak of the pandemic, job demand fell sharply, as did supply to a lesser extent. However, once the restrictions were eased, demand recovered rapidly, while supply lagged further behind. This led to significant tension in the labour market, which saw an excess in demand of 6.5 million jobs at the peak of the overheating in March 2022. At that time, there were almost two vacancies per unemployed person (the historical average is 0.65). The two main factors behind the weakness of supply were the fall in immigration, due to border closures during the pandemic, and the reduction in the labour participation rate, as a consequence of the so-called Great Resignation.2
- 2. For further details, see the Focus «Analysing the tightness in the US labour market» in the MR03/2023.
However, as supply recovered, the tensions in the labour market began to ease. In 2023 alone, the labour force grew by 2%, twice the pre-pandemic annual average, representing an increase of 3.6 million people (the total increase since March 2022 is 4.3 million). This is partly due to the end of the Great Resignation, once the savings accumulated during the pandemic were exhausted, and the rebound in migration flows since the end of 2020. The CBO estimates that between 2022 and 2023, net migration to the country was almost 6 million people: per year, that is more than double the annual average of the previous decade. While not all these migrants enter the labour market immediately, in 2023 almost 40% of the growth in the labour force came from foreign workers. The fall in the demand for labour of 0.3% (just 228,000 jobs) was negligible compared to this surge in supply. This increase in supply was key to filling the many vacancies, reducing the gap in the excess demand by 70%, to 1.9 million jobs, while the unemployment rate barely increased. This is particularly relevant given the context of high interest rates.
In 2024, the gap between supply and demand has continued to close. However, this time the improvement comes more from falling demand than from an increase in supply. In the last 12 months, the labour force grew
by 0.8% and the participation rate remained at around 62.7%, while demand fell by more than 1% year-on-year, the fastest rate of decline since 2021. This drop in demand is mainly due to the reduction in the number of job vacancies (of 18%), while the level of employment has remained relatively stable. The Establishment Survey (payrolls) points in the same direction, although it suggests a sharper fall in job creation: in 2024, the average monthly number of new jobs was 184,000. This is down from an average of 314,000 between 2022 and 2023 and is below the pre-pandemic average of 191,000. Both statistics suggest that the labour market is absorbing new entrants at a slower pace, which has been reflected in an increase in the unemployment rate.
It is worth considering whether this cooling reflects a normalisation or is a sign of recession. When the unemployment rate hit 4.3% in July, fears of economic recession were sparked by the signal provided by the Sahm rule. This historical pattern suggests that when the average unemployment rate for the past three months exceeds the three-month moving average of the last 12 months by 0.5 pps or more, we are on the verge of a recession.3 This is not a hard and fast rule, and the fact that in recent months the unemployment rate has set record lows could distort the message. In fact, Powell called it a «statistical coincidence» rather than a rule, as there have been episodes of «false positives», as seen in the third chart. A more in-depth analysis is needed that puts the indicator within the full economic context in order to reach a more precise conclusion.
In fact, other indicators suggest that the labour market is still in good health and that the cooling we are witnessing is due to a normalisation. For instance, the layoffs rate remains stable at around 1.0%, without any significant increases, and applications for unemployment benefits have not grown at a worrying pace either. Job demand has slowed, but has not collapsed. Moreover, an unemployment rate of 4.1% remains low in historical terms.
Our baseline scenario points to a moderation in economic activity growth towards its potential levels, but we do not anticipate a recession in the short or medium term. A slowdown in job demand falls within this moderation, but the risk is that it could cool down so much that it leads to a sharper increase in the unemployment rate. The Fed has been clear in its intention to avoid that scenario and has begun to relax its monetary policy, lowering interest rates by 50 bps and signalling a more aggressive approach than previously expected. It remains to be seen whether these measures will be sufficient to maintain the strength of the labour market in the face of the various risks affecting the economic outlook.
- 3. See C. Sahm (2019). «Direct stimulus payments to individuals». Brookings Institute.
