The US labour market in January

On 3 February, the US Bureau of Labor Statistics (BLS) released its report on The Employment Situation for January 2023. The punchline of the report was that the US labour market remained strong and seemingly unaffected by the progressive and sharp tightening in monetary policy that the Federal Reserve started in March 2022. Strikingly, the unemployment rate fell to 3.4%, the lowest rate recorded since 1969, from 3.5% in December.

Several other key measures of the state of US labour markets remained unchanged (after removing the effects of some annual adjustments to the data). These include the labour force participation rate, which remained at 62.4%, and the employment-population rate, which stayed at 60.2%. Nevertheless, these ratios are below their pre-Covid levels of 63.3% and 61.1%, respectively. The US labour market has in that sense not fully recovered from the pandemic.

There was much attention focussed on the non-farm payroll data, which showed that 517k new jobs were created in the month. These data, which stem from the Establishment survey, were revised following the annual benchmarking process. The new data shows that the US labour market generated 300k more jobs in 2022 than initially estimated. 

Figure 1. Non-farm payrolls

Source: BLS. Data as of 03 February 2023.

The increase in employment by 517k jobs in January is exceptionally large. While the labour market is undoubtedly strong, it seems likely that part of this increase may be a statistical artifact.

One possible explanation is that the seasonal adjustment factors were updated in January. Perhaps surprisingly, seasonal fluctuations in many economic time series often dwarf the non-seasonal component that is of interest to commentators.

The figure below shows that employment (not seasonally adjusted) normally falls by about 2.5 million in January following the busy end-of-year season. Indeed, the variability of employment growth in the seasonally unadjusted data (as captured by the standard deviation) is about seven times larger than the variability of the seasonally adjusted data. A small misestimate of the seasonal factors can therefore have a very large impact on the seasonally adjusted data on employment growth in January. 

Figure 2. Non-farm payrolls

Source: BLS. Data as of 03 February 2023.

To see how critical the seasonal adjustment is, the figure below shows the difference between the seasonally adjusted and the unadjusted data for the increase in non-farm payrolls in January for each year between 2000 and 2023. 

Figure 3. Difference between the seasonally adjusted and the seasonally unadjusted data for January

Source: EFG calculations on data from BLS. Data as of 03 February 2023.

The seasonal adjustment involves estimating and removing a seasonal factor, which is uncertain and can vary over time. The figure shows that the estimated seasonal factor increased the January data by between 1.35% and 1.65% in the period considered. With total non-farm payrolls around 155 million, an error of 0.1% translates into a miscalculation of 155k of the estimated increase in non-farm payrolls. While there is little doubt that the US economy generated more jobs in January, it makes sense to be a little cautious about the exact increase in payrolls until more data are available.