Dom’s Quick Take: Labour Market Remains Strong Despite Data Riddles

Summary

• I do not see the discrepancy between household and establishment survey employment, lower hours, and lower temp work as signalling labour demand weakness.
• Rather, my preferred leading indicators of labour market weaknesses and recession include Sahm’s rule and real corporate profits per worker, neither of which points at recession.

Market Implications

• The Federal Reserve is unlikely to cut in March.

Payroll/Household Discrepancies to Fall Over Time

The discrepancy between payroll and household surveys for January employment was large, reflecting data quality and definitions (Tables 1 and 2). The discrepancy between payroll and adjusted household data, which is based on the household surveys but uses payroll employment definition, was even larger, suggesting data quality was more of an issue than definitions.

The discrepancy between payroll and adjusted household employment changes tends to mean revert to close to zero (Chart 1). For instance, the discrepancy was 993,000 workers in October 2022 and -947,000 workers in December 2022. By comparison, the 12-month average to October 2022 was 24,000 workers.

Finally, plotting the ratio of NFP to household surveys over time shows no obvious pattern associated with recessions (Chart 2).

However imperfect, NFP tends to be a more reliable indicator of employment than the household survey, largely because the NFP sample size is twice as large as that of the household survey and its confidence interval is two-thirds smaller. That said, NFPs are only one of many labour market indicators, and I discuss two more: hours worked and temporary work.

Decline in Hours Worked Could Reflect Preference for Part-Time Work

Hours worked have been falling, but this likely reflects workers’ preferences rather than labour demand weaknesses. A fall in hours worked has accompanied all recessions, but hours worked has been falling outside of recessions (Chart 2). Specifically, there was a long-term downtrend in hours worked during the ‘60s, ‘70s, and ‘90s.

Hours worked rose during the pandemic due to the shortage of workers and peaked in mid-2021. Hours worked have since been declining and now stand 0.1 and 0.2 hours for non-supervisory and for all workers, respectively, below December 2019.

The decline likely reflects a decrease in the share of workers working full time (Chart 3).

Because the decline in full-time work is matched by an increase in non-economic (i.e., voluntary part-time work), it likely reflects workers’ preferences rather than labour demand weaknesses. Part-time work for economic reasons (i.e., unvoluntary part-time work) has just recovered to the levels prevailing at end-2019.

Decline in Temporary Work Could Reflect Employers Preferences

Another set of data that could signal labour demand weakening is temporary work.  Temporary employment typically contracts when the economy starts slowing. In 2000 and 2007, the peak in temporary work coincided with the trough in unemployment (Chart 5).

This time, though, the relationship has broken down. Temporary employment is well below its pandemic peak and below end-2019 levels, yet unemployment has barely budged. This could reflect the tight labour market and changing workers’ preferences as well as employers’ wishes for a more stable workforce. Temp agencies report record numbers of conversions from temporary to permanent employees.

What Would Concern Me?

Having listed variables that are not currently a source of concern, here are indicators that would change my positive view of the labour market.

The main one would be an increase in unemployment above, e.g., 4.5%. This is because, historically, unemployment has increased either a little or a lot, with a lot associated with recessions. This is the essence of Sahm’s rule that says a 50bp increase in the unemployment rate relative to the 12-month low signals a recession. The value of the Sahm rule is currently 0.3 and falling.

I am unsure that 50bp will be the correct threshold in the next recession, but the rule is intuitively appealing because it likely reflects threshold effects and the feedback loop between the consumption and production sides of the economy.

Another set of data I monitor is corporate profits per worker. Sharp falls in real profit per worker tend to precede recessions (Chart 6). This is also intuitively appealing since what drives hiring decisions is ultimately profitability. Q4 corporate profits will become available with the second estimate of GDP on 28 February.

Other labour market data (e.g., JOLTS, weekly claims, consumer surveys, labour market indices), tend to be more coincident than leading indicators.

Dominique Dwor-Frecaut is a macro strategist based in Southern California. She has worked on EM and DMs at hedge funds, on the sell side, the NY Fed , the IMF and the World Bank. She publishes the blog Macro Sis that discusses the drivers of macro returns.

(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)