The January jobs report defied most expectations and showed nearly half a million new workers were added to payrolls last month. Substantial re-estimates to the labor surveys also showed a workforce that is somewhat more robust than was previously assumed. Employers in January added 497,000 jobs, with private-sector payrolls gaining 444,000 jobs, while the unemployment rate rose to 4.0 percent. The labor force participation rate rose to 62.2 percent.

Here is a brief summary of the major economic indicators since the last jobs numbers:

• The Producer Price Index for final demand increased 1.0 percent in January;
• The Consumer Price Index increased 0.6 percent in January;
• Real average hourly earnings increased one cent from December to January;
• Orders for durable goods (including defense and aircraft) increased 1.6 percent in January;
• New home sales decreased 4.5 percent in January;
• The Price Index of U.S. imports increased 2.0 percent in January;
• ISM Services Index decreased 3.4 percentage points to 56.5 percent in February;
• ISM Manufacturing Index increased 1.0 percentage point to 58.6 percent in February;
• Consumer Confidence Index decreased 0.6 points from 111.1 to 110.5 in February;
• ADP reported private sector employment increased by 475,000 jobs in February.

Gordon’s Guesstimate: February Jobs

From time to time, it’s worth stepping back and considering how the Employment Situation, or “jobs report,” is discussed. The monthly report from the Bureau of Labor Statistics (BLS) is about 40 pages of narrative, tables, charts, footnotes, and addenda. As often as not, however, it is boiled down to a single or handful of indicators such as jobs created or the unemployment rate. These indicators are very much like a set of vital signs – they provide narrow glimpses in time, but when combined, illuminate a subject.

There are significant risks in relying on any single indicator to assess the state of the labor market. For example, from May to June 2020, over 4.5 million workers were added to the employment rolls – by far the largest monthly “jobs gain” in history. But of course, this was 2020, so there was a bit more to the story. To be sure, the substantial hiring in June of 2020 meant that millions of Americans went from unemployment to employment. But at the same time, 11 percent of the labor force was out of work, among the highest unemployment rates ever recorded. The monthly change in employment is a measurement of a labor market flow – a measurement of the change in employment over the course of an intervening month. The unemployment rate is a stock measurement and provides a snapshot of the labor market, rather than a measurement of the change in unemployment.

That change is meaningful (and in the report!) but tells a slightly different story.

The employment situation report attempts to characterize as much of the labor market’s vital signs as can be assessed on a monthly basis. But not all months are created equal. December necessarily involves Christmas spending, much as the early summer and early fall involve quite a bit of churn among teachers coinciding with the beginning and end of the school year. BLS uses seasonal adjustments to control for these foreseeable fluctuations, but it’s a moving target. The goal is to isolate aberrations and identify underlying labor trends. But these seasonal adjustments can under-correct or overcorrect in a given month. BLS revises monthly employment data twice after initial release, incorporating more observations and improving the integrity of the final published number.

But the revisions don’t stop there. Both major surveys in the Employment Report are updated at the start of the year, serving as a major reset for employment data. Last month’s report reflected a similar story, but it complicates how the data are discussed. Changes to seasonal factors, employment, and population levels similarly upend some prior understanding of the labor market. For example, the employment growth appeared to slow substantially in the last few months of 2021 – but other survey data suggested there was reason to doubt this deceleration. Sure enough, when the seasonal adjustment factors and other features of the surveys were revised, monthly employment growth in November-December was over 700,000 jobs higher than previously published. Similar challenges can attend the comparison of December household data to January household data. (see here for example)

In terms of the outlook for the month, notwithstanding this week’s theme of a return to normalcy, COVID-19 concerns were more prevalent in mid-February, when employment survey data is collected. There was a bit of a blip in initial claims that week which receded the following week. The ISM services employment index softened, but does not tend to correlate closely on a monthly basis with employment. The ADP index suggests strong private growth, but there again, ADP is not reliably correlated with the monthly employment gain either.

Fundamentally, the revised pace of job growth in the prior three months was strong, and stronger than expected. This guesstimator is assuming a 600,000 payroll increase in February and a decline in unemployment to 3.8 percent and an 8 cent, or 5.6 percent yearly average hourly earnings gain.