Current jobs numbers look very good, but past numbers aren't holding up. So be skeptical

The January US jobs report always has two important sets of data. The first involves the typical monthly updates and revisions to the prior two months, and those figures were quite strong for January.

Nonfarm payrolls increased by 225,000 jobs last month, with employment at construction sites increasing by the most in a year amid milder-than-normal temperatures, the government's survey of establishments showed. There were also strong gains in hiring in the transportation and warehousing industry.

Economists polled by Reuters had forecast payrolls would rise by 160,000 jobs in January. Data for November and December was revised to show 7,000 more jobs created than previously reported. Job growth in January exceeded the monthly average of 175,000 in 2019.

It wasn’t great in every area, because manufacturing shed 12,000 jobs and 8,300 jobs lost in retail. But 225,000 jobs is still pretty good, particularly given that the economy is heading toward its 11th straight year of expansion.

And even though the unemployment rate rose from 3.5% to 3.6%, it was because of a sizable increase in the number of people entering the work force, which meant that a key employment stat reached a multi-year high.

The labor force participation rate, or the proportion of working-age Americans who have a job or are looking for one, rose two-tenths of a percentage point to 63.4% last month, the highest since June 2013.

Wage growth also went up a tick after a soft month to round out 2019.

The tightening labor market is steadily driving up wages. Average hourly earnings increased seven cents, or 0.2%, last month after gaining 0.1% in December. That lifted the annual increase in wages to 3.1% in January from 3.0% in December.

But what’s not mentioned is that these January jobs numbers are from a smaller figure, because of the annual benchmarking, which is explained thusly by the Bureau of Labor Statistics.

In accordance with annual practice, the establishment survey data released today have been benchmarked to reflect comprehensive counts of payroll jobs for March 2019. These counts are derived principally from the Quarterly Census of Employment and Wages (QCEW), which counts jobs covered by the Unemployment Insurance (UI) tax system. The benchmark process results in revisions to not seasonally adjusted data from April 2018 forward. BLS revised seasonally adjusted data from January 2015 forward. In addition, both seasonally adjusted and unadjusted data for some series incorporate other revisions prior to 2015.

The total nonfarm employment level for March 2019 was revised downward by 514,000 (-505,000 on a not seasonally adjusted basis), or -0.3 percent. The absolute average benchmark revision over the past 10 years is 0.2 percent.

Those downward revisions include the first monthly loss in private sector jobs in 9 years in February 2019. With the year-over-year totals for the time before June 2019 being revised down, job growth sunk to the lowest levels in 8 years, both in overall growth and private sector growth.

But interestingly, the downward revision was concentrated in 2018 and the first half of 2019, while the last half of 2019 was revised up by a bit. This meant that job growth for last year was only reduced by 12,000 from the the 2.1 million that was previously reported.

However, I’ll give a note of caution here, as the stronger growth of the last half of 2019 and Jan 2020 are the months that have not been tracked by the QCEW. We’ll get the Q3 2019 QCEW in a couple of weeks, and I want to compare what that report says with the 2.02 million that is in the monthly jobs report. If it keeps coming up short, we should be very skeptical that job growth has jumped as much as has been reported in recent months.

What it means is that the current figures indicate that the declines in job growth that have gradually happened over the last 4 years have been limited, and there's been a slight bump up according to the monthly jobs reports. But because the last two years of revisions show that the jobs growth in the US has been consistently overstated, let's see what the refined QCEW data says before we think that things have truly turned around.