Tracking the movement in unemployment rates and jobs numbers month to month can resemble watching a dog chasing a squirrel — but without the fun.
They can be heartening or discouraging, confusing, surprising or misleading. And then the numbers get revised!
In September, Minnesota reported 9,900 jobs lost, but the following month that number was adjusted to a job loss of only 4,200. In October, the state initially reported 14,100 jobs but revised the number the following month upward of 16,000. November’s job loss of 5,100 did not get revised.
But state officials were clearly surprised by December’s “troubling” employment report showing 22,400 jobs lost in the month.
After substantially outpacing the national rate of job growth throughout the year, December’s losses reduced the state’s year-over-year job gains to 29,300, or a growth rate of 1.1 percent, more in line with the national rate of 0.8 percent. Since October, the unemployment rate held steady at 7.0 percent and total employment was essentially flat.
Steve Hine, director of research for labor market statistics and the Minnesota Department of Employment and Economic Development, warned that the numbers could change substantially. December will be “subject to revision” as state statisticians perform an annual benchmarking exercise, which “can change these numbers quite significantly.”
“I hope this 22,400 job loss is temporary statistically,” he added.
Nevertheless, the numbers put Minnesota forever in the record books as having the second-largest absolute month-to-month loss of jobs among the 50 states from November to December, according to the U.S. Bureau of Labor Statistics.
Nationwide, the U.S. unemployment rate fell 0.4 points in December, as 297,000 jobs were added. But that positive move was widely dismissed by many observers because it was offset by 260,000 who were counted as leaving the workforce. That accounted for about half the drop in the unemployment rate.
Monthly employment numbers are an estimate of payroll employment (those drawing a paycheck) based on a sample survey of companies, not an actual count. The survey does not count those who are self-employed, and as a sample, they are subject to statistical revisions month to month.
The unemployment rate, meanwhile, is a separate statistic based on a survey of households across the United States that asks if respondents are employed. If they’re not but have looked for work over the past four weeks, they get counted in the unemployment rate. If they have not looked for work for any reason, they are considered to have left the workforce. (See how the BLS measures unemployment here.)
So what to make of the “sawtooth” pattern of job growth and decline in the state and nation?
Economists caution not to make too much of one month’s numbers but rather to look at the trend over time.
One recent research report (PDF) by a Wall Street investment firm, Alliance Bernstein, took a look at December’s 0.4 point unemployment drop, comparing it with other large point drops in the past.
Joseph Carson, Bernstein’s chief economist, wrote, “We think there is ample evidence to support the view that the sharp fall in the jobless rate in December 2010 to a 30-month low was an inflection point rather than a statistical fluke.”
He added: “Large declines in the jobless rate are rare and the similarity of the recent data to other comparable periods suggests that the U.S. labor market is beginning to heal.”
Carson looked at the 22 instances over the past 40 years where the unemployment rate moved, up or down, by at least 0.4 points. While most of those occurrences were increases in the unemployment rate, on four occasions — not including December’s numbers — the unemployment rate fell 0.4 points or more.
In each case, that decline was driven in part by a reduction in the labor force, similar to the December decline and was followed in subsequent months by job growth. Two of the four occurrences marked an inflection point coming out of a recession, while the other two were in the middle of strong economic growth.
While acknowledging that 297,000 the December job gains were offset by a workforce decline of 260,000, he said: “[I]it’s impossible to determine why people opted to stop seeking jobs, as movements in and out of the labor force are volatile. Indeed, history says the most important early signal from the labor markets is the change in the unemployment rate, and not the granular data on job growth or labor force changes.”