Calculated Risk: Employment Report: 20K Jobs Lost, 9.7% Unemployment Rate

US employment/unemployment stats for January 2009 came in.  The headlines:  9.7% unemployment rate (down from 10.0) and 20,000 jobs lost.

At first glance this looks strange and counter-intuitive:  fewer actual people working, yet a higher unemployment rate.  Of course, we’ve seen this in the past with the change in the rate and the change in actual jobs giving conflicting signals.  In the past, the usual explanation is a change in the size of the labor force (the denominator in the unemployment rate) where large numbers of people got discouraged and dropped out of the labor force.  This time, though, the explanation for conflicting signals is more procedural.  The two numbers (unemp rate and # of jobs) come from two different statistical sampling surveys each month.  The unemployment rate is calculated from the responses of a sample of households – actual workers themselves.  The number of jobs estimate is a sum from two sources:  a survey of employers and an econometric model that estimates how many new businesses were created/failed each month.  When the reality in the economy is that there’s close to no change, then the two surveys can easily give conflicting signals.

Conclusion:  Like all macro stats, there’s a lot of “fuzziness” in the numbers. Judgement required. These are estimates based on statistical samples.  Judgement is required.  Using macro stats to monitor the health of the economy is  akin to using a child’s compass to find your way through the woods – you’ll eventually know the direction you’re headed, but there’s a lot of imprecision involved.  People (and politicians) get in trouble when they treat the numbers as if they’re like a speedometer on your car or like an accounting financial statement.

More numbers, links, and graphs below the fold:


From the BLS:

The unemployment rate fell from 10.0 to 9.7 percent in January, and nonfarm payroll employment was essentially unchanged (-20,000), the U.S. Bureau of Labor Statistics reported today. Employment fell in construction and in transportation and warehousing, while temporary help services and retail trade added jobs.Employment Measures and Recessions Click on graph for larger image.

This graph shows the unemployment rate and the year over year change in employment vs. recessions.

Nonfarm payrolls decreased by 20,000 in January. The economy has lost almost 4.0 million jobs over the last year, and 8.42 million jobs since the beginning of the current employment recession. (note: job losses were 7.2 million before benchmark revision).

The unemployment rate declined to 9.7 percent. (I’ll have more on that soon)

Percent Job Losses During Recessions The second graph shows the job losses from the start of the employment recession, in percentage terms (as opposed to the number of jobs lost).

For the current recession, employment peaked in December 2007, and this recession is by far the worst recession since WWII in percentage terms, and 2nd worst in terms of the unemployment rate (only early ’80s recession with a peak of 10.8 percent was worse).

via Calculated Risk: Employment Report: 20K Jobs Lost, 9.7% Unemployment Rate.

See also:

From Krugman:

Excellent, easy to read explanation with graphs of the two surveys & resulting issues in numbers:  Jobs and the Unemployment Rate via Calculated Risk.

One thought on “Calculated Risk: Employment Report: 20K Jobs Lost, 9.7% Unemployment Rate

  1. A small drop or rise in the unemployment rate means nothing, even in a crisis you get pent-up demand in some sectors resulting small spikes. The problem is that the GDP is highly overvalued as it contains the crisis-driven overspending which is not a sign of health rather the opposite! And with the variable mortgages being reset later this year that will all fall like a rock. As for trade statistics and other statistical data etc.: I have posted a Statistical Reference Inventory ( to my economics blog with economic and statistical data series, history, bibliographies etc. for students & researchers, probably the most comprehensive on the Internet. Currently over 300 meta sources, it will soon grow to over a thousand. Check it out and if you miss something, feel free to leave a comment.

Comments are closed.