Friday, Feb. 5
Written 11:30 a.m. EST
JOBS REPORT: The January employment report was a muddle. The economy shed more jobs than most economist expected. Yet the unemployment rate fell. Last year's employment numbers were revised, mostly downward.
Usually the mortgage market pays little attention to the headline unemployment rate, which fell from 10 percent to 9.7 percent. Instead, mortgage folks tend to pay more attention to the nonfarm payroll numbers, which fell 20,000. According to Briefing.com, the consensus forecast among economists had been that we would see an increase in nonfarm payrolls. Instead, we got a decrease.
As I'm writing this, late in the morning, mortgage rates have gone up a little. Most lenders either have raised rates by about an eighth of a percentage point, or you have to pay more in discount points to get the rate you want.
TWO REPORTS: As Calculated Risk points out, the nonfarm payrolls data and the unemployment rate come from separate surveys. The nonfarm payrolls number comes from the establishment survey, a query of about 400,000 businesses. The unemployment rate comes from the household survey of about 60,000 households.
The survey of businesses yielded that reported decline of 20,000 jobs. The survey of households showed an increase of 541,000 jobs. In this case, I'm tempted to find the household survey more credible, because my intuition tells me that jobs are being created by small businesses that aren't spotted by the establishment survey.
For what it's worth, I was part of the household survey in 1993 or 1994. Based on that experience, I'll tell you how the household survey worked. One day I got a knock on the door from a Census worker. She held a boxy electronic doodad that she entered data into. I let her in and she interviewed me for about 15 minutes, getting my demographic information as well as detailed information about my job -- I suppose so it could be classified correctly. I was a reporter for the newspaper in Toledo, Ohio.
She called me on the phone occasionally after that. If I remember accurately, she didn't call every month. Maybe it was every other month, or every quarter. She always asked about the previous week: Was I employed last week? Did I have the same job that I had when she first interviewed me? How many hours did I work last week, regular time and overtime?
She was precise about what constituted work hours. If I was on vacation the previous week, I worked zero hours. If I missed a day because of illness, I worked 30 hours instead of 37.5 (it was a union shop). She took a long time between questions because she had to enter the answers into her electronic box.
I think this lasted a year, and then I heard from her no more.