The Labor Department surprised economists by reporting this morning (Jan. 8) that the nation lost 85,000 jobs in December. Some economists had been predicting a rise in jobs, although November was revised upward to a gain of 4,000 -- the first increase in almost two years. But once again, I must report that the 85,000 number is actually worse. That's because the Labor Department has a computerized model that estimates the number of jobs created in new companies. Presumably, those jobs were not counted by the department's normal mechanism. It's called the birth-death adjustment model. In December, that model reported an increase in 59,000 jobs -- thus keeping the job loss from shooting well above 100,000.

The birth-death adjustment model lacks credibility. For example, the ADP report of private sector jobs, put out monthly by Automatic Data Processing, on Wednesday said that employment in small business (fewer than 25 workers) DECLINED by 25,000 jobs. It's not the same as the birth-death adjustment model, but it is close. For one thing, the government's model is an invention; it does not reflect real data. The ADP report does.

Overall, the ADP report found that the private sector lost 84,000 jobs in December. That was pooh-poohed by the economists who predicted the government report today would be positive. Now those economists are red-faced.