BLS Jobs Report, September 2010

The overall jobs picture for September 2010 painted by the BLS numbers is of an economy dead in the water with some further deterioration around the edges. As I often point out, the BLS report is based on two surveys, household (employees) and establishment (employers). The first measures people, the second, positions. The numbers are scaled up from these surveys and passed through models (which have not been working that well for the type of recession we are in) to give seasonally adjusted figures. It is these which are usually cited both in the reports and in the news. These are estimates and undergo revisions in the following two months and again at the end of the year.

Looking at the establishment data first, essentially the last of the temporary Census positions (77,000) ended in September. If we are to take the report at face value all job losses at the federal level were Census related. Local government cut 76,000 jobs of which 49,800 (65.5%) were in education. By contrast, the September data indicate that states cut very few jobs (7,000). With large state budget shortfalls I find that number surprising. Taken together and taking rounding into account, these sum to the 159,000 government job losses in September.

The private sector added 64,000 jobs. The primary gainers were 24,000 in the economically unsustainable healthcare sector; 16,900 in temp positions; and 34,000 in food services. An increase in temp workers can be an early indication of recovery, but here it is a sign of weakness, an indication of some redistribution of positions in the economy from full-time to part-time. The increase in food services is also not a terribly positive sign. You are not going to fuel a recovery by increasing the number of positions for burger flippers and waiters. It isn't just the number of jobs that is important but their quality. The quality of jobs in September left much to be desired.

159,000 jobs lost in the public sector minus 64,000 jobs in the private sector gives the overall 95,000 job lost figure for September cited in the report.

As a kind of crossover to the household data, we can estimate the number of jobs needed in September to stand still, taking into account population growth. The civilian noninstitutional population (CNP) over 16, the pool of all potential adult workers increased from 238.099 million to 238.322 million in September or 223,000. The employment-population ratio, those currently employed divided by the CNP, was 58.5% in September. This means the economy needed to employ or find jobs for 130,000 people in September just to stay where it was, but it lost 95,000. In other words, the jobs shortfall for September was 225,000. This is an important number which the BLS doesn't report and shows that the jobs situation for September was considerably worse than the surface numbers indicate.

The U-3 unemployment rate remained at 9.6% and represented 14.8 million unemployed. The U-6 measure of un- and underemployment increased from 16.7% to 17.1% representing an increase from 25.74 to 26.36 million Americans or 620,000 for September. This is an even bigger negative number which the BLS does not mention. The U-6 is a measure of what lambert calls disemployment and it got much worse in September.

Finally estimating the number of unemployed workers which the BLS undercounts, I have chosen a labor force participation rate of 67%, indicative of a strong expansion, and multiply the CNP by it. For September, this would come to 159.676 million. The actual BLS figure was 154.158 million. That is, by my estimate, the BLS undercounts the number of unemployed in the economy by 5.518 million, up 102,000 from the 5.416 million of the previous month. Add this to the 26.36 million of the U-6 and you get 31.88 million or an overall disemployment rate of 20% (19.97%).

The bottomline here is that the BLS report describes a jobs and employment situation that is essentially unchanged from August, bad but only slightly worse. However, if you look into the numbers, jobs and employment are significantly worse.

(Note: If you find numbers here not in the BLS report, they come from me directly reading the BLS tables or calculating off them.)