In its March report, the data from the Bureau of Labor Statistics remain squirrely. They’re up. They’re down. They’re sideways. The economy added 162,000 jobs. That’s good news. On the other hand, the expectations were for 200,000 or more so that’s not quite so good. The unemployment rate remained at 9.7% so that’s sideways. The U-6 number for un- and under-employment rose from 16.8% to 16.9% so that’s bad. By the way that’s 16.9% of 153.91 million (the size of the civilian labor force), which is 26 million, and that’s very bad.

If you look at the seasonally adjusted numbers for employment/unemployment, the potential labor force (those 16 and older) increased by 851,000 in the first 3 months of the year. The average participation rate (those in or trying to be in the workforce) over this period was 64.8% so this would mean 551,000 more people would need to be employed just to keep up with changes in the population. The number of employed increased by 1.113 million. So this would indicate that 562,000 more people, over and above the number just to keep up with where we are, were working. That’s 187,000 a month more people working in the first 3 months of the year. This is a really good number. The problem is that seasonally adjusted means it has been passed through a model to arrive at this result. So the result is only as good as the model.

When the data are all over the place, it becomes important to look at all of it, not just the usual stuff. I usually don’t look at the seasonally unadjusted numbers if the seasonally adjusted ones look comprehensible to me. This month I felt I needed to look at them.

The seasonally unadjusted numbers indicate that the potential labor force increased by 967,000 in the first 3 months of the year. The average participation rate was 64.667%, or 625,000 more people would need to be employed just to stay where we are (208,000 a month). But when we go to the numbers of those employed, there was an increase of only 30,000. Let me repeat that. The raw numbers show only an increase of 30,000 over 3 months, or an average of 10,000 a month. Compare this to the 1.113 million (187,000 a month) increase embedded in the seasonally adjusted numbers above. It is just very, very hard for me to see anything that would persuasively account for this variance between the two.

Now let’s look at the jobs numbers. These come from a different survey. So even at the best of times, the jobs numbers and the employment/unemployment figures will not track exactly, but they should still be in the same ballpark. The problem is they don’t always. This was especially true last year when the BLS was still using a model that consistently projected greater job creation or fewer jobs lost than was actually occurring. The BLS was supposed to have fixed this, but so far I am not seeing it. The model’s projections still undergo substantial revision in the 2 months after their initial release and it remains problematic marrying them to the employment/unemployment numbers.

According to the BLS and per its revisions, there were exactly zero jobs created in the first 2 months of 2010. So this month’s 162,000, if it holds up, represents all of the job creation for the first quarter of 2010. That’s 54,000 a month. Compare this to the 187,000-208,000 a month needed just to stay even in the employment/unemployment data above. Now let us look more deeply into this 162,000 number. The government increased its hiring by 48,000, essentially all of these are temp positions related to the Census. On the private side, 40,000 temp positions were also created. So 54% of the 162,000 figure is made up of temp positions. An increase in temp jobs is sometimes an indication of recovery, but the Census jobs are anomalous and the other temp jobs may be a redistribution of work from permanent to cheaper temp positions.

What I take away from this is that we have three sets of numbers that are widely at variance with each other. The seasonally unadjusted employment/unemployment numbers and the jobs numbers averaged for the first quarter are in closest, but not very close, agreement. They show a deteriorating picture where employment and jobs are not growing fast enough even to keep up with natural population growth.

This is, of course, not how it will be spun, best report in 3 years, making progress, but work still needs to be done, yadda, yadda, yadda. The numbers, as undependable as they are, tell a different story.