BLS Jobs Report, April 2011: The Economy is Either Up or Down

The BLS jobs report for April is out. I sometimes wonder why I even bother to analyze these reports anymore. It is like reading a fantastical story told with numbers that don't add up. This situation is encapsulated in the report's two big numbers. The report says the economy added 244,000 jobs in April. Yet the unemployment rate jumped from 8.8% to 9.0%. Now admittedly these two numbers come from different surveys. The jobs number comes from a poll of businesses whereas the unemployment figure is a poll of people.

Now the only way this can happen is if there is a big influx of people into the labor force, more than offsetting the increase in jobs, but the seasonally adjusted labor force participation rate (the ratio of the actual labor force –employed and unemployed– to the potential labor force, aka the civilian non-institutional population aka the NIP) was unchanged at 64.2%. Another possibility would be if the size of the NIP increased markedly in April, but it only increased by 146,000 in April, actually 3,000 less than in March. In addition, because of birth patterns, in the first 4 months of the year, so including April, the NIP increases more slowly than the following May-October period.

As in so many of these reports, we have a contradiction where one survey says "Look at all the pretty new jobs" and the other says "Oh, and unemployment went up again." So which is right? Well, I would search for corroborating data. Primarily, I would look at the average workweek. If the economy is adding on large numbers of employees, I would expect that the length of the work week would also increase, but it didn't. It was unchanged at 34.3 hours in April. Another place to look would be wages. But these increased only 0.1% in April or likely below what inflation will come in at for the month, meaning workers actually fell a little further behind in April. (The CPI data for April won't be out until next Friday).

Long story short, other BLS data do not support the 244,000 job number.

Diving into the numbers. As I said, the potential labor force increased by 146,000 in April. The employment-population ration was 58.4% (seasonally adjusted). Multiplying these two we get 85,264. Round to 85,000. This is the number of jobs needed to keep up with population growth in April. If we could believe the BLS jobs number and we can't, this would show that job creation in April was substantially higher than population growth.

The broader U-6 (seasonally adjusted) measure of un- and under employment increased in April from 15.7% to 15.9%. The marginally attached increased from 2.4 to 2.5 million. On the other hand, the unadjusted U-6 dropped from 16.2% to 15.5%.

The seasonally adjusted civilian labor force was 153.421 million. Multiplying by the U-6 (15.9%) gives us a number for the U-6 of 24.394 million. In March, it was 24.085 million. That's an increase of 309,000.

I define disemployment as the U-6 plus the BLS undercount. Following the BLS usage, its seasonally unadjusted measure for the undercount (Not in Labor Force, Want a Job Now) was 6.482 million, up 232,000 from the previous month. Since the marginally attached are counted both in this number and the U-6, to calculate disemployment, I subtract the 2.5 million of the marginally attached from the undercount. This gives 3.982 million. I add this to the U-6 to get 28.376 million. This is the measure of disemployment based on the BLS undercount. This yields a disemployment rate of 17.7%, unchanged from the previous month.

For my own calculation of the BLS undercount, I multiply the non-institutional population (potential labor force) 239.146 million by a participation rate of 0.67 indicative of a good recovery/expansion. This gives me an ideal labor force of 160.228 million. The labor force according to the BLS is 153.421 million. The difference between these two is 6.807 million and represents my number for the BLS undercount. This gives a number of 28.701 million for the disemployed and a disemployment rate of 17.9%, also unchanged from last month.

Seasonally adjusted the employed decreased by 190,000 and the unemployed increased by 205,000. Unadjusted, employment increased by 699,000 and unemployment decreased by 823,000. So depending which you look at, the economy either got better or worse.

Seasonally adjusted part-time workers increased from 8.433 million in March to 8.6 million in April. Interestingly, from September 2001 to November 2007, part-time workers were in the 4-5 million range. Their numbers increased after the December 2007 recession began and for the last 2 years have been between 8-9 million or twice their previous levels. At the beginning of a recovery, an increase in part-time workers is often expected but that the level of part-time workers has remained high over an extended period of time is rather a sign of economic weakness.

Finally, I just wanted to raise the issue of job quality. This is something the BLS really doesn't measure. We get only the broadest indications looking at wages and type of job, but it is really an important consideration to keep in mind. I think we would all probably agree that good jobs are few and far between, but it is difficult to quantify what this means. Still it is an added filter that we should always pass the BLS numbers through. Last month, for instance, 60,000 Mcjobs were added to the economy. They get the same weight as 60,000 manufacturing jobs at union scale or the same number of white collar jobs with benefits. Just something to keep in mind.