There are two key questions: (1) is the improvement in the job picture sustainable and (2) if it is, how long will it take get back where we started in December 2007? While there is considerable disagreement in terms of general economic policies, most agree that jobs are the key to the economic recovery.
Summary of the release
In November, the Establishment Survey showed that only 11,000 jobs were lost. There were some favorable revisions to October and September data.
The BLS also reports Household Data and there can be big differences between the Household and Establishment Surveys. For example, October, the Establishment Survey showed a decrease in nonfarm payrolls of (revised) 111,000. However, the Household data suggested a staggering increase of 558,000 to the ranks of the unemployed. In November, the news was much better on the Household Survey. The Household Survey showed that unemployment decreased by 325,000.
To get the unemployment rate, you take the Household Survey unemployment (15.375 million) and divide by the civilian labor force (153.877 million) and you get 10.0% (actually, 9.9992% — so we technically lost the ten-handle!)
There was other good news. The all-in unemployment rate (U-6) dropped from 17.5% to 17.2%. Temporary hiring (a good leading indictator continued to increase (adding 52,400 to the rolls). Finally, the average work week increased by 0.2. This means more money in the pockets of consumers.
All of this is good. However, there are some issues (which I am sure you expect from me).
Hiring and Firing
At some point, you cannot fire any more people and hope to have a viable business. The rate of firing has decreased. For a sustainable recovery, we need the rate of hiring to increase.
The participation rate has decreased from 66% in December 2007 (start of the recession) to 65% in November 2009. In numbers, there were 79.32 million not in the labor force in December 2007 and 82.866 million not in the labor force in November 2009. This is important. Population has been increasing. Lower participation is not a sign of economic health.
Duration of unemployment
At the beginning of the recession, the duration of unemployment was 16.5 weeks. Currently, it is 28.5 weeks. Note that the 28.5 weeks in November is an increase from October’s 26.5 weeks. Here is another way of looking at the data. In December 2007, 17.5% of the people looking for work were unemployed 27 weeks or greater. In November 2009, that proportion had more than doubled to 38.3% of the labor force being out of work for 27 or more weeks.
Seasonal factors are based on the average historical behavior in different months. The idea is to strip out variation due to the business cycle. For example, the fact that retail sales plunge every year from December to January has nothing to do with the health of the economy. In addition, these seasonal factors change through time and can be influenced by large observations — like we have seen over the past 18 months. Hence, we need to be especially cautious with the seasonally-adjusted data. That said, the not seasonally adjusted data also showed an improvement.
Why the inconsistency with ADP?
The ADP report on Wednesday suggested 169,000 job losses. For the past year, the ADP has been over-estimating job losses – but not by 158,000. The 158,000 miss is roughly double what their average overestimate is. Again, this is another reason for caution.
There are many that have given up looking for jobs. They will wait it out until conditions improve – like now. As a result, we will see a surge of people looking for jobs (i.e. re-entering the actively seeking work and unemployed). You have heard me make this argument before – but in the context of housing (the other key to a recovery). Many people want to sell their house but have taken it off the market because of frustration. They are waiting for markets to improve. Once the market improves and the wave of now-hidden inventory hits the market, a new downward pressure will be exerted on prices.
Mix of jobs and sustainability
There are 77 sectors that the BLS follows. Jobs were lost in 48/77 and there were gains in 29/77. The leading gainer was Administrative and Support Services. The category of Durable Goods jobs was the second biggest loser ranking 76/77. We need capital investment to increase to provide the foundation for a sustained recovery. This is key. The focus should be on not just getting a positive read on any particular month’s nonfarm payroll. We need the sort of investment that will lead to long-term, quality, job creation.
How long will it take to get back to where we started
Let’s do the math. First, the preliminaries. We have lost 7.2 million jobs. I went to the Bureau of the Census and pulled down file NP2008_D1.xls which shows the U.S. population by each year of age projected out many years. I added up the range of 16-65. The increment in this age group in 2010 is projected to be 1.844 million. Most of these people will be seeking work. I am being conservative here (if we count 16-70 years, the numbers increase to 2.177 million). Let’s say the participation rate is 65% (and it is way more empirically). This means that roughly 100,000 new people are entering the labor force every month.
Now we are ready to do the math. The question is the following: how many jobs need to be created, on average, every month over to put 7.2 million people back to work in three years. You take the 7.2 million and add 3.6 million (36 months times 100,000 average new entrants) and you get 10.8 million. 10.8million/36=300,000.
So, to be clear, November 2012, if we have averaged 300,000 jobs created every month for 36 months, we will have put the 7.2 million people that lost their jobs back to work.
This seems extremely unlikely. Over the past 20 years, i.e. 240 months, there have only been 55 months where non-farm payrolls have grown by more than 300,000. Over the past 10 years, only there have been only 4 months where we had more than 300,000 jobs created. Yet, we need 300,000 on average for the next 36 months to get back to where we started.
Indeed, nonfarm payroll growth has averaged more than 300,000 in only two years in the past 20 years: 1994 and 1984.
Well, if you don’t like 300,000 per month for three years, you need to add additional time to get those jobs back. For a five year horizon, roughly 200,000 jobs need to be created every month.
Note this analysis assumes no double dip and is very conservative on the number of people entering the labor force.
No matter how you look at it. We are in it for the long haul on the unemployment front. The persistent unemployment will feed back into slower growth over the next several years.