The October employment report came in weaker than expected as the country lost 190,000 jobs, rather than the 175,000 expectation. It was, however, an improvement over the 219,000 lost in September, but worse than the 154,000 jobs lost in August. Both the September and August job losses were revised sharply lower. As of last month it was thought that we lost 263,000 jobs in September and 201,000 in August. So in that context, missing expectations for October by 15,000 does not seem that bad. Of course, it is bad if you happen to be one of those losing your job. Based on the establishment survey we have now lost 7.3 million jobs since the recession started.
In general though, the pace of job losses has been slowing, especially if you step back and look at the big picture. Over the last three months, the economy has been dropping an average of 188,000 jobs a month, which is of course a disaster. However, in the prior three months, we were losing an average of 357,000 jobs a month and in the six months from November 2008 to April 2009 (inclusive), we were losing 645,000 jobs per month. In percentage terms, we have lost 4.41% of our jobs over the last year.
The graph below (from http://www.calculatedriskblog.com/) tracks both the unemployment rate, which is rapidly approaching a post war record (10.7% in 1983) and the year-over-year change in employment, which is actually starting to turn up (become slightly less negative). As the horrific job losses of last winter start to roll off, that measure should improve more in coming months. But remember: job losses have been going on since December 2007, which is closer to two years ago than one year ago. That can make the year-over-year change a bit misleading. Even with the improvement, the year-over-year job change is still far worse than previous recessions.
The job losses were wide spread by sector. In goods-producing industries we lost a total of 129,000 jobs, including 62,000 construction jobs and 61,000 manufacturing jobs. Since the recession started we have seen 1.6 million construction jobs go away and 2.1 million manufacturing jobs. Just to put in context how devastating the losses have been in those areas, there are only 5.966 million construction jobs and 11.675 million manufacturing jobs left in the country. The sharp loss in manufacturing jobs is in distinct contrast to the ISM survey earlier this week that indicated that manufacturing jobs were expanding. It does look like ADP got it pretty much right in their survey on Wednesday when they predicted a total of 203,000 jobs lost, including 65,000 in manufacturing.
The service sector lost a total of 61,000 jobs. The bulk came from retail, which lost 40,000 jobs. That’s not exactly what you want to see leading up to the holiday shopping season. Health care and education continued to buck the trend and added 45,000 jobs, including 29,000 in health care. In one encouraging sign, looking forward, temporary jobs increased by 44,000. At the bottom of a cycle those are often the first jobs to show an increase as employers realize they need more staff, but are not confident to add permanent employees. At least it is good news for firms like Kelly Services (KELYA).
The Unemployment rate, which is calculated from a different survey, painted a much gloomier picture, jumping to 10.2% from 9.8% in September and reaching its highest level since April of 1983. That survey showed the number of unemployed spiking by 558,000 in October, for a total of 8.3 million jobs lost since the start of the recession in December of 2007. Demographically all major groups saw their unemployment rate rise. For men, the rate is now 10.7%, an increase of 0.4% from September’s 10.3% and up from 6.4% a year ago. For women, the unemployment rate increased to 8.1% from 7.8% last month and 5.4% in October of 2008. Teen unemployment soared to 27.6% from 25.9% last month and 20.7% a year ago.
It is true that a job for most teens means money to put in the gas tank (or maybe to spend on clothing at Abercrombie and Fitch (ANF)), not a matter of paying the mortgage or keeping the lights on. However those first jobs teach important skills for the future, like the importance of showing up on time. Unemployment rates in the high 20s are something you associate with third-world countries, not the U.S.
By race, the unemployment rate among whites is now 9.5%, up from 9.0% in September and 6.0% a year ago. Hispanic unemployment is now 13.1%, up from 12.7% in September. Among blacks, 15.7% are now unemployed, up from 15.4% last month and 11.3% a year ago.
If one is looking for silver linings in the report, there are a few. The average manufacturing work week increased by 0.1 hour to 40.0, and average overtime increased by 0.2 hours. Employers will usually lengthen the number of hours their current employees are working before they start to add new staff, especially if those hours had already been drastically cut back. Also, for those with jobs, average hourly earnings increased by $0.05 or 0.3%, versus just a 0.1% increase last month. Over the last year average hourly earnings have increased by 2.4%. However, since people are working fewer hours, average weekly earnings are up just 0.9%. That does not provide a lot of firepower for people to take to the malls this Christmas season. The number of people working part time for economic reasons was unchanged at 9.3 million. That number had been increasing rapidly, so perhaps there is a little bit of stabilization on that front.
Still there are other measures in the internals of this report that are a troubling. One number that does not get nearly enough attention is the ratio of people working to the total population, what I like to call the employment rate. While it will never get close to 100%, unless you plan on eliminating child labor laws and have people in their 90s work, it is a very important measure. Ultimately it is the employed in a society that support everyone else, either directly or indirectly. The employment rate fell to 58.5% in October from 58.5% in September and 61.7% a year ago. The high for this cycle, was 63.4% in December 2006.
This cycle broke a long string of higher highs and lower lows stretching back over 50 years. That massive uptrend was largely the function of demographic changes that have largely played themselves out. Women are pretty much fully integrated into the work force and the baby boomer pig has pretty much moved through the python.
However, view this graph in conjunction with the first one. When a related measure, the civilian labor force participation rate was rising (it was unchanged in October at 65.1%), the rate of job creation (red line above) had to be much higher to generate the same unemployment rate (blue line above). The employment rate is now down to its lowest level since October 1983. Back then we still had large numbers of baby boomers that were trying to enter the job market for the first time, and far fewer women worked outside the home.
This has easily been the worst recession from the point of view of employment since the Great Depression, even though the unemployment rate was marginally higher under Reagan, hitting 10.7%. Consider the next graph (also from http://www.calculatedriskblog.com/). It shows the job losses as a percentage of the prior peak employment levels for every recession since the end of WWII. We have now exceeded the prior record set by the 1948 recession in terms of depth. We were going through a huge structural adjustment in 1948 as we demobilized the wartime economy and shifted back to civilian production. Nothing of the sort is occurring now.
Also in all but four of the prior recessions, not only had we stopped losing jobs by this point, but we had actually gained back all the jobs lost and had more people working than at the start of the recession. Next month we will pass that mark for the 1958 recession. The scary part is that the three recessions that took longer than 22 months to get back above the prior peak employment levels are the three most recent downturns, the ones that stated in 2001, 1990, and in 1981. Further, each of those lasted longer than the one before. Since we are still losing jobs at a fairly rapid rate, it could be October 2011 before we have more people working than we did in December 2007, and that is if we were to match the record of 2001. If the pattern of lengthening the time to get back to even continues, it could be far longer than that.
One of the most disturbing signatures of this recession is the length of time that people are out of work. There are now 5.594 million people who have been out of work for more than six months. The median duration of unemployment is now 18.7 weeks, a jump of 1.4 weeks from September’s 17.3 weeks. The previous record was set in June at 17.9 weeks, but prior to this downturn we have never seen anything remotely comparable to this level. The only other post-war recession that challenges this one in terms of severity is the 1981-83 downturn. The peak in the median length of unemployment in 1983 was just 12.3 weeks. A year ago we were at 10.6 weeks. The average (mean) duration of unemployment tells a very similar story. Since it is impossible to be unemployed for fewer than 0 weeks, the mean will always be higher than the median. On average if a person is out of work, he or she has now been out of work for 26.9 weeks, up from 26.2 weeks in September and 19.8 weeks a year ago.
As the chart below shows, it has historically been very rare for the average to exceed 20 weeks (and for the median to exceed 10 weeks). The only good sign I could find in the unemployment duration numbers was that the ratio of long term (over 26 weeks) to short term (less than 5 weeks) unemployed dropped slightly to 1.78 from 1.83 in September. However, let’s not get too excited about that. A year ago it stood at 0.74, and prior to this cycle had never gone above 0.80. The reason for the decline was a 6.1% increase in the number of short term unemployed, which is disturbing since that number had been trending down. The number of long-term unemployed increased by 2.9% on the month and is up 146% from a year ago. If those long-term unemployed were members of the middle class, it is unlikely that they are any more, and as the length of their unemployment continues to grow, their chances of rejoining the middle class diminish. This long term unemployment is going to cause a serious increase in poverty.
All in all, a disappointing jobs report. Yes we are making some progress, but it is not nearly fast enough. The unemployment rate is now far above the peak seen even in the “more adverse” scenario of the bank stress tests. People who are out of work for extended periods of time are going to have a hard time paying their mortgages. This means more foreclosures are going to be in the pipeline, and hurting the earnings of major mortgage lenders like Bank of America (BAC) as well as Fannie (FNM) and Freddie (FRE). By extension, it is also going to hurt us, the taxpayers, who now own 80% of both of them.
Read the full analyst report on “ANF”
Read the full analyst report on “KELYA”
Read the full analyst report on “BAC”
Read the full analyst report on “FNM”
Read the full analyst report on “FRE”
Zacks Investment Research
