In July, the unemployment rate remained unchanged at 9.5% even though the economy lost a total of 131,000 jobs. The unemployment rate a year ago was at 9.4%. However, the private sector added 71,000 jobs. The big reason for the overall loss of jobs was that 143,000 temporary census workers completed their work and were laid off.
Last month, a total of 221,000 jobs were lost, with the private sector gaining only 31,000 and 225,000 census workers being laid off. The private sector job gains for last month were revised sharply lower from an original reading of 83,000 jobs gained. The total jobs lost in June was revised up from 125,000.
The reason the economy could lose jobs and have no change in the unemployment rate, even in the face of a growing population, is that the percentage of people participating in the labor force -- the civilian participation rate -- has been on a steady downtrend. In July it fell to 64.6% from 64.7% in June, and down from 65.4% a year ago.
The participation rate will never get close to 100%. For that to happen, even infants and old people in nursing homes would have to be working. But it does indicate that even though the unemployment rate is only slightly higher than it was a year ago, the overall employment picture is significantly worse than last year.
Last year, though, things were deteriorating at a rapid pace, and now the direction is mixed on a lot of indicators. The percentage of people who are actually working, or the employment rate (also known as the employment population ratio) was 58.4% down from 58.5% in June and 59.3% a year ago. The unemployment rate is really about the interaction between the participation rate and the employment rate.
Participation Rate vs. Employment/Unemployment
The graph below shows the long-term history of the participation rate, the employment rate and the unemployment rate. Note that there was a huge secular uptrend in the participation rate starting in the early 1960’s and lasting until the end of the century. The participation rate peaked in April 2000 at 67.2%. It started to fall just before the 2001 recession never really recovered before starting to plunge again.
Even during the secular rise in the participation rate, it would flatten out or decline slightly during economic downturns. When it is hard to find a job, people get discouraged and leave the workforce, with older workers facing bleak employment prospects they opt for early retirement (also more people go on disability).
There were two great demographic forces that pushed the participation rate higher. The first was the Baby Boom, which started in 1946. By the mid-1960’s, those babies started to enter the workforce (if they were not in Vietnam). The tail end of the Baby Boom was in 1964, meaning that by the early 1980’s all of the Baby Boomers were in the work force.
The second huge demographic force was the entry of women into the labor force. If in the mid-1960’s you read an article about women and labor, the odds were that it was about childbirth. In July, women were 49.7% of all production and non-supervisory employees. They are fully integrated into the labor force and not likely to increase their participation rate significantly in the future (at least relative to men). The leading edge of the Baby Boom is now starting to retire.
Employment Rate
The employment rate is much more volatile than the participation rate (if it wasn’t, the unemployment rate would be very stable over time). It, too, was on a secular uptrend from the early 1960’s through the end of the century, reaching a higher high in each expansion, and the lows during each recession were not quite as low as the previous downturn.
That pattern was decisively broken in the last expansion. The employment rate peaked in April 2000 at 64.7%, and in the 2001 recession and its aftermath it fell to a low of 62.0 in September of 2003. The subsequent recovery was feeble and the highest the employment rate ever got to was 63.4% in March of 2007 before starting to slip again.
While we had a little bit of a recovery in the employment rate in the spring, and the current rate is still above the 58.2% level we hit in December, we are starting to fall again. The December rate was last seen in October 1977.
In other words, if the participation rate had not been falling, the unemployment rate today would be much higher. Then again, the unemployment rate during the entire Bush presidency would have been significantly higher as well. While it is the unemployment rate that gets all the headlines, the employment rate is just as important.
The unemployment rate among adult men fell to 9.7% from 9.9% in June and 9.8% a year ago. For adult women, it ticked up to 7.9% from 7.8% in June and from 7.6% a year ago. Men were particularly hard hit in this downturn, but perhaps the pain is starting to be spread a bit more equally among the genders.
It is also worth noting that on average women are still paid less than men (particularly if they are married), so perhaps employers have been holding onto their lower-cost employees and get rid of the high-cost ones. However, more of the difference comes from the fields that men dominate, such as construction, being harder hit in this downturn than fields where women have higher representation, such as health care.
Unemployment by Demographics
Teen unemployment has been particularly severe, rising to 26.1%, up from 25.7% in June and 24.5% a year ago. Some of that increase was due a rising participation rate among teens, but it is still at just 34.6% up from 34.1% in June and down from 37.9% a year ago. The employment rate for teens was just 25.6% down from 28.6% a year ago. That fact goes a long way towards explaining why the comp store sales at the teen retailers like Abercrombie & Fitch (ANF) have been so lousy.
Teen employment is important not just for the disposable income it generates so they can go on dates and buy clothes. For many, it helps pay for their education. Beyond that, it helps teach them the basic work skills (like showing up on time) that will help them out later in life.
The unemployment rate for whites was unchanged at 8.6% from June, and down from 8.7% a year ago. The participation rate was unchanged from last month at 65.1%, but it was down from 65.9% last year.
The unemployment rate for blacks increased to 15.6% in July from 15.9% in June and is up from 14.7% last year. The news for blacks is even worse than that appears. The participation rate fell to 61.5% in July from 61.9% in June and 62.5% a year ago.
For Hispanics, the unemployment rate fell to 12.1% in July from 12.4% in June and also 12.4% a year ago. The Hispanic participation rate was unchanged from last month at 67.4% but down from 68.5% last year.
The group with the worst job prospects by far is black teens. Their unemployment rate is an astounding 40.6%, up from 39.9% in June and 36.2% a year ago. That is in the face of a very low participation rate of just 24.2%, down from 27.8% a year ago. The employment rate for black teens is just 14.4%...that is the employment rate, not the unemployment rate!
Unemployment & Education
In July, the unemployment rate improved for those without much education and deteriorated slightly for the better educated. However, the absolute rate of unemployment is much worse for those who don’t have much in the way of formal education. "Stay in school" remains good advice, particularly when the economy gets soft.
The unemployment rate among those who did not even get a high school diploma dropped to 13.8% from 14.1% in June and 15.3% a year ago. High school graduates saw an even bigger improvement, with the unemployment rate falling to 10.1% from 10.8% last month, but up from 9.4% a year ago. For those who have some college experience, or an associates degree, the rate ticked up to 8.3% from 8.2% and is up from 8.0% a year ago. Those who have a bachelors degree (or higher) had an unemployment rate of 4.5% in July, up from 4.4% in June but down from 4.7 a year ago.
Perhaps the best news in the employment report is that we finally seem to be making some progress on the long-term unemployed. The duration of unemployment measures are still terrible, but at least this month they moved in the right direction. The median length of unemployment fell to 22.2 weeks from an astounding 25.5 weeks in June. That, however, is up from 15.9 weeks a year ago, and as the graph below shows, at that time, it was an all-time record high.
The average duration fell to 34.2 weeks from 35.2 weeks in June, but up from 31.2 weeks last year. Prior to the Great Recession, the highest the median duration ever reached was 12.3 weeks in May of 1983. The highest pre-Great Recession average duration was hit a month later in June 1983 at 20.8 weeks. Thus, while the improvement is very welcome, the absolute level is still just plain awful.
Long-term unemployment is a very different experience than short-term unemployment. Under normal circumstances, state-based unemployment insurance only lasts for 26 weeks. During recessions, the Federal Government has always stepped in with extended unemployment benefits. Despite the off-the-charts level of the long-term unemployed, it was a major struggle to get unemployment benefits past a Senate filibuster, but it was eventually overcome.
Since the unemployed are severely liquidity constrained (i.e., they don’t have any money and are going to have a hard time borrowing it) they will tend to spend the unemployment benefits very quickly, thus stimulating the economy. On a dollar-spent per-job-saved basis, it is hard to find a more effective stimulus program than extended unemployment benefits.
In July, there were a total of 14.599 million unemployed, down from 14.623 million in June but up from 14.534 million a year ago. Of those, 6.572 million or 44.9% of the total had been out of work for more than 26 weeks. That is a significant improvement from the 6.751 million (45.5% of the total) that had been out of work for more than 26 weeks in June, but up from “just” 4.972 million (34.2%) a year ago.
Short-term unemployed (fewer than five weeks) though increased to 2.839 million (19.4%) from 2.769 million in June, but down from 3.181 million (21.9%) a year ago. Those out of work for between five weeks and 14 weeks fell to 3.060 million from 3.121 million in June and from 3.539 million last year. Those out of work between 15 and 26 weeks fell to 2.151 million from 2.208 million in June and 2.847 million a year ago.
The graph below shows the number of unemployed by each length of unemployment group since 1948. Under normal circumstances, the bulk of the unemployed are out of work for just a short time, fewer than five weeks, followed by those out of work between five and 14 weeks. The number of unemployed who are out of work for fewer than five weeks (blue line) is not really affected all that much by the state of the economy.
Even in boom times there will be those who are laid off, but they can quickly find a new job. When the economy is strong, people are more likely to simply quit their existing job and try to find a new one. It is longer-term unemployment that is really the hallmark of a weak economy.
Long-Term Effects of the Great Recession
During a recession the number of people out of work for more than 15 weeks -- and especially those out of work for more than 26 weeks (orange line) -- really shoots up. The Great Recession has been exceptional in this regard. Since the official start (12/07) of the recession, the number of long-term unemployed has increased by almost five fold, while in most recessions it “merely” doubles.
Note that the number of long-term unemployed continues to rise well past the official end of the recession, and that the peak in long-term unemployment has been coming further and further past the end of the recession with each passing business cycle.
If the decline in the number of long-term unemployed we saw this month really does mean that we have hit a peak, that is very encouraging news indeed. The sharp decline in the number of people out of work between 15 and 26 weeks (green line) means that there are fewer people entering the long-term unemployment pipeline and makes a current peak seem plausible. However, one month does not make a trend, and it will take a few more months before we can declare victory on this front.
The decline in the participation rate is probably the greatest among the long-term unemployed. You are far more likely to get discouraged and give up looking for a job when you have been out of work for nine or ten months of rejection and no response to the resumes you send out than you are when you have only been looking for nine or ten weeks. Thus, the decline in the number of long-term unemployed does not mean that all those people got jobs.
As I noted at the beginning of this article, more than all of the jobs lost in July were government jobs, mostly temporary census jobs. However, the government payroll shrank more than the loss of census workers. The federal government shed a total of 154,000 workers, so with 143,000 census workers laid off it means that the rest of the federal government cut its civilian payroll by 11,000.
State governments dropped a total of 10,000 workers and local governments dropped 38,000 workers, including 27,000 teachers. The recent state aid bill that just passed the Senate will help limit the losses at the state and local level, but is not enough to counteract the massive budget deficit the state and local governments are facing. Since they are not allowed to run operating deficits by law, that means that they have to either raise taxes or cut spending, and cutting spending means cutting jobs.
Given the mood of the country, local elected officials are afraid that if they raise taxes that they are the ones who will be looking for a new job, not the teachers or cops or health inspectors. Look for more job losses at the state and local level in the months to come.
Private Sector Short of Expectations
The private sector job gains were a little bit short of expectations of about 80,000 job gains. The big disappointment was the sharp downward revision to just 31,000 private jobs gained in June rather than the 83,000 reported last month.
The goods producing sector added 33,000 jobs, up from a loss of 3,000 jobs in June (originally -8000). The services sector added 38,000 jobs versus 34,000 in June, although the June number was revised down from a gain of 91,000 jobs. Relative to the state of things a year ago, though, these are very good numbers. Last July, the economy lost a total of 346,000 jobs including 297,000 in the private sector, 128,000 in the goods producing sector and 169,000 in the service sector.
Within the goods producing sector, manufacturing added 36,000, up from a gain of 13,000 in June. That makes seven straight months of manufacturing job gains. The last time that statement could have been made was back in March of 1998. Since the start of the year, the economy has added a total of 183,000 manufacturing jobs.
As the graph below shows, that does not begin to reverse the secular decline that manufacturing employment (blue line, left scale) has been in since 1979. The peak in manufacturing employment was back in July 1979 at 19.531 million. The decline in factory jobs really accelerated in 2001; in January 2001 there were still 17.211 million factory jobs in the country. Even with the recent rebound, today there are just 11.717 million manufacturing jobs.
There have been two forces at work here. First is automation, which means employers can make the same or more stuff with fewer workers. The second is the shipping of jobs overseas where labor costs are much lower. Since we no longer make the stuff here we import it.
The trade deficit is entirely on the goods side; we actually regularly run a small trade surplus in services. While about half the goods trade deficit is due to oil, the rest is from all the imported stuff that fills the aisles of Wal-Mart (WMT) and Target (TGT). While over the years foreign auto makers such as Toyota (TM) have gained significant market share from domestic auto makers like Ford (F), many of those foreign cars have just as much domestic cars.
Construction Still Down
The construction industry (red line, left scale) has not been in the same secular decline as manufacturing. By its very nature it is impossible to outsource construction jobs. However with the popping of the housing bubble it is an industry that has really taken it on the chin in the Great Recession.
When the recession started, construction jobs had already been drifting down, but there were still a total of 7.535 million of them in November of 2007. With this month’s loss of 11,000 more jobs, the construction industry is now down to just 5.573 million, a decline of 1.962 million or 26.0%.
Since November of 2007, the economy as a whole has lost a total of 7.709 million jobs. In other words, more than one in four jobs lost in this downturn has come from the construction. At the start of the recession, the construction industry accounted for fewer than 5.5% of all jobs.
Services
The service sector added 38,000 jobs in July on top of 34,000 added in June. The June number was revised sharply lower from a gain of 91,000. The service sector is usually much more stable than the goods producing industries like manufacturing and construction, as shown by the green line on the graph. Since it is so much larger than the other sectors, it is shown on the right scale.
Usually the service sector merely flattens out during recessions rather than showing outright declines in employment. Within the service sector, health care added 27,800 of the jobs. It has consistently added jobs each month throughout the downturn.
Temporary Workers
One area of concern within the numbers for the service sector is a 5,600 decline in the number of temporary workers. These are the temps that employers get when they call Kelly Services (KELYA) or Manpower (MAN), not to be confused with the temporary census workers. They are an important leading indicator of the labor market.
When demand first picks up, an employer will not be sure if it is just a quick blip, or the start of a turnaround in his business. The initial response will be to work the current employees longer and harder, especially if they have had their hours cut back when the downturn hit. The next response though will be to bring on temporary help, rather than hire paramagnet employees (particularly if the firm has fringe benefits like health insurance).
The number of temporary workers had been in a nice uptrend since the start of the year, so the decline this month is a bit disconcerting, but is still relatively small and not a disaster -- just something that bears watching in the future.
Average Work-week
In addition, the other side of the early response, the average work-week ticked up to 34.2 hours from 34.1 hours in June and 33.8 hours a year ago. While that works out to just 6 minutes a week per worker for the month and 24 minutes a week for the year, when you multiply it across the 130.242 million people working according to the establishment survey it adds up to the equivalent of many more people working.
The establishment survey is the one that is used when people talk about the number of jobs gained or lost for the month. The unemployment rate is based on a separate household survey. The household survey shows a much higher total number of people working since it will pick up the self employed who will not be counted in the establishment survey.
In Summation
All in all, this report was a bit on the disappointing side, but not a disaster. It certainly does not show enough strength in the labor market to seriously bring down the unemployment rate. The declines we have seen in the unemployment rate since it peaked out in October at 10.1% have been mostly a function of people dropping out of the labor force -- a decline in the participation rate -- rather than from robust job creation.
The loss of jobs from the census layoffs should be mostly disregarded, just as the big jump in jobs in May due to bringing them on board was disregarded. The private sector has added jobs for seven months in a row this year, but at a very slow rate. Still, that is better than the implosion in the labor market and the overall economy we were seeing in late 2008 and early 2009.
There were encouraging signs in the duration of unemployment numbers, but they remain at levels that are absolutely disastrous and unprecedented. The leading indicators of the average work-week and the number of temporary workers are sending divergent signals. This report was very much of a mixed bag. Nothing to get excited about, but then again, no reason to get despondent.
Last month, a total of 221,000 jobs were lost, with the private sector gaining only 31,000 and 225,000 census workers being laid off. The private sector job gains for last month were revised sharply lower from an original reading of 83,000 jobs gained. The total jobs lost in June was revised up from 125,000.
The reason the economy could lose jobs and have no change in the unemployment rate, even in the face of a growing population, is that the percentage of people participating in the labor force -- the civilian participation rate -- has been on a steady downtrend. In July it fell to 64.6% from 64.7% in June, and down from 65.4% a year ago.
The participation rate will never get close to 100%. For that to happen, even infants and old people in nursing homes would have to be working. But it does indicate that even though the unemployment rate is only slightly higher than it was a year ago, the overall employment picture is significantly worse than last year.
Last year, though, things were deteriorating at a rapid pace, and now the direction is mixed on a lot of indicators. The percentage of people who are actually working, or the employment rate (also known as the employment population ratio) was 58.4% down from 58.5% in June and 59.3% a year ago. The unemployment rate is really about the interaction between the participation rate and the employment rate.
Participation Rate vs. Employment/Unemployment
The graph below shows the long-term history of the participation rate, the employment rate and the unemployment rate. Note that there was a huge secular uptrend in the participation rate starting in the early 1960’s and lasting until the end of the century. The participation rate peaked in April 2000 at 67.2%. It started to fall just before the 2001 recession never really recovered before starting to plunge again.
Even during the secular rise in the participation rate, it would flatten out or decline slightly during economic downturns. When it is hard to find a job, people get discouraged and leave the workforce, with older workers facing bleak employment prospects they opt for early retirement (also more people go on disability).
There were two great demographic forces that pushed the participation rate higher. The first was the Baby Boom, which started in 1946. By the mid-1960’s, those babies started to enter the workforce (if they were not in Vietnam). The tail end of the Baby Boom was in 1964, meaning that by the early 1980’s all of the Baby Boomers were in the work force.
The second huge demographic force was the entry of women into the labor force. If in the mid-1960’s you read an article about women and labor, the odds were that it was about childbirth. In July, women were 49.7% of all production and non-supervisory employees. They are fully integrated into the labor force and not likely to increase their participation rate significantly in the future (at least relative to men). The leading edge of the Baby Boom is now starting to retire.
Employment Rate
The employment rate is much more volatile than the participation rate (if it wasn’t, the unemployment rate would be very stable over time). It, too, was on a secular uptrend from the early 1960’s through the end of the century, reaching a higher high in each expansion, and the lows during each recession were not quite as low as the previous downturn.
That pattern was decisively broken in the last expansion. The employment rate peaked in April 2000 at 64.7%, and in the 2001 recession and its aftermath it fell to a low of 62.0 in September of 2003. The subsequent recovery was feeble and the highest the employment rate ever got to was 63.4% in March of 2007 before starting to slip again.
While we had a little bit of a recovery in the employment rate in the spring, and the current rate is still above the 58.2% level we hit in December, we are starting to fall again. The December rate was last seen in October 1977.
In other words, if the participation rate had not been falling, the unemployment rate today would be much higher. Then again, the unemployment rate during the entire Bush presidency would have been significantly higher as well. While it is the unemployment rate that gets all the headlines, the employment rate is just as important.
The unemployment rate among adult men fell to 9.7% from 9.9% in June and 9.8% a year ago. For adult women, it ticked up to 7.9% from 7.8% in June and from 7.6% a year ago. Men were particularly hard hit in this downturn, but perhaps the pain is starting to be spread a bit more equally among the genders.
It is also worth noting that on average women are still paid less than men (particularly if they are married), so perhaps employers have been holding onto their lower-cost employees and get rid of the high-cost ones. However, more of the difference comes from the fields that men dominate, such as construction, being harder hit in this downturn than fields where women have higher representation, such as health care.
Unemployment by Demographics
Teen unemployment has been particularly severe, rising to 26.1%, up from 25.7% in June and 24.5% a year ago. Some of that increase was due a rising participation rate among teens, but it is still at just 34.6% up from 34.1% in June and down from 37.9% a year ago. The employment rate for teens was just 25.6% down from 28.6% a year ago. That fact goes a long way towards explaining why the comp store sales at the teen retailers like Abercrombie & Fitch (ANF) have been so lousy.
Teen employment is important not just for the disposable income it generates so they can go on dates and buy clothes. For many, it helps pay for their education. Beyond that, it helps teach them the basic work skills (like showing up on time) that will help them out later in life.
The unemployment rate for whites was unchanged at 8.6% from June, and down from 8.7% a year ago. The participation rate was unchanged from last month at 65.1%, but it was down from 65.9% last year.
The unemployment rate for blacks increased to 15.6% in July from 15.9% in June and is up from 14.7% last year. The news for blacks is even worse than that appears. The participation rate fell to 61.5% in July from 61.9% in June and 62.5% a year ago.
For Hispanics, the unemployment rate fell to 12.1% in July from 12.4% in June and also 12.4% a year ago. The Hispanic participation rate was unchanged from last month at 67.4% but down from 68.5% last year.
The group with the worst job prospects by far is black teens. Their unemployment rate is an astounding 40.6%, up from 39.9% in June and 36.2% a year ago. That is in the face of a very low participation rate of just 24.2%, down from 27.8% a year ago. The employment rate for black teens is just 14.4%...that is the employment rate, not the unemployment rate!
Unemployment & Education
In July, the unemployment rate improved for those without much education and deteriorated slightly for the better educated. However, the absolute rate of unemployment is much worse for those who don’t have much in the way of formal education. "Stay in school" remains good advice, particularly when the economy gets soft.
The unemployment rate among those who did not even get a high school diploma dropped to 13.8% from 14.1% in June and 15.3% a year ago. High school graduates saw an even bigger improvement, with the unemployment rate falling to 10.1% from 10.8% last month, but up from 9.4% a year ago. For those who have some college experience, or an associates degree, the rate ticked up to 8.3% from 8.2% and is up from 8.0% a year ago. Those who have a bachelors degree (or higher) had an unemployment rate of 4.5% in July, up from 4.4% in June but down from 4.7 a year ago.
Perhaps the best news in the employment report is that we finally seem to be making some progress on the long-term unemployed. The duration of unemployment measures are still terrible, but at least this month they moved in the right direction. The median length of unemployment fell to 22.2 weeks from an astounding 25.5 weeks in June. That, however, is up from 15.9 weeks a year ago, and as the graph below shows, at that time, it was an all-time record high.
The average duration fell to 34.2 weeks from 35.2 weeks in June, but up from 31.2 weeks last year. Prior to the Great Recession, the highest the median duration ever reached was 12.3 weeks in May of 1983. The highest pre-Great Recession average duration was hit a month later in June 1983 at 20.8 weeks. Thus, while the improvement is very welcome, the absolute level is still just plain awful.
Long-term unemployment is a very different experience than short-term unemployment. Under normal circumstances, state-based unemployment insurance only lasts for 26 weeks. During recessions, the Federal Government has always stepped in with extended unemployment benefits. Despite the off-the-charts level of the long-term unemployed, it was a major struggle to get unemployment benefits past a Senate filibuster, but it was eventually overcome.
Since the unemployed are severely liquidity constrained (i.e., they don’t have any money and are going to have a hard time borrowing it) they will tend to spend the unemployment benefits very quickly, thus stimulating the economy. On a dollar-spent per-job-saved basis, it is hard to find a more effective stimulus program than extended unemployment benefits.
In July, there were a total of 14.599 million unemployed, down from 14.623 million in June but up from 14.534 million a year ago. Of those, 6.572 million or 44.9% of the total had been out of work for more than 26 weeks. That is a significant improvement from the 6.751 million (45.5% of the total) that had been out of work for more than 26 weeks in June, but up from “just” 4.972 million (34.2%) a year ago.
Short-term unemployed (fewer than five weeks) though increased to 2.839 million (19.4%) from 2.769 million in June, but down from 3.181 million (21.9%) a year ago. Those out of work for between five weeks and 14 weeks fell to 3.060 million from 3.121 million in June and from 3.539 million last year. Those out of work between 15 and 26 weeks fell to 2.151 million from 2.208 million in June and 2.847 million a year ago.
The graph below shows the number of unemployed by each length of unemployment group since 1948. Under normal circumstances, the bulk of the unemployed are out of work for just a short time, fewer than five weeks, followed by those out of work between five and 14 weeks. The number of unemployed who are out of work for fewer than five weeks (blue line) is not really affected all that much by the state of the economy.
Even in boom times there will be those who are laid off, but they can quickly find a new job. When the economy is strong, people are more likely to simply quit their existing job and try to find a new one. It is longer-term unemployment that is really the hallmark of a weak economy.
Long-Term Effects of the Great Recession
During a recession the number of people out of work for more than 15 weeks -- and especially those out of work for more than 26 weeks (orange line) -- really shoots up. The Great Recession has been exceptional in this regard. Since the official start (12/07) of the recession, the number of long-term unemployed has increased by almost five fold, while in most recessions it “merely” doubles.
Note that the number of long-term unemployed continues to rise well past the official end of the recession, and that the peak in long-term unemployment has been coming further and further past the end of the recession with each passing business cycle.
If the decline in the number of long-term unemployed we saw this month really does mean that we have hit a peak, that is very encouraging news indeed. The sharp decline in the number of people out of work between 15 and 26 weeks (green line) means that there are fewer people entering the long-term unemployment pipeline and makes a current peak seem plausible. However, one month does not make a trend, and it will take a few more months before we can declare victory on this front.
The decline in the participation rate is probably the greatest among the long-term unemployed. You are far more likely to get discouraged and give up looking for a job when you have been out of work for nine or ten months of rejection and no response to the resumes you send out than you are when you have only been looking for nine or ten weeks. Thus, the decline in the number of long-term unemployed does not mean that all those people got jobs.
As I noted at the beginning of this article, more than all of the jobs lost in July were government jobs, mostly temporary census jobs. However, the government payroll shrank more than the loss of census workers. The federal government shed a total of 154,000 workers, so with 143,000 census workers laid off it means that the rest of the federal government cut its civilian payroll by 11,000.
State governments dropped a total of 10,000 workers and local governments dropped 38,000 workers, including 27,000 teachers. The recent state aid bill that just passed the Senate will help limit the losses at the state and local level, but is not enough to counteract the massive budget deficit the state and local governments are facing. Since they are not allowed to run operating deficits by law, that means that they have to either raise taxes or cut spending, and cutting spending means cutting jobs.
Given the mood of the country, local elected officials are afraid that if they raise taxes that they are the ones who will be looking for a new job, not the teachers or cops or health inspectors. Look for more job losses at the state and local level in the months to come.
Private Sector Short of Expectations
The private sector job gains were a little bit short of expectations of about 80,000 job gains. The big disappointment was the sharp downward revision to just 31,000 private jobs gained in June rather than the 83,000 reported last month.
The goods producing sector added 33,000 jobs, up from a loss of 3,000 jobs in June (originally -8000). The services sector added 38,000 jobs versus 34,000 in June, although the June number was revised down from a gain of 91,000 jobs. Relative to the state of things a year ago, though, these are very good numbers. Last July, the economy lost a total of 346,000 jobs including 297,000 in the private sector, 128,000 in the goods producing sector and 169,000 in the service sector.
Within the goods producing sector, manufacturing added 36,000, up from a gain of 13,000 in June. That makes seven straight months of manufacturing job gains. The last time that statement could have been made was back in March of 1998. Since the start of the year, the economy has added a total of 183,000 manufacturing jobs.
As the graph below shows, that does not begin to reverse the secular decline that manufacturing employment (blue line, left scale) has been in since 1979. The peak in manufacturing employment was back in July 1979 at 19.531 million. The decline in factory jobs really accelerated in 2001; in January 2001 there were still 17.211 million factory jobs in the country. Even with the recent rebound, today there are just 11.717 million manufacturing jobs.
There have been two forces at work here. First is automation, which means employers can make the same or more stuff with fewer workers. The second is the shipping of jobs overseas where labor costs are much lower. Since we no longer make the stuff here we import it.
The trade deficit is entirely on the goods side; we actually regularly run a small trade surplus in services. While about half the goods trade deficit is due to oil, the rest is from all the imported stuff that fills the aisles of Wal-Mart (WMT) and Target (TGT). While over the years foreign auto makers such as Toyota (TM) have gained significant market share from domestic auto makers like Ford (F), many of those foreign cars have just as much domestic cars.
Construction Still Down
The construction industry (red line, left scale) has not been in the same secular decline as manufacturing. By its very nature it is impossible to outsource construction jobs. However with the popping of the housing bubble it is an industry that has really taken it on the chin in the Great Recession.
When the recession started, construction jobs had already been drifting down, but there were still a total of 7.535 million of them in November of 2007. With this month’s loss of 11,000 more jobs, the construction industry is now down to just 5.573 million, a decline of 1.962 million or 26.0%.
Since November of 2007, the economy as a whole has lost a total of 7.709 million jobs. In other words, more than one in four jobs lost in this downturn has come from the construction. At the start of the recession, the construction industry accounted for fewer than 5.5% of all jobs.
Services
The service sector added 38,000 jobs in July on top of 34,000 added in June. The June number was revised sharply lower from a gain of 91,000. The service sector is usually much more stable than the goods producing industries like manufacturing and construction, as shown by the green line on the graph. Since it is so much larger than the other sectors, it is shown on the right scale.
Usually the service sector merely flattens out during recessions rather than showing outright declines in employment. Within the service sector, health care added 27,800 of the jobs. It has consistently added jobs each month throughout the downturn.
Temporary Workers
One area of concern within the numbers for the service sector is a 5,600 decline in the number of temporary workers. These are the temps that employers get when they call Kelly Services (KELYA) or Manpower (MAN), not to be confused with the temporary census workers. They are an important leading indicator of the labor market.
When demand first picks up, an employer will not be sure if it is just a quick blip, or the start of a turnaround in his business. The initial response will be to work the current employees longer and harder, especially if they have had their hours cut back when the downturn hit. The next response though will be to bring on temporary help, rather than hire paramagnet employees (particularly if the firm has fringe benefits like health insurance).
The number of temporary workers had been in a nice uptrend since the start of the year, so the decline this month is a bit disconcerting, but is still relatively small and not a disaster -- just something that bears watching in the future.
Average Work-week
In addition, the other side of the early response, the average work-week ticked up to 34.2 hours from 34.1 hours in June and 33.8 hours a year ago. While that works out to just 6 minutes a week per worker for the month and 24 minutes a week for the year, when you multiply it across the 130.242 million people working according to the establishment survey it adds up to the equivalent of many more people working.
The establishment survey is the one that is used when people talk about the number of jobs gained or lost for the month. The unemployment rate is based on a separate household survey. The household survey shows a much higher total number of people working since it will pick up the self employed who will not be counted in the establishment survey.
In Summation
All in all, this report was a bit on the disappointing side, but not a disaster. It certainly does not show enough strength in the labor market to seriously bring down the unemployment rate. The declines we have seen in the unemployment rate since it peaked out in October at 10.1% have been mostly a function of people dropping out of the labor force -- a decline in the participation rate -- rather than from robust job creation.
The loss of jobs from the census layoffs should be mostly disregarded, just as the big jump in jobs in May due to bringing them on board was disregarded. The private sector has added jobs for seven months in a row this year, but at a very slow rate. Still, that is better than the implosion in the labor market and the overall economy we were seeing in late 2008 and early 2009.
There were encouraging signs in the duration of unemployment numbers, but they remain at levels that are absolutely disastrous and unprecedented. The leading indicators of the average work-week and the number of temporary workers are sending divergent signals. This report was very much of a mixed bag. Nothing to get excited about, but then again, no reason to get despondent.
Dirk van Dijk, CFA is the Chief Equity Strategist for Zacks.com. With more than 25 years investment experience he has become a popular commentator appearing in the Wall Street Journal and on CNBC. Dirk is also the Editor in charge of the market beating Zacks Strategic Investor service.