The September 2010 employment report released this morning by the Bureau of Labor Statistics again showed positive but totally inadequate private sector employment growth, with the addition of 64,000 private-sector jobs in September. State and local governments, their budgets crunched, lost 83,000 jobs. The pain of the state and local budget problems are clear in these numbers: of the 83,000 state and local jobs lost, 58,000 were in education, as teachers and other education workers were not called back for the new school year. September was the first month this year where, barring changes in temporary Census workers, the labor market lost jobs. The unemployment rate held steady at 9.6%.
While the private sector gained 64,000 jobs, the total number of payroll jobs decreased 95,000 in September due to government employment losses. The federal government shed 77,000 temporary 2010 Decennial Census workers in September, while state and local governments lost 83,000 jobs. With only 6,000 temporary Census workers remaining, the effect of the Census on the payroll employment numbers has now run its course. But with state and local government budget problems continuing, losses in public employment will continue.
Today’s report included a preliminary estimate of the annual benchmark revision to payrolls. The estimates show that when the revisions ultimately occur (with the release of the January 2011 data), they will likely show a downward revision of around 366,000–in other words, that the establishment survey overestimated job growth between April 2009 and March 2010 by 366,000. This compares to a downward revision of over 900,000 last year.
Wages and Hours
Average hourly wages increased in September by a penny, from $22.66 to $22.67. Nominal hourly wage growth had been generally slowing over the course of the recession. For example, from 2010Q1 to 2010Q2, nominal wages grew at a 1.1% annualized rate, compared to a 3.2% rate during the year before the recession started. However, wage growth picked up somewhat in the third quarter of this year, growing at a 1.9% annualized rate. With unemployment expected to remain high for a long time, recent wage growth improvements are not likely to continue over the longer-term.
The length of the average workweek held steady in September, at 34.2 hours. There has been no improvement in average hours since May, though average hours (for all private employees) are up 1.5% from their low of 33.7 last fall. However, at the start of the recession in December 2007, the length of the average workweek in the private sector was 34.7 hours, so there remains a great deal to make up. Simply restoring average hours worked by all 108 million private-sector workers from 34.2 back to 34.7 would be equivalent to adding 1.6 million new jobs. The gradual restoration of average hours will therefore be an ongoing drag on new hiring.
Weekly wages, which combines the impact of changes in hourly wages and average hours, increased slightly in September, from $774.97 to $775.31. After falling faster than hourly wage growth for the year and a half of the recession as hours were cut back, weekly earnings growth saw improvements from the summer of 2009 to the spring of 2010, driven in part by the increase in average hours between the fall of 2009 and the spring of 2010. However, as average hours have stalled out since May, so have weekly wages. Weekly wages have grown at an annualized rate of 1.6% since May, after growing at an annualized rate of 3.2% in the six months before that.
Labor Force
The labor force increased by 48,000 in September, leaving the labor force participation rate unchanged at 64.7%. The labor force participation rate is still far below its prerecession level of 66.0% in December 2007, so the pool of “missing workers,” that is, workers who dropped out of (or didn’t enter) the labor force during the downturn, remains large. We can estimate its size in the following way. The labor force should have increased by around 3.8 million workers from December 2007 to September 2010 given working-age population growth over this period, but instead it grew by 289,000. This means that the pool of missing workers now numbers around 3.5 million. None of these workers are currently reflected in the official unemployment count, but as they enter or re-enter the labor force in search of work, this will contribute to keeping the unemployment rate high.
Long-term unemployment
The share of unemployed workers who have been unemployed for over six months dropped from 42% to 41.7% in September. This improvement likely reflects workers dropping out of the labor force after exhausting unemployment insurance benefits. Despite this, the long-term unemployed share remains the seventh-highest on record, and there are still 6.1 million workers who have been unemployed for longer than six months. These dramatic figures are unsurprising given that there are still 4.6 unemployed workers per available job. The median, or typical, unemployment spell rose from 19.9 to 20.4 weeks, and the average unemployment spell dropped from 33.6 to 33.3 weeks.
Underemployment
The “underemployment rate,” or the U-6 measure of labor underutilization, is a more comprehensive measure of labor market slack than the unemployment rate because it includes not just the officially unemployed, but also jobless workers who have given up looking for work and people who want full-time jobs but have had to settle for part-time work (note, however, it does not include people underemployed in the sense that they have had to take a job that is below their skills, training, or experience level). This measure increased by 0.4 percentage points to 17.1% in September, meaning that more than one in six U.S. workers was either unemployed or underemployed. The number of involuntary part-time workers increased by 612,000, while the number of “marginally attached” workers (jobless workers who have given up looking for work), increased by 139,000. In September, there were a total of 26.8 million workers who were either unemployed or underemployed.
Demographic breakdowns
All major demographic groups have experienced substantial increases in unemployment over this downturn, though men, racial and ethnic minorities, young workers, and workers with lower levels of schooling have gotten hit particularly hard.
- In September, unemployment was 17.9% among workers age 16-24, 8.7% among 25-54 year olds, and 7.2% among workers 55 and older (increases of 6.1, 4.6, and 4.0 percentage points, respectively, since the start of the recession in December 2007).
- Unemployment was 16.1% among black workers, 12.4% among Hispanic workers, and 8.7% among white workers (increases of 7.1, 6.1, and 4.3 percentage points, respectively, since the start of the recession).
- Unemployment was 10.5% for men, compared to 8.6% for women (increases of 5.4 and 3.7 percentage points since the start of the recession).
- For workers age 25 or older, unemployment reached 10.0% for high-school-educated workers and 4.4% for those with a college degree (increases of 5.3 and 2.3 percentage points, respectively, since the start of the recession).
Industry sectors
All of the gains in private-sector jobs were in service-providing industries—service-producing industries added 86,000 jobs while goods-producing industries lost 22,000 jobs. Restaurants and
bars added 33,900 jobs, one real bright spot. Health care added 23,900 jobs, on par with what it added in the prior three months. Temporary help services added 16,900, close to what it added in August. Retail trade added 5,700 in September, an improvement over its performance in the last four months, in which it averaged a loss of 3,200.
Construction lost 21,000 jobs, after adding an average of 7,000 jobs a month over the prior three months. Manufacturing lost 6,000 jobs in September, after adding an average of 18,000 a month for the first eight months of the year. September’s loss was all in non-durable goods.
In the public sector, aside from changes in temporary Census jobs, the shedding of jobs at the state and local level remains an ongoing drag on employment growth. In September, state and local governments shed 83,000 jobs (-7,000 state, -76,000 local). The pain of the state and local budget problems are clear in these numbers: of the 83,000 state and local jobs lost, 58,000 were in education, as teachers and other education workers were not called back for the new school year. Since their peak in September 2008, state and local governments have shed 410,000 jobs (-57,000 state, -353,000 local).
Conclusion
As the chart shows, the labor market remains an estimated 8.1 million payroll jobs below where it was at the start of the recession in December 2007. This number includes both the 7.8 million jobs lost in the payroll data as currently published plus the announced preliminary benchmark revision of -366,000 jobs to last March’s employment level. And even this number understates the size of the gap in the labor market by failing to take into account the fact that simply to keep up with the growth in the working-age population, the labor market should have added around 3.4 million jobs since December 2007. This means the labor market is now roughly 11.5 million jobs below the level needed to restore the pre-recession unemployment rate (5.0% in December 2007). To get down to the pre-recession unemployment rate within five years, the labor market would have to add around 300,000 jobs every month for that entire period. In September, excluding changes in temporary Census hiring, the labor market lost 18,000.
The Congressional Budget Office estimates that without the American Recovery and Reinvestment Act of 2009, the unemployment rate would be up to 2 percentage points higher than it is right now, and we would have up to 5.2 million fewer full-time equivalent jobs. In other words, ARRA worked, but was never big enough given the scale of the crisis. With a deficit of 11.5 million jobs, a 9.6% unemployment rate, a private sector failing to provide robust job growth, and a recovery act now fading out, Congress and the Federal Reserve Board need to take action to stimulate the economy and create jobs.
—Research assistance from Kathryn Edwards and Andrew Green