the Employment Situation report for September 2012
Last Friday at 8:30 am EDT, the Bureau of Labor Statistics of the Labor Department released its monthly Employment Situation report for September.
September job gains
The main employment figures were, at best, unexciting. According to the BLS, the economy added +114,000 new jobs last month. This figure breaks out into +104,000 positions in the private sector and +10,000 in state and municipal governments.
revisions for July and August
More significant, in the first of the usual two monthly revisions, the results of +96,000 jobs (+103,000 private, +7.000 government) for August were revised up by +46,000 positions to +142,000 (+97,000 private, +45,000 government).
The final July figures are also in. They’re +181,000 jobs (+163,000 private, +18,000 government). That’s also a positive revision. The initial report had put them at +163,000 (+172,000 private, -9,000 government). The first revision had pared that to +141,000 (+162,000 private, -21,0000 government).
the bottom line
The US economy needs to create around +150,000 new positions each month simply in order to absorb to the workforce people just finishing school and looking for their first jobs. Roughly speaking, that’s all the economy is doing. In other words, we appear to be making no headway in creating enough new positions to re-employ the several million people who lost their jobs during the recently ended Great Recession.
Yet,
the unemployment rate dropped below 8%, for the first time in years.
That’s the eye-popping figure in the BLS report.
Two aspects of this number are odd.
First of all, a surge in job creation usually results in a rise in the unemployment rate, not a fall. This is a quirk in the way the unemployment figures are calculated. People who become discouraged by their inability to find work and stop looking for work are not counted by the government as unemployed (no, I’m not making this up). When these people perceive that job hunt conditions have improved, they begin looking again–thus swelling the ranks of the unemployed.
Also, if we figure the US workforce is 150 million, then a 0.3% drop in the unemployment rate suggests the economy has added a net +450,000 jobs. That’s almost quadruple the job gain figure reported for September.
Coming as close as it does to next month’s presidential election, the unemployment rate has become a focal point for politicians on both sides of the aisle. Democrats are claiming the drop to be the first fruit of their economic policies. Republicans are saying the administration is cooking the books.
what’s really going on?
survey differences
The first point to note is that the monthly job gains and losses and the unemployment rate figure are compiled from different sources.
The BLS conducts a monthly survey of about 140,000 businesses, from which it estimates the job gains and losses. The sample is relatively large but the focus is narrow. The figures don’t include agricultural workers, the self-employed and unpaid family workers. Only changes larger than +/- 100,000 are statistically significant.
The Census Bureau conducts a monthly telephone survey of 60,000 households for the BLS, from which the unemployment rate is estimated. This sample is smaller and contains the upward biases that come with telephone interviewing. It does collect information about the groups mentioned above. Changes are only significant if they’re +/- 400,000 or more.
September’s household survey
Based on the telephone interviewing, the Labor Department estimated that just over 143 million people are working in the US at present–out of a workforce of 155 million.
The workforce was up by 418,000 vs. the August report. But there were enough new positive responses about work that 873,000 people found new jobs last month. This follows a two month period where the number of employed had dropped by about 300,000.
Do we know anything about what these new jobs are like? Not that much. It looks like about two-thirds of them are part-time. The sharp rise in this month’s household survey estimate appears to be based on positive responses by an extra 300 or so interviewees.
what to do
If there were any change in the interview questionnaire or any new instructions to interviewers that might server to elicit more positive responses in the survey, I doubt that could be kept secret for long. As a result, I think any attempt to manipulate the survey results for political gain would backfire.
My hope would be that the household survey is correct. My take is that this reported large employment number is a statistical anomaly that will disappear in the next survey. That’s scheduled to appear on the Friday before the election.
From an investment point of view, I think it’s much better to continue to position holdings in anticipation of a slow but steady upward tone to the economy rather than to adopt a more bullish posture. In fact, I’d be tempted to sell a bit into any strength–although none appears to be forthcoming so far.
Next month will be soon enough to reassess.