I confess to being stumped by the contradictory information circulating today.
From a survey of households, you get evidence that unemployment is dropping: From August to September, on a seasonally adjusted basis, the number of unemployed dropped by 12,000 while the number of employed grew by 13,900, resulting in an 8.4% unemployment rate.
From a survey of employers, you get evidence that jobs are being eliminated: From August to September, on a seasonally adjusted basis, the number of jobs declined by 20,900.
What gives here? Could it be that people are giving up and dropping out of the workforce? I don’t think so — the first link, the survey of households, actually shows the labor force expanding by 1,900 during the period.
So, I’m left with three potential, non-exclusive explanations:
1) One or both of the surveys is off.
2) The seasonally adjustment mechanism is introducing more uncertainty into the process than an non-expert like myself can understand. Looking at the unadjusted results for the survey of households, you can see that the summer season adds about 50,000 people into the labor force. Given that the movement from August to September is a fraction of that, an inaccurate adjustment can distort the actual numbers.
3) There’s some semantic tension between what each survey is measuring — one is employed people and the other appears to be jobs. There’s a difference and that difference might account for some of the contradictory evidence.
Anyone else have any ideas? Both campaigns are busy duking it out over what these results mean. It’d be nice to have a clear explanation.