Tuesday, February 10, 2009

Recession And Job Loss In Charts

These bright and colorful charts are full of dire predictions and disheartening info, and they've been making the rounds online of late.

First, this chart gives the job losses from the most recent peak for the past three recessions. Very scary, though Ezra Klein is right to point out that it's a bit unfair since the economy was larger to begin with this time around and therefore had more jobs to begin with and more jobs to lose.

Klein then pointed to this chart:

Which shows job loss as a percentage for the past six recessions. Klein is scared that we're already pretty much as far down as the worst of these and have no idea how much further things will fall this time around.

Nate Silver made an equally (and potentially more) important observation when he pointed out that over the past six recessions on this chart each one has taken progressively longer to recover the jobs lost. Which would mean that we should expect a long lasting recession. Also very scary, probably scarier since this is better data.

And to top it all off Dean Baker explains with this chart that job loss reports are often too low initially (Floyd Norris reports on just such a revision from last week) and that we should assume that the latest report of 580,000 jobs lost in January is low as well...

That may be hard to believe, but the economy almost certainly lost more jobs in January than the 597,000 job loss reported by the Bureau of Labor Statistics (BLS). The reason is that BLS imputes jobs for new firms that are not included in its sample.

The formula used for calculating this imputation is backward looking, meaning that it depends on growth in prior quarters. When the economy takes a sharp turn in either direction, as it did last fall, the imputation is likely to be too high or too low, depending on the direction of change.

The chart below compares the imputed job gain/loss in new firms in last four months with the imputation for the corresponding month one year earlier. (These data are not seasonally adjusted.) It is simply not plausible that more jobs have been generated in new firms in the last four months than in the same months of last year. The true rate of job growth is likely 40,000 to 60,000 less per month than current data indicate.

No comments:

Post a Comment