Monday, January 12, 2009

Stimulus Projected to Save 3.675mm Jobs... 3mm Too Little?

Christina Romer and Jared Bernstein have released projected jobs created / saved due to the Obama stimulus plan... a cool 3.675 million. The charts below show where these jobs are projected to come, by sector and by method (i.e. direct or indirect).

The most jobs, not surprisingly, are in construction. Other top sectors include retail (from stimulating consumption), leisure (gotta do something when laid off right?), and manufacturing.

Table 2:

How? Well, the majority are expected to be indirectly created not from the plan itself, but by the recycling of dollars / demand created from those created directly. Projections are for state relief (i.e. fund projects that would otherwise be cut) to save / create the most, followed by the protection of those jobs vulnerable in the downturn, and tax cuts.

Table 5:

The question is obvious... is this enough? According to Christina Romer and Jared Bernstein's "R/B" own conclusion:

  • The recovery plan needs to be large to counter the tremendous job loss that is likely to occur
So is it? Paul Krugman says it clearly is not:
Here’s one way to look at it: R/B show the effects of the plan rapidly fading out during 2011. Yet at the end of 2011 the unemployment rate is still 6.3%. Meanwhile, the CBO estimates the natural rate, aka “full employment,” at just 4.8%. Why does the plan go away with the job undone?
In other words, R/B and Paul all conclude that the plan needs to be large enough to offset the jobs lost, but R/B themselves project that it won't. By their own analysis, employment is expected to rise to 9% with no stimulus. Comparing employment figures from December 2007 (unemployment at that time was 4.8%, which equals the CBO "full employment" level) with the projected employment figures for 2010 using R/B's 9% unemployment figure, I project the those unemployed to rise 6.7mm from December 2007. This is 3mm higher than the number of jobs the stimulus plan is projected to create / save (the difference between the loss in employment and gain in unemployment is the population growth) or to be blunt, not enough.


  1. Simple question: How does one objectively measure something that does not happen (job loss)?

    Does the fact that my next door neighbor didn't die in a house fire last night (which didn't happen) mean I saved his life?

    Maybe there is some quantum physics here that I'm not understanding.....

    And, by the way, where's my medal for saving my neighbor's life?

  2. ha... i like it.

    my problem has more to do with the fact that their projected jobs savings don't equal their expected job losses.

    this is like saying, you think your neighbor will die due to a fire that needs a "projected" 1000 gallons of water to put out, so lets splash 500 gallons on it.

    in other words, the solution is "expected" to fail