I’ve probably made this point before, but it’s worth making again. There’s a lot of snark going around about the job-creation figures released last week on Recovery.gov. The conservative National Review, for example, jokes that the data shows an “embarrass[ing]” $533,000-per-job performance by the economic stimulus bill.
That $533,000 figure comes from dividing the total amount of money spent so far on contracts, $16 billion, by the number of jobs they created, 30,000. $533,000 is more than 10 times the median national income — so if it takes that much money to create a job, the stimulus bill must be wildly inefficient, right?
Wrong. The $533,000-per-job figure is wildly misleading, for two reasons.
The Recovery.gov data is full of errors. This is another problem, of course; the stimulus bill’s vaunted transparency isn’t very useful if the government is releasing bad data.
But this was the first batch of recipient reporting data, and everyone — including the Recovery Accountability and Transparency Board, which oversees the stimulus — expected problems.
I’m looking around the DC area, for example, and I see dozens of contracts (some for $1 million or more) that don’t include any job creation numbers. It’s not that they say “zero jobs created” — which some contracts do. The contractors evidently didn’t provide any figures at all.
As data quality improves, which the RAT board promises it will, I suspect you’ll see the jobs numbers increase markedly.
Direct vs. indirect jobs. The data also only captures direct job creation; recipients don’t report indirect effects.
What’s the difference? If HUD gives you a million-dollar contract and you hire 20 construction workers, you’ve directly created 20 jobs. Those 20 contractors then turn around and spend some of their salary, stimulating the economy. Maybe they take their kids out to dinner; maybe they buy a new car. In any event, their spending creates indirect jobs.
The Recovery.gov data doesn’t capture those indirect jobs. The Obama administration’s economic team estimates, though, that each direct job creates one indirect job — and even if they’re overstating the case, the stimulus bill will create far more jobs than reported on Recovery.gov.
