The Liberty Papers

First, let me state categorically that I don’t believe government statistics. When they talk about $X Billions of dollars spent and Y00,000 numbers of jobs created, it’s clearly BS. The government rarely knows (or cares) how much of allocated money is actually spent until they’ve run out, and I think we all know that the job count is inflated in any statistically possible way to make themselves look better.

But I just can’t get over the analysis (and sorry Jason, such as this post) that simply divide the number of billions of dollars of stimulus money spent by the number of jobs to come up with a “per job” cost. In this case it was $746K “per job”.

There’s an implicit charge there, suggesting that if you spend $746K “per job”, it’s a complete and total waste of money. The charge, of course, is that the labor cost is the “per job” amount — or should be if it weren’t “squandered”.

And let’s also be clear. I’m not suggesting this money was spent well or efficiently. But that’s not the point.

Let’s say, for example, that a multinational company wants to build a skyscraper. And building that bridge required hiring a general contractor who farms work out to a bunch of subcontractors who employ in total 1,000 workers on the building. The construction cost of the building was $1 Billion. You could easily suggest that the cost of the project was $1 Million per job. But would that matter in any way, shape, or form? Not only does it not really account for the capital costs of all the equipment — cranes, trucks, tools, etc that those workers must use, it doesn’t account for the capital costs of the building materials themselves!

If I want to build a skyscraper, the vast majority of the cost is for the steel, drywall, wiring, piping, elevators, etc — materials. The cost of the workers is a slight fraction of the total. Likewise, if the government wants to build a bridge, or construct 15 miles of freeway, or engage in pretty much any other infrastructure spending, far more money will be spent on materials than on actual workers.

Now, don’t get me wrong. I am not trying to defend the stimulus. I’m not trying to defend the government’s numbers on job creation (because, of course, borrowing the money they needed to pay for the project may have crowded out actual productive enterprises that could offset those “created” jobs). But simply dividing spending by # of jobs to come up with some arbitrary (and absurdly high) “per job” cost is a cheap rhetorical device. It might sound great on Rush Limbaugh’s show or whip up outrage amongst people who don’t know better, but it certainly isn’t a serious analysis of the policy.