Short-term stimulus

There have been a lot of claims made about jobs “saved or created” by the so-called “stimulus” (and how you measure a job “saved” when you don’t know the might-have-beens, I have no idea); but has it occurred to you to wonder how long those jobs have lasted? Apparently, not very long in some cases:

How much are politicians straining to convince people that the government is stimulating the economy? In Oregon, where lawmakers are spending $176 million to supplement the federal stimulus, Democrats are taking credit for a remarkable feat: creating 3,236 new jobs in the program’s first three months.

But those jobs lasted on average only 35 hours, or about one work week. After that, those workers were effectively back unemployed, according to an Associated Press analysis of state spending and hiring data. By the state’s accounting, a job is a job, whether it lasts three hours, three days, three months, or a lifetime. . . .

At the federal level, President Barack Obama has said the federal stimulus has created 150,000 jobs, a number based on a misused formula and which is so murky it can’t be verified.

When even the AP is noticing that Democratic politicians are playing games with the numbers, you know it’s hard to ignore. (Though it’s worth noting that the AP appears to be trying to hide the story, judging by the fact that this link is down.) It should of course be pointed out that Oregon’s behavior here is uniquely egregious:

Oregon’s accounting practices would not be allowed as part of the $787 billion federal stimulus. While the White House has made the unverifiable promise that 3.5 million jobs will be saved or created by the end of next year, when accountants actually begin taking head counts this fall, there are rules intended to guard against exactly what Oregon is doing.

The White House requires states to report numbers in terms of full-time, yearlong jobs. That means a part-time mechanic counts as half a job. A full-time construction worker who has a three-month paving contract counts as one-fourth of a job.

That said, the response from the state to that criticism is telling:

Oregon’s House speaker, Dave Hunt, called that measurement unfair, though nearly every other state that has passed a stimulus package already uses or plans to use it.

“This stimulus plan was intentionally designed for short-term projects to pump needed jobs and income into families, businesses and communities struggling to get by,” Hunt said in a statement. “No one ever said these would be full-time jobs for months at a time.”

But wasn’t that the implication? After all, when the President talked about “3.5 million jobs saved or created,” he didn’t add the caveat “but only for a little while”; an extra week’s worth of work is not nothing, to be sure—I’ve been a temp, I know the drill—but if that’s the best the government can do, your job hasn’t been saved, your job loss has just been delayed a bit. And when most people talk about “job creation,” temp work is most certainly not what they have in mind.

The truth is, this story from Oregon highlights how fuzzy and dubious these job claims are even when the politicians aren’t playing games with them. As the Reason Foundation’s Anthony Randazzo points out,

The problem remains that there is still no good way of counting exactly the number of jobs that wouldn’t have been lost because of the savings, and there is no way the government is going to track the number of jobs that have been lost because of stimulus spending (such as lost jobs in traditional energy because of green spending).

Put another way, all such claims depend on a knowledge of the might-have-beens—if we hadn’t done this, what other things might we have done instead, and what results would they have produced? And what would have happened if we hadn’t done anything at all?—and that’s knowledge we don’t actually have in any reliable way in most cases, and particularly when you’re talking something as complex and interconnected as the national economy.

HT: David Riddle