This harkens back to an uncompleted argument I had with Eric Rauchway and never finished; I've been meaning to write this post for months. And since I think we're going to be having this argument again and again over the next months or years, it seems like a good time to do it.
Whether you should count relief jobs as part of the unemployment figure really depends on what you're trying to measure. If you're trying to measure something about the worker experience, it makes sense to count relief jobs: they go there and they pick up a paycheck. But if you're trying to measure how robust the economy is, you shouldn't measure those jobs, because relief jobs do not vary in the same way as the underlying job market.
This is, I must point out, not an argument about whether relief jobs are good or not--I certainly think they're preferable to other interventions, like long term welfare. It's just an argument about what about the economy the unemployment rate is supposed to be a proxy for. But we have other figures that already measure peoples' material conditions pretty well--the poverty level, f'rinstance. On the other hand, we don't have any very good figure to tell us the strength of the private labor market.
Here's my core problem with including relief jobs in the unemployment rate: the unemployment figures start telling you as much about the political situation as the underlying economy. The government can push the unemployment level down to nearly any arbitrary level it wants by putting anyone without a job on the dole and making them do some not-very-valuable task in order to cash their check. Just as you can push GDP up by the simple expedient of borrowing money (which does not show up in the GDP figures) and having the government spend it, you can lower unemployment by borrowing money and spending it on creating jobs with a value too low to justify creating them in the private sector.
But ultimately, the health of the economy relies on the ability of the private sector to generate growth, and the private sector to generate jobs. It's the tax revenue from that growth, those jobs, that allow the government to create government jobs. In the short term, the government can borrow and spend, but if the private sector and the labor market are not growing, eventually lenders will get worried about the ability to repay and raise the interest rates they charge. Result: well, there are a lot of deposed Latin American heads of state who could describe the result in some very vivid language.
If we include relief jobs in the unemployment figures, we lose track of the strength of the underlying private labor market that has to ultimately support those jobs.
Regardless of what the government does, our economy will not have recovered from the current mess until the private markets start growing again. The government spending may (or may not) be a good way to alleviate the pain of a weak labor market, but unless either the economy recovers on its own, or the massive stimulus shocks it back into full output, the stimulus will eventually have to be cut back, and we'll be back into the bad old equilibrium with a lot of extra debt payments making us even poorer. It behooves us, then, to keep a closer eye on the underlying figure than on what the government is doing in the short term.
And that's approximately the place the US found itself in at the end of the 1930s, with unemployment back up to 19% after falling into the low teens. The underlying economy was not sound. Perhaps the liberals are right and the problem is that FDR didn't spend enough--but that is a hypothetical. What we know is that the US economy was still in shockingly bad repair until World War II turned it around.
This article available online at: