Broken-Window Alert

How many jobs has the “stimulus” package created? Sensible economists of course know this question cannot possibly be answered. Government spending directs resources toward politically favored projects at the expense of others, but the net effect on resource allocation, including the net effect on “jobs”—a heterogeneous category that depends on hours worked, job characteristics, and other small details—cannot be measured with precision. There are also important intertemporal effects—e.g., expectations of inflation, tax increases, new government programs, etc., may cause firms to delay hiring workers they otherwise would have hired.

Not to worry, says John Irons of the Economic Policy Institute. Indeed, the Obama Administration’s silly attempt to count “jobs saved” may actually underestimate the beneficial effects of stimulus, because “that construction worker [hired by stimulus funds] then goes out and spends money at a local diner, at a local McDonald’s or the local movie theater. That could very well mean an additional job.” Well, yes, but there’s the construction worker who wasn’t hired because investors, spooked by regime uncertainty, wouldn’t fund the construction project; the construction worker who won’t be hired tomorrow when taxes are hiked to pay for today’s stimulus; the construction workers who will sit idle at home when hyperinflation destroys the construction industry; and many other members of Bastiat’s “unseen” who would have spent money at diners and McDonald’s and movie theaters. But we never see those lost jobs, so—poof!—they don’t exist, and shouldn’t be counted against the free lunches wished into existence by Stimuluspalooza.

Update: My colleague Mike Sykuta beat me to it.