Zero marginal productivity

August 11, 2010 at 4:30 am

| Gabriel |

One of the intellectually exciting but practically horrifying ideas being discussed recently on econ blogs is the “zero marginal product” worker (see here and here). That is, labor that contributes nothing to the firm, or at least doesn’t contribute enough to justify the minimum wage and overhead. If you take this seriously and game it out, you’re faced with the prospect of large-scale permanent unemployment which is not only a bad thing for the underclass themselves but tends to produce politics of redistribution that look a lot less like TH Marshall or John Rawls than Gaius Gracchus or Hugo Chavez.

Anyway, I was particularly interested in this post (h/t MR) because it makes explicit that once you take into consideration the possibility of catastrophic failure, you can even have workers whose marginal productivity is negative, at least in expectation. The logic underlying negative productivity gets into theoretical issues reviewed in my paper (with Esparza and Bonacich) on spillovers in Hollywood, but the theoretical Ur-cites for this downside model are Jacobs in sociology and Kremer in economics. Basically, every so often a worker will create a catastrophic failure. If the probability of creating a screw-up is a function of skill then it makes sense to avoid low skilled workers, even if on the median day they would have nontrivial positive productivity. The scary thing about negative productivity (in expectation) is that it implies that there might still be some nontrivial permanent unemployment even if we tried every plausible policy intervention to make hiring low-skilled workers more attractive: a lower minimum wage, eliminate the payroll tax, single-payer (which would make moot the current stigma on not providing health benefits), reasonably generous wage subsidies, etc.

Entry filed under: Uncategorized. Tags: .

Some ways Stata is an unusual language Getting long flat-files out of field-tagged data

The Culture Geeks

%d bloggers like this: