Steve Pearlstein’s recent article suggesting that we need wage cuts to correct unemployment have gotten a lot of attention. To the extent that this is sold as a way to fix the general problem of unemployment, I agree with Yglesias that because this will increase the real value of household debt it’s probably not the best way of going about it. If we were in an inflationary environment, or even getting a normal level of inflation, and unemployment were persisting due to structural issues, then wage cuts could help as a general solution.

On the other hand, to the extent that wage cuts lead to more hiring it is identical to job sharing that so many progressives seem so fond of. I’m guessing those that support job sharing but not wage cuts are worried that the link between cuts and hiring will be broken if it is optional for employers. But then the problem isn’t debt deflation per se, but debt deflation without enough of an employment increase to offset it. In this case it’s really an empirical question that can’t be waved away by theory alone.

Overall I think the case of wage cuts as a general solution is not strong, but I agree with Pearlstein when it comes to particular industries. The auto industry for example is clearly undergoing structural changes; GM and Chevy did submit restructuring plans  after all. You can point to an undervalued Yuan as a cause for our lack of overall global competitiveness, but anywhere that tradeable goods industries have wages above domestic market levels devaluing the dollar will only get you so far.  Industries with strong labor presence will be among these- and yes there are fewer and fewer union workers but 13% of the private labor force is nothing to sneeze at. Minimum wages, other labor market restrictions, and non-regulatory labor frictions like debt induced labor immobility would apply here as well.

We’re not just concerned about employment in tradeable goods however, which is why, as Dean Baker points out, we should be looking to high skilled service industries for places where wages are above market level (although he overstates the case with banks).  We should be looking harder at the laws that cause these high wages and finding ways to loosen or repeal them.

We shouldn’t just be talking about private labor markets either. Matt Yglesias asks how we know when public sector workers are overpaid, and suggests a lack of a general problem, e.g. “[w]hether or not Michigan is overpaying janitors at state office buildings has no logical relationship to the appropriate compensation level of federal bank regulators.”

But we do know that a union wage premium exists in public sector just like it does in the private sector, so that anywhere that public employees belong to a union, wages are going to be higher than they otherwise would. Well how high would they otherwise be? This will be a function of how much local politicians can convince voters that they should be willing to pay for the services they receive, and what kind of wage/quality tradeoff the bureaucrats responsible for the particular public service decide. This won’t be perfect, but I can’t see any convincing reason why this would be systematically too low such that unions would improve it. Also, keep in mind here that unionization doesn’t just come with a wage premium, but a union preferred wage setting mechanism which can undo any quality increase that the wage premium might get you. Teachers unions come to mind here.

Finally, I’ve noticed that we are now talking about taking energy subsidies designed to ameliorate climate change out of the hands of the democratic system that brought us ridiculous ethanol subsidies and other debacles, and placing into a technocratic evidence based institution like the National Institute of Health. Maybe states will get desperate enough to bring solutions like this to labor markets, and appoint technocratic institutions to design labor market policies. Democracy fundamentalists will shudder at the notion, but governance is getting worse just when we need it to be getting better. Drastic measures seem in order. I’m not suggesting this is necessarily the best way to go, but it’s something we should be considering.