Yes, Wages are Sticky; No, I’m not a Keynesian
Bryan Caplan believes that sticky wages are partially the product of government intervention, but much more significantly a product of behavioral psychology. He has recently gotten excited about it and challenged anti-keynesians to register their responses to Truman Bewley’s magnum opus on sticky wages with the presumption being that significant and persistent unemployment caused by non-legislatively induced wage stickiness suggests expansionary monetary/fiscal policy as the only reasonable solution. If you’re a GDP fetishist, you’re an interventionist plain and simple, if you oppose involuntary unemployment (lost gains from trade), these facts would seem to recommend intervention as well. I see this as a watershed in the liberty vs. efficiency debate, where the advocates of efficiency break toward Keynes and the advocates of liberty must bear some lost gains from trade to respect the rights of savers, ill-positioned (vis a vis stimulus) businesses, and, most interestingly, the unemployed and those businesses who refuse to hire them at lower wages.
First off, you might be wondering how I think the liberty vs. efficiency debate even applies here. I take liberty to imply both complete freedom to use those resources to which you are entitled and complete responsibility for what comes of those decisions. One thing people are entitled to under any reasonable version of entitlement theory is their psychological disposition, so long as they bear the full cost to themselves and others. The disposition at issue is the common unwillingness to be a reasonably productive worker when you (wrongly) feel you’re being undercompensated. The cost of that disposition to you is “involuntary” unemployment. If I could wish that cost or disposition away so you could rejoin the workforce and add to society’s bounty, I absolutely would. Unfortunately, the Keynesian remedy comes at a greater cost than a passing thought and pair of crossed fingers; it requires a reallocation of resources away from the distribution currently geared toward those who’ve managed to justify their wages and it punishes those people who appreciate the difference between nominal and real wages enough to bid their numbers down without sacrificing productivity (and those who’ve cultivated the willingness/ability to recognize this from a hiring perspective).
The end result? You’ve tricked a relatively undesirable portion of the workforce into thinking they’re worth more than they are by diminishing the real compensation of the relatively productive. And you’ve done all this without the consent of those who’ve borne the cost. There are two inclusive arguments against my point here 1) If the employed had all the relevant information, they’d consent to it because 2) re-employing unutilized resources meaningfully grows the economy and compensates for a very large part of any transfers that took place. Liberty and efficiency!
The first point doesn’t quite measure up: the proposal is a risk and a difficult-to-gauge one at that; surely there’d be plenty of risk-averse wage earners not willing to bear the costs of misallocation and/or inflation. On the second, the size of this compensation is not at all clear, and there are plenty of reasons to doubt it would recoup the social cost. Among these is the fact that, in most cases, businesses have discretion with regard to who gets laid off, and the body of the long-term unemployed population will largely be comprised of low skill/wage ratio individuals. That means higher nominal wage targets will need to be met with more aggressive (inflationary/distorting) stimulus to add some of the most relatively low-productivity workers to (hopefully) relevant/valuable sectors of the new economy. The act itself subsidizes their problematic dispositions, and the results are not obvious boons either. You’ve just about tricked a handful of people into not being zero-marginal-product workers; not only does this violate liberty, it likely comes well short of clearing the – thoughtfully construed – efficiency bar as well.