Walter Block criticizes Bernhardt and Scoones for discovering a peculiar “market failure” in the labor market.

The alleged failure invokes the idea that giving an employee a raise will “signal” to other entrepreneurs that the worker thereby blessed is more valuable than previously thought with the result that those entrepreneurs will have a renewed reason to try to steal the worker away.

Thus, entrepreneur A may reason as follows:

Smith is getting $60K / year. I think he’s a very productive guy and is rather underpaid and knows it. I don’t want to lose him. I think entrepreneur B could lure Smith away with a promise of $70K per year. To prevent this outcome, I myself shall offer Smith a $10K raise. But wait a minute. If I raise his salary, then B will surely notice this and come under the impression that Smith is a better worker than B has thought before. So, B will decide to offer Smith $80K / year. I can’t pay that much. So, I lose Smith either way, but at least if I don’t give him the raise, I keep the $10K. So, I won’t give him the raise.

The result is that the worker will not be promoted “as quickly or often as is socially optimal.”

Block offers a number of intriguing arguments against this claim. Some of them fail, like the idea that A can create a “poison pill” by promoting “employees who are totally underserving of this accolade, in the hope that B will snatch them up — and lose out thereby.” Wait, Walter, what? This requires that a highly implausible situation exists, namely that out of millions of firms out there, the poison pill will be swallowed by A’s direct competitor Q quickly enough and for long enough that the harm to Q will outweigh the harm to A. Most of the time, A will simply fire the malcontent.

Here’s my own argument. Consider the following illustration.

I took my cat to the vet recently and was told that the cat is getting fat in his almost-senior years, and one way to improve his health is to switch him to wet canned food diet. So, I went to PetSmart and was impressed by a large variety of canned food with prices ranging from ¢50 per can to $1.30 per can. Suppose then that I reasoned as follows: “I love my cat and want the best for him. The expensive food wouldn’t be such unless it were better. So, that’s what I’ll buy.” Obviously, no sane consumer reasons like that.

And indeed, I asked one of the store girls whether the cheaper brand was good. She was a little reluctant to affirm its “goodness” but told me it was their most popular brand. I was not surprised at this validation of the law of demand. Further investigation revealed some possible reasons why the more expensive cat food existed: some was “grain-free,” and so contained fewer carbs and more protein, some had purer ingredients, etc. I then decided whether the benefits outweighed the costs. (In the end, I bought a few different brands to test them out.)

The simple counter to Bernhardt and Scoones is that the price of a good is a very poor signal of the quality of the good. If B picked his employees based on how much they earned in other firms, B would quickly find himself staffed with expensive incompetents and disappear.

There are then two effects: (1) Smith comes to value his job with A more because of the raise; (2) B comes to consider Smith to be more productive than before because of the raise. Of these, reason and experience dictate that (1) should be far stronger than (2), such that the marginal benefit to A of a $1K raise to Smith outweighs its marginal cost, overall diminishes Smith’s job search activity, and so will be selected for.

Suppose, however, that I am wrong, and effect (2) is stronger than effect (1). Then Smith will still get his $10K / year raise by changing his job to come to work for B! (Even if B is now less aware of Smith’s abilities, Smith must himself suspect that he could do better. Otherwise, if neither B nor Smith himself think Smith is underpaid, then A has no reason at all to give Smith a raise. A will just sit back and enjoy his quasi-rent on Smith.) With their positions now reversed, A will now wonder whether Smith was a better worker than A imagined before Smith deserted him. Perhaps A will in turn offer Smith $80K to come back! Smith will still receive his raise via this somewhat circuitous route.

Suppose now that Smith does not deserve to be promoted by A. A’s marginal revenue of employing him is close to Smith’s marginal cost. If B steals Smith away and pays him $70K, then if Smith justifies his new salary with increased productivity in B’s enterprise, then we have an entirely economic and efficient re-allocation of a resource.

In short, I agree with Walter that there is no market failure, though for slightly different reasons.


Leave a Reply

Your email address will not be published. Required fields are marked *