Bryan Caplan  

Resentment, Information, and Unemployment

My Private-Sector Complaint... Testing Unflattering Claims Ab...
There's one main problem with Tyler's post on the unemployment of the poor.  He writes:
Perhaps low-skilled workers cannot be employed at lower wages because their resentment at the low wage would be so high that they would impose unacceptable morale costs on the organizations employing them.  In other words, insult them with a sub-par wage offer and they turn destructive toward the entire organization.  Companies of course prefer to keep these workers at arms' length under this hypothesis.
The problem?  From context, it sounds like there's symmetric information about resentment, so each individual's resentment causes his own unemployment.  In reality, though, there's asymmetric information about resentment, so the average resentment poor workers feel causes unemployment for poor workers in general, including poor workers who are perfectly content.

Think about it like this.  A firm asks job candidates, "You earned $15 per hour in your last job.  This job pays $10 an hour, but you'll work side-by-side with other workers who earn $15 an hour to do the same job you'll be doing.  Are these conditions going to make you resentful, reducing your productivity to less than $10 an hour?"  Almost every candidate will respond, "Of course not!  I just want to work."  But many of these candidates are lying - to the interviewer and possibly themselves.  Since firms have trouble telling the liars from the truth-tellers, hiring anyone may be unprofitable.  If so, a worker who genuinely feels zero resentment, who is worth more than the market wage, ends up involuntarily unemployed.

How the math works: Suppose that in the preceding example, resentful workers produce $5 per hour, while content workers produce $11 per hour.  There are two resentful workers for every content worker.  With symmetric information, the firm will hire the one-third of workers who are content, and spurn the rest.  If the firm can't distinguish the two types of workers in advance, however, the firm will look at the expected hourly productivity of $(2/3*5+1/3*11)=$7 and hire no one.

Why not just cut the wage to $7?  The minimum wage aside, cutting the wage to $7 will provoke even more resentment.  Resentful workers' productivity might fall to $2, and the share of content workers might fall to 10%, reducing expected productivity to $2.90.  And so on.  As I've said before, "Encouraging individual workers to be more flexible in their wage expectations is still helpful advice.  But it would be even more helpful for the average worker if the average worker became more flexible."

COMMENTS (4 to date)
AS writes:

If it was profitable to pay all workers at least the minimum wage, firms would already do so absent coercion.

ThomasH writes:

This suggests that in addition to sensible monetary policy (targeting NGDP growth) and eliminating the wage tax on the first x thousand of of wages, it might be good policy to raise the EITC during periods when NGDP is below target.

Matt Moore writes:

I've always found the Shaprio-Stiglitz no shirking condition to be one of the most compelling arguments for efficiency wages.

I don't find this one compelling. In very low wage jobs, you can quickly observe each worker's output, which undermines the whole model.

YS writes:

This is only a narrower manifestation of a more general problem - how to keep track of productivity in an organization. Suppose a worker is paid his reservation wages. How do you know something else in the organization or his personal life does not make him resentful and unwilling to work to the extent of productivity suggested by the wage? He may resent his female boss, the nature of work, the need to take "sensitivity training" or he may be under the impression (correct or incorrect) that everyone else is dragging his feet .....

Comments for this entry have been closed
Return to top