The reasoning in the previous post implies that an individual worker’s market power is probably lower than what we’d assume or estimate (however you’d operationalize market power) using homogeneous human capital. Caveat for what follows: I’m not a wizard with game theory and I know very little about bargaining in particular.
Roberts said that a worker’s market power derives from the size of that worker’s choice set. Let’s operationalize market power as “the ability to willfully influence one’s own terms of employment.” I can think of two jointly sufficient conditions for a worker to have any market power at all:
- the worker can credibly threaten to defect from wage bargaining,
- the employer would rather continue to bargain than let the worker defect.
It might be hard to prove, but I suspect that these conditions are also necessary. If they are, then workers’ market power might not be that large in general. Net of social institutions, how much can a typical worker ask of an employer? This framework could probably be generalized to any bargaining situation, and I’d be shocked if nobody has written about it.
One interesting “prediction” it generates: consider a type of workers who are afraid of losing their jobs, and who employers know which workers are of this type. How credible is their threat of defection? How much market power do they have?
I think there are a lot of workers that fall into this category at least partially.