Note to Megan McArdle

Contrary to what you say here, the monopsony model of labor markets does not require employer collusion. All it requires is that there are important frictions in labor markets (such workers' heterogeneous preferences, incomplete information, and mobility costs), and that employers set wages. Or, in technical terms, that the supply of labor to an individual term is not infinitely elastic -- i.e., if an employer cuts wages by one cent, all the workers won't immediately quit. 

You might want to take the trouble to try to understand the monopsony model (this book is the standard work on the subject) before you mischaracterize it this way.

A few other points: although the monopsony model challenges the standard model of perfect competition, it's a model that I believe is very much within the neoclassical tradition, in that it depends on rational actors, maximizing behavior, and decisions based on the margins. And the economists of my acquaintance who believe that monopsony may, under many circumstances, be a more appropriate model for labor markets than perfect competition are by no means of "the left," contrary to your characterization.

Finally, the monopsony model is basically a theory of how labor markets work. To refer to it as "a monopsony model of minimum wage employment" doesn't make a whole lot of sense, because the model can be used to analyze many features of labor markets -- not just the the minimum wage.

Comments

[Megan McArdle]

Mmmmm, yes. But I find the idea of substantial search friction in the retail and fast food industries even less plausible than some sort of signalling collusion. You're talking about industries with annual turnover well in excess of 100%; I've heard numbers over 1000%.

[Megan McArdle]

Also, I'm specifically referring to the debate over Card and Krueger, which at this point is often a debate about whether there is monopsony in local minimum wage markets. I am well aware of the limits of my expertise, but it's pretty hard to get out of half a semester of Micro without knowing that monopsony is a general model not only applied to labor markets, but anything demanded. I was referring--perhaps insufficiently clearly--to a model of minimum wage markets that would permit Card and Krueger's finding that employment had risen after a minimum wage increase to be an actual relevant datum rather than an obvious fluke.

[mq]

Megan McArdle shows up to demonstrate her ignorance of economics yet again. Megan, are you or are you not aware that rapid turnover creates employment frictions that replicate the key aspect of monopsony (namely: a supply curve to the firm that is less than infinitely elastic), and can therefore lead to a monopsony result? That Card and Krueger outlined exactly this model almost fifteen years ago in the book they wrote about the minimum wage? That it has been referred to and discussed numerous times in the economics literature since then?

It is painfully obvious that as you say, your only economics training is half a semester of intro micro. But haven't you taken advantage of any of your time and leisure since then to do any reading or learn anything about the topic you supposedly specialize in?

Unfortunately, you're far from aware of the limits of your expertise -- you don't even know enough economics to know what you don't know.

[Tim Fowler]

mq - A less than infinitely elastic supply curve doesn't amount to a monopsony. If it did almost all markets would be monopsonies. Sure its a requirement for the monopsony to actually be a problem (if the supply curve for labors is such that a penny's decrease in wages would cause everyone to quit than the monopsony employer wouldn't be able to drive down wages), but its not a requirement for a monopsony to exist, or an indication that a monopsony does exist.

Its obvious that you don't strictly speaking have a monopsony in the market for low end labor in most places. A monopsony is a SINGLE buyer, and that's rare. You could have a dominant employer who is a near monopsony buyer of labor, or you could have a cartel, but those aren't common either.

Infinite elasticity is rare, but you can't reasonably say any situation without it is a monopsony, or even a problem. The employer's demand for labor isn't infinitely elastic either, and if an employee doesn't like the pay or conditions a particular job there are other employers in most situations.

The standard model is obviously not entirely correct, you don't have perfect competition, and in a number of situations you don't have near perfect competition, but "less than perfect competition" isn't the same as "monopsony" (which would be no competition at all).

Now the monopsony model for employment, doesn't actually require a literal monopsony (only one buyer for labor), but in most locations, the number of places you can get a low wage job is rather high, and in many cases they cluster to a great enough degree that if you can check wages at one you can check wages at another without too much difficulty. Its possible some specific individuals, in rural areas, do face something close to an effective monopsony buyer for their labor, but such relatively rare examples wouldn't be the driving force behind the wages for most low wage employees.