Monday, February 18, 2013

Some thoughts on the minimum wage: Theory

A lot of people act like economic theory provides an obvious answer to the minimum wage: it's supply and demand. If you put a price floor on something you get a surplus. Everyone knows that. Everyone teaches that. I taught that to freshman too - not just the price floor point, but using the minimum wage as an illustration.

So aren't people denying basic economics when they express doubts that the minimum wage hurts employment?

Well, no. We lie to students all the time. That's called "teaching" if it's a useful lie. Teaching is just giving students a useful set of ideas for understanding the world, and all human ideas for understanding the world are to some extent contrivances or simplifications. Introducing students to supply and demand is a very useful lie. There are a lot of senses in which economists have moved past the simple supply and demand story that are relevant to the minimum wage, and I'll review a few here.


We'll start with one of the oldest. Everyone knows about monopoly. Behaviorally, firms with monopoly power are no different from firms without monopoly power: they both set output at a level that maximizes profit. The difference is that when a monopolist does this any change in its output level changes the price, while a competitive firms output decisions have no appreciable impact on the price. So monopolists can increase their profits in two ways: by making prices higher or costs lower. Competitive firms typically only have the option of minimizing their costs. The result is that monopolists use both tools at their disposal and end up producing less at a higher price.

The comparable case for a firm that is the only demander in a market is monopsony. This is relevant if we're thinking about labor markets because firms demand labor. You don't have to think about complete monopsony - you can have a few firms with monopsony power (oligopsony) or a lot more with some monopsony power (monopsonistic competition). The latter is probably the most realistic, but the principle is the same.

Like monopolists, monopsonists can earn profits by either pushing down the cost of labor or by pushing up the productivity of labor. Competititve firms can only adjust the amount of labor they demand to push up the productivity of labor. The critical point here is that if you were to push up the cost of labor (say, through a minimum wage law), the monopsonist would not suddenly hire less workers. They would just lose some of the excess profits they received from their monopsony rents. In fact, they will hire more workers assuming they are below their most efficient utilization of labor (which is a fair assumption).

It seems to me this is very reasonable to assume for teenage labor. Teenagers are unskilled and usually unable to go far afield for jobs. They often are doing other things (like school), they are often not able to move out of their parents' house, and they may be constrained when it comes to transportation. They also don't have much of an employment history that gives them independent advantage in a labor market, so their current employer has relatively more monopsony power over them. He or she holds private information about the teenage employee. Of course, the teenage job market is also not particularly hard to break in to (teenagers have to be able to break into it, after all), so it is an empirical question. But there is a considerable case for why employees would hold market power over teenagers, and that would lead to a minimum wage that does not negatively impact and may even positively impact employment.

It's perfectly legitimate economic theory. It's not fancy economic theory either. Robinson wrote about this back in the 1930s, and monopsony is taught to undergraduates as well (often in Econ 101, just a couple weeks after teaching about price floors. So don't let anyone pull the wool over your eyes on that.


Monopsony theory suggests that there's something to this idea that "firms can afford it", because the minimum wage is just cutting into monopsony rents earned by the firm, which it can do without threatening employment. A more sophisticated approach to this is to think in terms of a bargaining model. Employment contracts are struck when a firm and a work find an employment relationship mutually beneficial. But if there is positive surplus associated with employment, there are a variety of wages that would be found mutually beneficial. Economists often construct bargaining models that determine how that surplus is divided (this, admittedly, is not in Econ 101). In the context of a bargaining model, we can think of the minimum wage as the teenage worker bringing the federal government to the bargaining table. Now if the minimum wage is too high, of course some contracts might not be beneficial for employers anymore. But presumably there is some surplus available (and presumably employers are enjoying a lot of it, so a minimum wage would just shift some of that surplus towards teenagers.

This is also good, standard economics. There's nothing fishy or leftist or heterodox about bargaining models. It is somewhat higher level than monopsony.

Search and matching

The Dube, Lester, and Reich work I noted in the last post brings turnover costs into the discussion offering a good opportunity to talk about search models. This is also not fancy or leftist or heterodox economics - it's as legitimate as the stripped down supply and demand and it has had contributions from Chicagoans like Robert Shimer and Obama nominees like Peter Diamond. If there are notable job search frictions, a minimum wage can influence employment through a variety of mechanisms. It may impact the rate at which labor contracts disolve because of an efficiency wage situation where the higher minimum wage makes the worker less likely to either shirk or quit. The higher wage rate may force firms to be more careful about their matches, forming more durable matches and reducing turnover. Declining turnover in general could reduce unemployment caused by congestion and other structural problems. I did some quick googling and search theory approaches seem to be able to come up with a range of answers, as one might expect. That of course makes it an empirical question (and we discussed that in the previous post). The point is that search theory - good, standard economics - provides complications to the simplistic freshamn supply and demand argument.

A bad theory: higher marginal propensity to consume

Something you'll hear is that teenagers consume more of their income so a higher minimum wage will increase labor demand. I think this might be a good macroeconomic reason to support a minimum wage increase, but I don't think it gets us anywhere on the labor economics side of thinking about the employment effect of the minimum wage. It's the same mistake that Henry Ford makes in justifying a higher wage so that workers can buy back the product: teenagers don't spend a lot of their money buying burgers at the same restaurant where they're employed and auto workers are not going to be accounting for much of Ford's revenue. It is legitimate to talk about inequality, consumption, and aggregate demand. It is not legitimate to talk about these effects in the teenage labor market specifically. If you employ teenage labor the labor costs associated with your employees far outweigh the revenue coming from your employees' wallets.


  1. Concerning the last point, ain't that another convenient lie to push our thinking in the right direction? In reality nothing on the microeconomics level of complexion is insulated from the changes of the macro-variables, and macro is not independent of what is happening with micro either. For this reason I don't think there is a lot of sense in resolutely rejecting macro-concerns in the microeconomic questions.
    Also, I don't think you analogy with Ford's management decisions is correct simply because Ford was only a single businessman but the proposal for the minimal wage is country- and so economy-wide.

  2. "In fact, they will hire more workers assuming they are below their most efficient utilization of labor (which is a fair assumption)."

    Can you explain this part? I can see how it could have no effect on the hiring of workers, but I can't see why it would cause more to be hired.

    1. Monopsonists face the market supply curve and only demand less than the socially optimal because it's profit maximizing. Raising the wage rate puts employers on a portion of the supply curve that is higher than the profit maximizing point but lower than the socially optimal point.

  3. I think the higher MPC reason doesn't make much sense for another 2 reasons:

    1) Consumption is not the same thing as AD.

    2) Boosting AD is good when there is a demand shortfall. Not all the time.

    3) if the minimum wage did not exist, we might see more people employed at an even lower rate and therefore with an even higher MPC.


All anonymous comments will be deleted. Consistent pseudonyms are fine.