I promised a few weeks ago to "write a further note explaining a more-sophisticated way of understanding the harmful effects of the minimum wage." This isn't it.
The reason is that three issues came up in the comments and e-mails and I want to handle them first. The promised follow-on post will come in another few days.
The three issues were:
1. Arguments for the minimum wage as a way of helping some low-wage workers even if the employment effects are negative.
2. Questions about why imposing a minimum wage in a monopsony situation could increase employment and efficiency.
3. A question about why unskilled workers are actually the least likely to face monopsony.
Here are my responses in order.
1. I should have specified in my original post, but didn't, that I was giving the student who asked the original question the best (and really only) argument one could give for a minimum wage increasing efficiency. It's easy to come up with arguments for a minimum wage when you equally weight the dollar losses to those who lose their jobs and the dollar gains to those who get higher wages. But then you're leaving out the main losers: employers and, ultimately, consumers who pay somewhat higher prices for the goods and services produced by the minimum-wage workers. I was excluding these kinds of arguments because, to repeat, I was looking for only pro-efficiency (in the economist's sense of efficiency--gains from the change in $ terms exceed losses in $ terms) arguments.
2. The student at Susquehanna University told me by e-mail that he didn't understand the monopsony argument. I still have not got around to learning how to put a supply demand graph on Econlog. I will learn, but not today. Fortunately, I don't have to reinvent the on-line monopsony graph. Here's the best one I have found.
In the figure, Figure 14.9, if the employer is a monopsonist unconstrained by a minimum wage, he will hire Lm of labor because at that point, his marginal revenue product (MRP) equals marginal factor cost (MFC). Reading down from the intersection of the two lines to the supply curve (S), we get that the wage is $4 an hour.
But the efficient amount of labor is the amount where the supply curve intersects MRP. That point, unfortunately, is not labeled in the graph. You can tell, though, just by eye-balling, that that point is at a wage of approximately $4.70 an hour. So if the government imposes a minimum wage of $5.00 an hour, it will cause the MFC curve to be at $5 up to the point where $5 and the supply curve intersect, and then will jump to the old MFC. With this new MFC curve, they will employ L2. Notice that is more than the number of people employed when there was no minimum. QED.
The government could make the situation even more efficient by setting the minimum wage at $4.70 an hour rather than $5.00 an hour. I leave that as an exercise for the reader.
3. In my previous post, I wrote:
I don't find the monopsony claim persuasive. Monopsony requires that the employer have no or few competitors trying to hire the same kind of labor. It is precisely the fact that the workers are unskilled that gives them many potential employers.
In response to this, Sam Raptis wrote: Could anyone explain the reasoning behind Henderson's claim toward the end of the article that "It is precisely the fact that the workers are unskilled that gives them many potential employers." is true? I'm not doubting it, but I found it confusing since I would tend to think the opposite.
Yes. Many, many potential employers want relatively unskilled labor for unskilled tasks. So there are many potential employers for the unskilled. Go to Home Depot sometime and watch how many people drive up with cars and pickup trucks and hire (in my area, anyway) Hispanic workers for a day. Now imagine that the worker is an astronaut. This is highly skilled. How many potential employers are there? Or, imagine that the worker is a professional basketball player and is highly skilled. How many potential employers are there?