At long last, Kathy G comes back on the minimum wage.

Most of the post is responding to an argument I have not made: that the minimum wage decreases employment. To be sure, I might make that argument--but then again, I might not. I think the empirical evidence is ambiguous, and though the bulk of it reinforces the intuitive belief that if you raise the cost of labor people will use less of it, there is a plausible argument that this is the result of publication bias. (Just to be clear, I haven't investigated this argument, so I am not endorsing it; I just think it's possible.)

I still think that the minimum wage is terrible policy, but that rests more on the fact that most of the people who receive it are not poor, and we have much better, more direct ways to target poverty than possibly introducing structural rigidities into the labor market, or encouraging employers to make working conditions worse. Whatever unemployment effect there is of US-sized changes in the minimum wage is too small to be detected amidst statistical noise.

So saying, for example, that Card and Krueger redid their study and found "no decrease in employment" is not useful. In fact, it's a big shift in the goalposts: before we were arguing about whether employment increased.

There's been a lengthy back and forth over Card and Krueger, and I'm about as sick of it as Card and Krueger are. People who want to believe that you can jack up wages ad nauseum without creating unemployment will abuse Card and Krueger to support ACORN's living wage programs. People who think that raising the minimum wage by a penny could plunge the economy into a new Great Depression are going to dispute the thing even if Jesus Christ Himself walks across the Potomac to endorse its conclusions. I can no more alter their fixed opinions than I could stop an earthquake by holding onto the ground. Especially since my personal opinion--"most work on price controls suggests that they're probably wrong, but it's hard to say for sure"--leaves me in a poor position to haul out the sword of righteousness and start laying into the infidels.

So back to the original argument: does the low wage labor market look like a monopsony? That is, does it act as if there is essentially one employer? My answer remains . . . ummmm, no.

Ms. G's argument relies heavily on the question that invariably gets deployed by the commenters I call The Jedi Masters of Econ 201: "If you lower wages by one penny, will everyone quit? If not, you've got a monopsony". The technical answer is that we don't know, because the government keeps employers from finding out.

But the real answer is that this is trivial. Almost no markets in existence look like the models in the economics textbooks; if Ms. G is going to reject generally applicable market principles in any situation that doesn't exactly correspond to perfect competition, then she should stop taking economics classes now; none of the models will be any use to her.

We don't have much trouble identifying Standard Oil or DeBeers as monopolies even though they are not actually the only companies in their markets and did not actually have perfectly unlimited pricing power. We have little trouble understanding that the US car market is no longer an oligopoly, even though the actual number of sellers has not actually expanded that much. Even commodity markets do not always look like perfect competition.

But despite the fact that the housing market is not perfectly competitive, rent controls still degrade the housing stock and restrict supply. Cap and trade works a zillion times better than simple fiat legislation. Etc. Etc. Etc.

Since wages rarely fluctuate by a penny in these inflationary times, the correct question is not whether such small fluctuations engender a perfectly competitive reaction. The right inquiry is whether noticeable wage gaps cause workers to move. Last time I looked, they did.

Likewise, noting that some workers have heterogenous preferences does not get you very far. Virtually all real world markets have a spectrum of customers with varying degrees of price sensitivity, depending on their affinity for other variations in the product. Nonetheless, they tend to function basically as you expect them to: when companies raise the price of their products, customers react by consuming less of them.

Basically all of these arguments seem like quibbling. "Monopsony doesn't require that turnover rates be low"--no, it doesn't, but you'd better have a good explanation for why it's so high if preferences are so heterogenous. You'd also better be able to demonstrate that you can keep wages low while others raise them and still have an adequate labor force. There may be data showing that this is true in the broad mass of the minimum wage market, but I have not seen it.

The core question is: if minimum wage labor markets are a monopsony, how come we don't find more evidence of statistically significant increases in employment following minimum wage increases? One answer is that it is a monopsony, but employers are not using their pricing power to supress wages/employment--in which case, I'm not sure why we care. Or possibly Ms G, like me, thinks that the changes in American minimum wage policy are generally too small to pick the employment effect out of statistical noise. So perhaps a different way to think about it is this: how much money would you be willing to bet that if a city enacted a $12 minimum wage, employment in that city would grow faster than employment in the surrounding area? I'd put $1,000 on the opposite happening. And I'm not much of a gambler.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.