≡ Menu

Does the minimum wage reduce employment?

In the Intelligence Squared debate I participated in last night on the minimum wage, Jared Bernstein argued (and argues again here in his follow-up at the Huffington Post), the impact of the minimum wage on employment is an empirical question. Based on his review of the literature, Bernstein finds that the impact is zero:

The whole point of the new research — and I’m talking about work that finds both positive and negative impacts — is that those impacts hover around zero, which should lead objective observers to be highly skeptical that phasing out the minimum wage would lead to large employment gains.

Could be. Large is a relative term. Certainly, the lower the minimum wage, the smaller the effect and the harder it will be to observe its impact in the data. Many minimum wage studies focus on teenagers or younger workers or workers without a high school diploma. But even among these workers, most of them earn much more than the minimum wage. So the minimum wage is not binding. But that only means the effects are small and hard to identify empirically, not that they are zero.

My argument is that most workers don’t need the minimum wage so getting rid of it will have a limited effect, lowering compensation among the least skilled of the least skilled. Most of us earn much more than the minimum, even teenagers and workers without a high school diploma. Were we to abolish the minimum wage, most of us would be unaffected. But while some would get paid less if the minimum wage were abolished, there’s a big benefit–others would be able to get a job in the first place. Why would you want to legislate higher wages for the poorest workers, knowing that the main group that loses out is even poorer workers?

So while I am glad that the minimum wage is only $7.25 because its harm is relatively small and hard to measure, it is still worth abolishing and certainly should not be increased.

Here is how Bernstein couches our argument:

…allow me to summarize the opposition’s argument, half of which is, I think, a fair point from a libertarian perspective, though one with which I deeply disagree.

Their first point is that the minimum wage hurts a lot of people. But a) that’s not what the research shows (even the bulk of the work that finds some negative impacts shows that the vast majority of affected workers benefit from the policy), and b) half the time they argued that it’s a small policy that affects few people so getting rid of it won’t be a big deal. Karen and I tried to figure out how a small, ineffectual program could be so damaging to America that it had to be abolished but I thought they were quite muddled on this point.

This is an interesting straw man. Maybe I wasn’t clear enough. The point is that small isn’t small when it’s your job that’s lost or it’s you looking for work. The fact that hundreds of thousands of workers can’t find work in an economy of 135 million workers is going to be very hard to uncover empirically while the few million that earn the minimum wage or less (and ignoring that many of them earn tips) are very easy to see. But just because the numbers of unemployed are small relative to the entire labor force doesn’t mean the policy is no big deal.

I guess if you think the empirical results that find zero effects are the truth (rather than a failure of econometric precision), then the minimum wage is a free lunch. I don’t believe those empirical results. First, there are lots of criticisms of those results. But the point I tried to make last night is that the reason low-skilled workers have struggled so much in recent decades is because they have found themselves competing with robots and automation and computers and foreign workers. Everyone, on the left and the right, agree that employers are eager to save costs and will substitute machines for workers or outsource production if those changes are profitable. Why will artificially higher wages created by minimum wage legislation not lead to similar substitutions?

The weird part of the debate over the magnitude of the employment effects, is that when someone uses the reductio ad absurdum of a minimum wage increase to say $50 or $100 an hour, everyone understands that won’t work because it would destroy the labor market. So where do those disemployment effects kick in? If the minimum wage is small enough so that it doesn’t cause job losses, then it can’t be having much of an effect boosting wages.