Either AEI Has Forgotten Basic Statistics or They Are Advocating For More Collective Bargaining

Admittedly, I don’t read the American Enterprise Institute’s blog very much. Today, however, someone alerted me that they published this post suggesting that minimum wage laws are at least partially to blame for Europe’s employment woes. In it, Dr. Mark Perry presents a table of countries in Western Europe, their country-wide statutory minimum wage, if they have one, and their respective “jobless rates.”1 He then asserts:

“[M]inimum wage proponents… argue that other countries have minimum wage laws apparently without any adverse consequences on employment levels or jobless rates. The empirical evidence from Western Europe seems to suggest otherwise. Labor markets in the group of countries with no minimum wage laws are much healthier and doing much better than the group of countries with minimum wage legislation, measured by the jobless rates in Western Europe.”

To paraphrase, “Look: minimum wage, higher joblessness; no minimum wage, lower joblessness… just saying.” Although this argument has something of a timeless quality to it (pdf), it’s wrong, tissue-paper thin, and not something we would expect from any serious researcher.

First of all, as we learned in Statistics 101, there’s a difference between correlation and causation. Even if there appeared to be some pattern between minimum wages and unemployment, that wouldn’t mean that one is in any way causing the other. The only way to try to identify causality is to isolate as many—ideally all—other factors that might play a role in the suspected relationship through statistical regression methods. This is what researchers mean when they say they are “controlling” for other variables. Perry does none of this. Obviously, we wouldn’t hold a blog post to the same rigorous standard that we would expect of a more formal study, but Perry’s post is intellectually disingenuous. Fortunately, numerous other researchers have looked at this very question with the more rigorous and appropriate methods, and the best research shows that minimum wages have little to no effect on employment levels.

 I could stop right there, but since Dr. Perry has taken us down this rabbit hole, let’s go with it. Here’s the table from the AEI blog post:


Set aside for a moment the countries listed as “no minimum wage,” and take a look at the countries that do have minimum wages. If minimum wage laws do lead to higher joblessness, as Perry suggests, one would expect that the higher the minimum wage, the higher the jobless rate. According to this table, that’s not the case in Western Europe. The figure below is a simple scatterplot of the minimum wage rates and the jobless rates from the table. As you can see, under the superficial approach that Dr. Perry is viewing these data, higher minimum wages actually imply lower jobless rates. (To be clear, I am not—and would not—encourage anyone to view this chart as evidence of a causal relationship because I have not controlled for any of the numerous other factors that could affect these countries’ labor markets, and the sample is tiny.)

Now you’re probably saying “Yeah, but what about all those countries with no minimum wages?” Yes, I’ve excluded them from my graph, but not without reason.2 To say that these countries have “no minimum wage” is, at best, a half-truth. All of the countries listed here as having “no minimum wage” have broad collective bargaining, where industry-wide unions negotiate a minimum wage for all workers in their industry. This industry-wide collective bargaining doesn’t exist across all industries in all of these countries, but it is widespread. For example, in Denmark, the average negotiated minimum wage for all workers in both the public and private sectors was roughly $20 per hour (pdf). The Occupational Safety and Health Administration in Finland reports minimum wages set by collective bargaining in construction, metalworking, cleaning, and the hotel and restaurant industry to be $13.10, $11.65, $11.27, and $13.26, respectively. Germany’s lowest minimum wage, in the commercial cleaning industry, is roughly $9.50 per hour. I could go on, but I think the point is clear: claiming that these countries have no minimum wages is either woefully ignorant or intentionally deceptive.

Bottom line: There’s no “empirical evidence from Western Europe” that minimum wage laws have damaged labor markets. If Dr. Perry wants to argue that having broad, industry-wide collective bargaining to determine wages is better policy than country-wide statutory minimum wages, I would be intrigued to hear that argument—preferably with some rigorous analysis behind it.

1. Essentially the same thing as unemployment rates, but countries may measure unemployment slightly differently than one another.

2. Just for fun, even if you add the “no minimum wage” countries to the scatterplot with a minimum wage value of zero—again, not accurate—you get a trendline of y=0.0026x + 0.0873, with an R-squared value of 0.0385. In other words, even under the dubious claim that these countries have no minimum wages, you get a relationship that is essentially nonexistent.