In my last post, one anonymous commenter challenged me on the data of minimum wage. Rather than burying what is an important response deep within a post, I will first copy the critique, and then provide a detailed response.

Dr. Hammond,

You paint a pretty alarming picture of the effects of a minimum wage increase on unemployment and business operation, but the available research doesn’t seem to support your conclusions. I’m sure you are aware of the quantity of studies that have been done on this topic and how the results can at best be described as inconclusive. There are studies which claim to show negative effects of minimum wage increases on unemployment and there are many others which show either no effect or minimal effect. The main example you give is the restaurant industry, but the study cited here https://news.cornell.edu/stories/2016/01/restaurant-industry-unharmed-modest-minimum-wage-hikes found no negative effect on restaurant closings as a result of modest minimum wage increases. Also, the US has a very low minimum wage compared to other advanced countries with similar unemployment rates. Overall, the preponderance of evidence seems to show that fears of mass unemployment over minimum wage increases may be exaggerated to some extent. Given that, isn’t the principle that one should not work 40 hrs a week and still be in poverty upholding the biblical teaching of the importance and dignity of work? If work is to be valued and upheld as a virtue, shouldn’t that be recognized by compensating the worker with an amount that is at least adequate for them to live off of?

As Mark Twain said, and more recently popularized with former President Ronald Reagan, “It’s not what people don’t know that’s the problem; it’s what they know that ain’t so.” Anonymous, (not unlike most of us when we don’t discipline ourselves), is engaging in confirmation bias–listening to the studies which support his/her point of view. When you look at this issue systematically, you’ll see the usual economic result–there is a reason why the vast preponderance of economists think that a minimum wage is going to have negative consequences, to some degree, to those that are the least skilled. Here is the abstract of the most recent systematic study (Jan 2021):

The disagreement among studies of the employment effects of minimum wages in the United States is well known. What is less well known, and more puzzling, is the absence of agreement on what the research literature says – that is, how economists even summarize the body of evidence on the employment effects of minimum wages. Summaries range from “it is now well established that higher minimum wages do not reduce employment,” to “the evidence is very mixed with effects centered on zero so there is no basis for a strong conclusion one way or the other,” to “most evidence points to adverse employment effects.” We explore the question of what conclusions can be drawn from the literature, focusing on the evidence using subnational minimum wage variation within the United States that has dominated the research landscape since the early 1990s. To accomplish this, we assembled the entire set of published studies in this literature and identified the core estimates that support the conclusions from each study, in most cases relying on responses from the researchers who wrote these papers. Our key conclusions are: (i) there is a clear preponderance of negative estimates in the literature; (ii) this evidence is stronger for teens and young adults as well as the less-educated; (iii) the evidence from studies of directly-affected workers points even more strongly to negative employment effects; and (iv) the evidence from studies of low-wage industries is less one-sided.

But let me take this further–it goes well beyond a disagreement with the data. First, we need to understand why there are dissenting studies to orthodoxy. If you are trying to get published, which study is more likely to get attention: Dog bites man, or man bites dog? We know the answer; no one cares that a dog bites a man–that happens routinely. But a man biting a dog? Now that’s a headline. The same logic goes in academic study. If I just do some analysis proving that the minimum wage is harmful, how many journals are interested in publishing that? Unless I have a novel econometric technique, or some sort of more narrowly-focused area or research method, my study is not likely to capture much interest. The bottom line is that there is a bias towards the counter-intuitive in journal publications. Now that’s not necessarily bad, as we want any discipline to have its orthodoxy to be challenged, that’s how we grow. Yet we need to ask more detailed questions about these counterintuitive studies. For example, the most famous paper going against orthodoxy on minimum wage, Card-Krueger, has some important critiques against it, as captured in accessible summaries like this and this.

Further, I think there is something more fundamental. Empirical evidence must be coupled with sound theory, otherwise, “if you torture the data long enough, it will confess.” There is a massive problem with replication across many disciplines, and the most famous early critique of what’s going on is still valid (Ed Leamer’s paper “Let’s take the con out of econometrics,” even with improvements in study design. Modern statistical packages allow researchers to repeatedly run different data sets, and there is simply nothing to stop them from running thousands of sets until they find something that “works,” i.e., fits the counterintuitive narrative, as “we’re all p-hacking now.” Given the incentive structure of publish or perish, and the fact that counter-intuitive results will be more likely accepted for publication, and that modern statistical software makes it easy to hunt for something that could fit a story, are we really surprised that there are more studies showing not much negative effect? And in the case of minimum wage, there is almost no theoretical basis for thinking that there is no impact from a minimum wage hike. The only theory that could really have any hope is that of monopsony, and the conditions that would be necessary for it to have explanatory power are highly unrealistic (specifically, that every monopsony (single buyer of factor input, e.g., labor) is also a monopolist).

A final consideration is that even with a utilitarian moral framework (which I condemned), the net benefits compared to losses will increase the larger the minimum wage increase is. To raise the minimum wage by more than doubling it, and thinking you won’t cause more problems than a simple 10% increase is gravely mistaken. And even if you don’t see this in the data, you have likely already started the damage, as employers don’t wait for this–as the public pressure for higher minimum wages rises, employers anticipate it and respond by more automation. The response to $15/hr wages is already happening–count on it–and the losers in this will be the least skilled.

And how many of you think that Anonymous, when he looks at this linked report, will suddenly agree with me? I put the likelihood at less than 1%. @ anonymous: love to hear your response.