David R. Henderson  

Boudreaux on Theorizing about Facts

Return to Dorms Bleg... School Networking Bleg...

"[M]uch of today's empirical work on the effects of minimum-wage legislation is in fact highly theoretical."
--Don Boudreaux

"The data never speak for themselves."
--W. Allen Wallis, the statistician and economist who was pals and colleagues with Milton Friedman, George Stigler, and George Shultz.

Don Boudreaux has an excellent post today on theorizing about the minimum wage. An economics professor had written him about his various posts on the harmful effects of the minimum wage. Boudreaux writes:

I opened an e-mail from a correspondent (an assistant professor of economics) who coincidentally expressed to me his wish that I would be more "humble in face of the facts about the minimum wage." This correspondent upbraided me for what he takes to be my stubborn, ideologically driven refusal to admit that "the facts prove" (his words) that minimum wages do not have the negative employment effects that standard economic theory predicts. Not surprisingly, this correspondent recommended that I read more carefully the works of Arindrajit Dube - a scholar whose research does indeed offer up evidence that, contrary to standard theory's prediction, minimum-wage legislation does not harm low-skilled workers.

Pretty devastating, right? After all, Dube has the facts to prove that the minimum wage does not harm low-skilled workers.

Well, not quite. Boudreaux writes:

These are just two samples among many possible ones that I could give here. Google the authors' names and find their papers related to the minimum wage. Read those papers. You'll see that they are largely devoted to theorizing about how best to empirically measure the consequences of minimum-wage legislation.

I am not criticizing the authors of these papers. Quite the contrary. I applaud them. Thinking straight about how to do good empirical work is both important and difficult. My point, instead, is that much of today's empirical work on the effects of minimum-wage legislation is in fact highly theoretical. For this reason alone (although other reasons can also be cited) it is simply naive to insist that empirical research alone can settle the matter of the effects of minimum-wage legislation. It is uninformed and, frankly, deeply unscientific to accuse those who use theory to question whatever position you take about the effects of minimum-wage legislation as being unwilling to look at the data or as elevating theory above 'the facts.'

Nicely said. The whole post, which is not long, is worth reading.

Comments and Sharing

COMMENTS (21 to date)
Unlearning writes:

Surely there's a distinction between pure theory and frameworks which analyse empirical evidence?

love actuary writes:

It takes a germane quotation to bring home the fact that 'data' is a plural noun. *bam*

Nick writes:

He paraphrases the email he was sent, and he does it in a way that strikes me as uncharitable. Of course I can't know this, not having read the actual email. But I doubt a professor of economics recommending Dube's research would say "the facts prove that the minimum wage has no negative employment effects," and if he did then he misspoke. The minimum wage literature might be divided into three camps, the Neumark and Wascher camp which claims to find disemployment, the Card and Krueger camp which claims to find the opposite effect or no effect, and the Dube camp, which sits between the two, and argues that the problem is a difficult one, and that if there are any effects, they are probably small and hard to detect.

The argument therefore is not that the 'facts prove' that there is no negative effect, but rather that if there is an effect, it is very hard to see. If you think minimum wages have large disemployment effects, Dube's research should result in a moderation of your views, and it should be a similar refutation for people who believe minimum wages will have a large effect in increasing employment.

What I suspect the email writer was responding to was the lack of this moderation that I read both in your posts on the minimum wage and in Don Boudreaux's posts. For example, take this post:


I think this post shows both an ignorance of the empirical evidence about the minimum wage, in that "little or no negative effect on employment" is a pretty good summary of the empirical research so far, and an ignorance of theoretical research about the minimum wage, which introduces plausible factors such as market power or equilibrium effects to explain why minimum wages might not have a large impact on employment.

I share your dislike of the minimum wage as a policy instrument. But even I have to concede that the evidence is just not there to support the standard supply/demand story about minimum wages reducing employment.

Don Boudreaux writes:

Nick: My understanding of the basic theoretical case against minimum-wage legislation is that it never predicts the magnitude of the disemployment effects of such legislation. It says only that, as a general rule, raising the minimum wage will reduce the quantity demanded of low-skilled workers. That effect might be monstrous; it might be middlin', it might be minuscule. Yet as a matter of policy, most economist opponents of the minimum wage (including myself) oppose such legislation not because we assume that lots of low-skilled workers will lose their jobs, but, rather, chiefly because we believe that some - perhaps an absolutely small number - will lose their jobs.

While I've read several economists express support for minimum-wage legislation on the grounds that the gains to the workers who earn higher wages outweigh (what they believe to be, and what might in fact be) the relatively small disemployment effects of such legislation, I can think of no economist who opposes the minimum wage only because he or she believes that the disemployment effects are large. That is, I can think of no economist opponent of the minimum wage who would favor it (or even not oppose it) if he or she were convinced that only a relatively or even absolutely small number of workers would be thrown out of work because of it. I believe that this claim is true for Milton Friedman, Thomas Sowell, Walter Williams, Armen Alchian, Jim Buchanan, Deirdre McCloskey, Jim Gwartney, Vernon Smith, Gary Becker, Pete Boettke, Russ Roberts, Dwight Lee and many other opponents of such interference with the pricing mechanism in the labor market.

Again, the foundational case against the minimum wage is not that its negative effects are large (by whatever standard) but, rather, that there are negative effects that fall mostly upon the very workers who are said to be the intended beneficiaries of such legislation.

Don Boudreaux writes:

Perhaps it might seem unduly academic or ideological or unpragmatic of me, in my previous comment, to say that I would oppose minimum-wage legislation even if I were convinced that only a small number of workers would lose their jobs while a much larger number of workers would enjoy higher hourly wages. (I overlook here negative consequences of minimum-wage legislation that are often overlooked even in the academic literature - negative effects such as worse job conditions or fewer opportunities for overtime work.) And perhaps I am guilty as charged. But here's my response to such a charge:


JLV writes:


Would any amount of evidence that there is no disemployment effect convince you or any of the people listed to not oppose minimum wages? Or is the assumption that there are negative effects a dogmatic prior?

(And am I right in assuming you have leximin preferences in loss space, but no preferences in gain space?)

JLV writes:

Sorry, not "no preferences", but rather indifference over any gains.

Don Boudreaux writes:

JVL: Well, you rather bias my choices by labeling one "a dogmatic prior."

My position is this: the law of demand is so foundational to economic science that any claim that a particular good or service is not subject to this law - or (say, because of monopsony power) does not observably appear to be subject to it - must be supported by (1) overwhelmingly convincing empirical evidence and (2) a sound theory for why that good or service is an exception to such a foundational rule.

The empirical evidence that the minimum wage does not reduce the job opportunities open to low-skilled workers isn't close to being overwhelming. A great deal of empirical studies have in the past, and continue today, to find evidence in support of the standard textbook prediction. (Again, that prediction isn't about the magnitude of the effect; it's about the direction of the effect.)

And the theoretical reasons offered for why low-wage labor is somehow exempt from the standard operation of the law of demand strike me as being very weak and ad hoc. Monopsony power - the granddaddy of such reasons - just doesn't work practically to explain the labor markets that I observe in America. Certainly I find highly implausible the claim that markets for low-skilled labor in the U.S. are chronically shot through with such levels of monopsony power to make a government-imposed minimum wage a wise economic policy. (If such monopsony power were in fact present today - and has been present for decades - why does it not attract other employers into the market in such numbers as to rid the market of this monopsony power? Entry barriers into such industries as retailing, maid service, and lawn-care services are not very high in the U.S.)

So because (1) the empirical evidence against the standard economic prediction isn't overwhelming or consensus-building, and (2) there is no solid theoretical explanation given for why the market for low-skilled workers is consistently exempt from the standard operation of the price mechanism, and - I'll add also - (3) the number of different margins that employers and employees have available to adjust to a higher minimum wage are many and often invisible to analysts (such as employers working their low-skilled workers harder), I believe that it would be wholly mistaken to reject the standard textbook, supply-and-demand analysis of the minimum-wage legislation.

ThomasH writes:
much of today's empirical work on the effects of minimum-wage legislation is in fact highly theoretical.
is nicely said but I don't understand the point of saying that it is "theoretical" if that is meant as over against "empirical."

It seems that the nub of the issue over minimum wages is how much additional income do those a) who remain employed earning the minimum wage earn in relation to the income lost by those who b1) are not employed because of the minimum wage, b2) who earn smaller profits and b3) who pay higher prices. Of course what kind of models one builds to estimate these effects is based on theory, but in that sense every empirical result is and ought to be "highly theoretical."

Don Boudreaux writes:

ThomasH: With respect, I don't think that what you identify as the nub of the issue is in fact the nub of the issue - at least not for most of us economists who oppose minimum-wage legislation. For me - and, I'm pretty sure, for the likes of Milton Friedman, Gary Becker, Thomas Sowell, Bruce Yandle, etc., - the nub in fact is this: Does minimum wage legislation reduce the job opportunities open to low-skilled workers? If so, then as a matter of economics it's an undesirable policy.

That is, while some economists can (and do) support the minimum wage because, on their cost-benefit calculation, the gains to the winners outweigh the losses to the losers, opponents of the policy object to government-imposed losses on the losers, regardless of the gains to the winners. (See my post above that contains a link.)

Stepping back, in the general public the debate about the minimum wage typically is conducted as if there are no negative benefits inflicted on the workers who are its intended beneficiaries. General-public supporters of the minimum wage, if they recognize tradeoffs at all, see those tradeoffs as between greater benefits to workers vs. greater benefits to employers. The negative employment effects are largely overlooked.

ThomasH writes:

I'm going to repeat my meta comment expressing surprise over the amount of space the minimum wage issue gets in this blog compared to many other instances of interference in the market with much greater effect.

Even if we restrict ourselves to labor markets what about financing SS and Medicare with payroll taxes rather than a progressive consumption tax? Surely that also harms some of the most disadvantaged people the law is designed to help.

Capt. J Parker writes:

"The data never speak for themselves."

And, if you the torture the data long enough they will confess to anything.

Jon Murphy writes:

If I may take a stab at JVL's question:

My biggest thing, even if the data were overwhelming or consensus-building, would be why the Law of Demand doesn't apply to labor markets in the same way it does to virtually every other market out there. What is it about the labor market that makes it so unique that it is the exception, rather than the rule?

That is the big thing I'd like to see before I am willing to overturn my professional views on minimum wage.

Don Boudreaux writes:

ThomasH: With respect, I believe that you incorrectly identify the nub of the issue. The nub of the issue as I see it is whether or not there are negative effects of minimum-wage legislation that are borne by workers who are meant to be helped by that legislation. The question is: Does such legislation causes the quantity demanded of low-skilled workers to fall - or, more generally, cause the employment opportunities available to low-skilled workers to shrink? If so, then the popular case for minimum-wage legislation loses much of its appeal.

The nub of the debate has never been, in my long-time engagement with this policy issue, whether or not the benefits to the workers who earn higher wages offset or not the losses to those who must endure greater costs (including, perhaps, forced unemployment) because of minimum-wage legislation.

Greg G writes:

I am not a fan of the minimum wage. I believe it is the market, not the minimum wage that puts the real effective floor beneath wages. I accept the idea that a legislated minimum wage will result in at least a slightly smaller number of jobs being available to low income workers.

But there is something very strange about this debate. All the agonizing about the suffering that the minimum wage causes to poor people seems to come from relatively wealthy people. The minimum wage is quite popular among poor people.

When will people start coming forward and self identifying as actually wanting to exercise their liberty to work for less than the minimum wage? Even the long term unemployed don't seem to do this. Quite an extraordinary amount of ink is spilled defending a freedom that so few really seem to want to exercise.

Levi Russell writes:

If you want to see the long-term effects of continual hikes in the minimum wage, look at the LFPR for the younger age groups.

Furthermore, as Bob Murphy indicated in his article last year, recent work shows that there's a significant negative effect on the growth rate of employment and that if this is the primary effect of the minimum wage, we might not see much effect on the level of employment in our models.

As always, Don has some interesting insight into this issue.

Nick writes:


Monopsony power - the granddaddy of such reasons - just doesn't work practically to explain the labor markets that I observe in America.

Is this a theoretical claim, or an empirical one? What do you think of the more common argument that low wage employers have market power not because they're monopsonists but because low wage workers can't easily search for other jobs because they're too poor to live off their savings?

Again, that prediction isn't about the magnitude of the effect; it's about the direction of the effect.

What do you think the elasticity of labor supply is? Labor demand? From personal experience I think neither is close to zero. Here is a summary of estimates:


Harold Cockerill writes:

I find it amusing there's this huge debate about the effects of small increases in the minimum wage when one of the original purposes of the minimum wage was to increase unemployment among southern blacks that were moving into the northeast in search of jobs. It was certainly clear to the unions pushing the minimum wage what the effect would be and that they were correct is born out by unemployment figures for blacks after the passing of minimum wage rules.

Nick writes:


I appreciate your thoughtful response. I can't figure out how to comment on your blog, so I'm doing it here.

If I understand you correctly, you are arguing that (a) half of minimum wage workers are under 25 years of age, and 20% are in their teens and so are especially likely to have a family supporting them, or to not have a family to support.

That leaves 50% who are over 25 years of age, and while it may seem likely that the remaining 50% have no family to support or a family to support them, you didn't cite any statistics on that. So I don't find that these statistics compelling, they tell me only that somewhere between 0% and 50% of minimum wage workers probably have an external source of support.

By the way, I wouldn't use the term 'desperately poor' as you did. The argument is not that these workers will starve to death if they leave their job, but rather that, at the margin, they are willing to trade off lower wages today in exchange for not having to risk some number of months of unemployment in the future, months which will be particularly unpleasant, perhaps because they have low savings or will have to live off their family. This may give firms market power as if they were monopsonies.

But I don't want to get bogged down in statistics -- as I said, I agree with you that the minimum wage is bad policy.

Where I disagree is in your distinctions between theoretical and empirical evidence. Here we have two competing theories of the labor market. On the one hand, we have the century-old model of perfect competition, on the other hand we have a model of search and matching or a model of monopsonies. In the model of perfect competition, minimum wages unambiguously decrease employment. In a search and matching model, the effect is ambiguous for the reasons described above (also see here). In a model of monopsonies, employment is increased.

You feel strongly that the model of perfect competition better captures our labor market, other people favor the other models.

But as far as I'm aware, nobody, not even your correspondent, is making the extreme claim that the facts speak for themselves independently of theory.

Rather, the claim is that the facts seem currently to favor the other models. The fact is that we haven't really been able to identify large effects of the minimum wage on employment, and while this does not rule out the model of perfect competition, it should lead us as rational thinkers to place more weight on the models of search and matching or monopsony.

But, I don't see any evidence of this in your blog posts. I'm not saying you should capitulate and embrace the minimum wage. I think more nuance would be nice though. Judging from the comments on your post, the people reading your blog do not really understand either the search and matching or monopsony argument, and if they don't understand why these alternate theories are good theories, then they are poorly equipped to argue that the minimum wage is a bad policy.

Floccina writes:

@Nick, if they are working so much that they do not have time to put in an occasional application then they are probably doing fine. When I was young I would often work 60 hours a week (sometimes more.)

One big problem that I see with min wage is that I see lots of people working for much less that minimum wage. In some of these direct sales outfits (like Vector Marketing http://www.cutco.com/company/careers.jsp) many people end up paying to work and bother their family and friends on top of it. The kids who try that might be better off working for $4 or $5 per hour.

Comments for this entry have been closed
Return to top