The Empirical Literature on the Minimum Wage

by Russ Roberts on January 5, 2007

in Work

There have been dozens of articles in the academic literature on the impact of the minimum wage on employment, particularly on the employment of low-skilled workers. Virtually every one of these articles finds that an increase in the minimum wage reduces employment. As Don points out, this is just one of the negative impacts of the minimum wage. There are other effects that are harder to observe and measure—on the job training, the expected level of effort to be put forth on the job and so on.

Economists have done an immense amount of work trying to measure these effects and the overall impact on the poor. The overwhelming consensus has been that minimum wages serve the poor very poorly. The standard finding is that a 10% increase in the minimum wage reduces employment among low-skilled workers from 1% to 3%.

This consensus was challenged in 1993 in a series of papers by Card and Krueger. Using a very different methodology from previous research, they found virtually no effect on employment and some evidence that an increase in the minimum wage might increase employment among low-skilled workers. Card and Krueger’s work generated a critical response questioning the reliability of their findings.

I do not find the Card and Krueger findings compelling. Some do.

Even if you don’t, you can still think it a good idea to put 1% or 3% of low-skilled workers out of work in return for a 10% increase in the wage of those who keep their jobs. Personally, I find that to be a very unattractive trade-off, especially when you consider the non-employment effects, but that is a judgment call.

There have been many surveys of the literature on the impact of the minimum wage. The only one I could find on the web that is publicly available is this survey from the Joint Economic Committee. Scroll through the bibliography at the bottom. The author has an ax to grind (the author doesn’t like the minimum wage) but it will give you an idea of how much effort has been made to discover the actual impact of the minimum wage. Here is a pro-minimum wage survey of articles.

None of the articles are casually done. Most or all of them involve a complex set of statistical techniques to try and hold constant the other factors that affect employment. This is hard to do. Reasonable people disagree over when it is done well and when it is not. A pretty persuasive argument can be made that is nearly impossible. (If you have access to JSTOR, you can find John Kenna’s superb article reviewing of the difficulties.) 

A few more thoughts. Politicians like the minimum wage because the cost of financing it is paid by three groups—the workers (in the form of lower employment),  employers (in the form of lower profits) and consumers (in the form of higher prices). Missing from that list is taxpayers—so for politicians, if the negative effects are hidden from most voters, the minimum wage is close to a free lunch. So it is preferred by politicians to the earned income tax credit (EITC) which costs tax dollars. Most (all?) economists argue that the EITC is a much better way to help the poor. The other benefit of the minimum wage for politicians is that it makes low-skilled labor more expensive and boosts the demand for close substitutes, often union workers.

I find it strange that those who favor an increase in the minimum wage often are the same people who complain about outsourcing, or the moving of factories to low-wage countries or the greed of corporations such as Wal-Mart eager to squeeze every last penny out of their employees by paying only what the market will bear. Surely, such greedy and enterprising organizations will find a way to avoid the impact of the minimum wage by hiring fewer workers and finding other ways to reduce the cost of workers who are suddenly more expensive yet no more productive. I’m a big believer  in the profit motive. I again commend to you this article at Coyote Blog that shows how a real business responds to a sudden increase in the cost of labor.

Finally, let’s keep it civil here, folks. One of the things I like about this blog is the thoughtfulness of the comments on both sides of the aisle. Help us keep it that way by reducing ad hominem attacks on your opponent’s motives and IQ. 


Be Sociable, Share!



Add a Comment    Share Share    Print    Email


spencer January 5, 2007 at 11:24 am

I suggest you read this interview of Card at the Minneapolis Fed.

What he says is the simple model you use is not the relevent model. That labor economists almost across the board use a very different model to explain job search and employment.

If you want to discuss the impact of minimum wage I would like to see your expanation of why we should use your model rather then the job search model most economists use.

Randy January 5, 2007 at 11:31 am


I have yet to hear of a single case of an employer who increased hiring "because" the minimum wage increased. Have you?

spencer January 5, 2007 at 11:38 am

In the dynamic, flexible, very high trunover job market as it exist at the lower end of the US job market a 1% to 3% job loss from a higher minimum wage is insignificant and is less then the rounding error. The data says that in our dynamic economy the individuals displaced by a higher minimum wage have no trouble finding other employment. The difference in our assesment is you assume that the displaced employees become permantely unemployeed. But that is just not the case in the US economy. Do you have any evidence to show that the displaced workers never find alternative employment because the data just does not show that.

Also, you argue that employeers change the way people work after a rise in the minimum wage so that the remaining employees are worth their higher pay. That is generally called higher productivity and it is the way that we raise everyones productivity.

The only economist I have ever seen argue that higher productivity is a bad thing are those from George Mason when they discuss the minimum wage.

spencer January 5, 2007 at 11:40 am

Randy — if you expect to be taken seriously you have to do better then that comment.

Russ Roberts January 5, 2007 at 11:57 am


Consider two ways to increase productivity. One is to have higher skills. The other is to pass a law requiring workers to work harder–take fewer breaks, no leisure on the job (chatting with employees, surfing the web now and then) and so on. The first makes workers and employers better off. The second makes employers better off and workers worse off. It is that latter increase I'm against.

JohnDewey January 5, 2007 at 12:01 pm


I have never read anything from George Mason economists saying that higher productivity is bad. Where did you read that? or did you read something else they wrote and then add your interpretation to their writing?

Don Boudreaux January 5, 2007 at 12:01 pm

Spencer says above:

"Also, you argue that employeers change the way people work after a rise in the minimum wage so that the remaining employees are worth their higher pay. That is generally called higher productivity and it is the way that we raise everyones productivity.

The only economist I have ever seen argue that higher productivity is a bad thing are those from George Mason when they discuss the minimum wage."

Fifteen years ago I was offered a job at a law firm. The salary I was offered was $20,000 more per year than I was offered by the employer (Clemson University) whose job offer I accepted instead.

One reason I turned down the law-firm job is that I didn't like the work conditions — I didn't want to work as hard and under as much stress that young lawyers must work.

Economically, I would have been more productive working as a lawyer than as a college professor. But I knowingly and happily rejected the law-firm offer.

Would I have been made better off if government forced me to accept the law-firm's job offer? Would society have been better off?

spencer January 5, 2007 at 12:14 pm

John Dewey — Tyler Cowen at Marginal Revolution made a big point about employeers making people wark harder several months ago. He was even quoted at aysmmetrical Information.

Part of his thesis is that employeers would turn off the air conditioning, etc., to make up for the higher labor costs. In comments I pointed out that he was arguing that employeers would react to higher wages by doing things to reduce productivity and this did not sound very rational or economical. He responded by saying that was not what he meant.

Don? How does your law job have anything to do with the minimum wage discussion? I have never seen anything about forcing people to take jobs in any minimum wage discussion.
Why are you trying to change the discussion?

Do you have a response to the Card interview or am I safe in making the assumption that you agree with his discussion of job search models?

Randy January 5, 2007 at 12:16 pm


Just trying to eliminate irrelevant variables. I think it is clear from the lack of employers who increase hiring "because" of an increase in the minimum wage, and the presence of employers who do decrease hiring "because" of an increase in the minimum wage, that an increase does "cause" at least some unemployment.

So I guess I'm wondering why anyone would favor the minimum wage over the EITC, which would help all lower income workers, rather than helping some at the expense of others.

Morgan January 5, 2007 at 12:21 pm


I think you're right that most people won't become permanently unemployed as a result of a hike in the minimum wage. I have no doubt that some people earning the minimum wage are underemployed in the sense that they could be doing more valuable work. If a change in the minimum wage causes them to lose their current jobs, they are likely relatively quickly to find permanent, more productive employment that justifies the higher wage. To the extent that happens sooner than it otherwise would have, it might actually be a benefit accruing to a hike. These people are probably better off long-term because of the dislocation (depending on the length of their search), and the economy as a whole is better off.

I doubt, though, that *all* employees who lose their jobs because of an increase in the minimum wage will quickly find pemanent, more productive employment. More likely, they'll flow in and out of jobs in which they are marginally employable, which slows their accumulation of skills.

And there is always a next generation of low-skilled workers who also won't have any jobs offering wages that match their productivity. They, too, are likely to flow in and out of jobs for which they are marginally employable until they establish a sufficient skill base to become permanently employed. And again, their development is delayed.

To say that these people aren't harmed is wrong, so there is still a tradeoff, even if it isn't as stark as in Prof. Roberts's example.

JohnDewey January 5, 2007 at 12:31 pm

"Do you have a response to the Card interview or am I safe in making the assumption that you agree with his discussion of job search models?"

Why would you assume that a lack of immediate response to your demands indiciates someone agrees with you? Maybe someone doesn't respond to your demands because they have better things to do with their time.

Don Boudreaux January 5, 2007 at 12:41 pm


Suppose government raised the minimum-wage to a level that would have made me unattractive to Clemson University (and to any other similar employer). Having no job offer from a University, I might then have accepted the law-firm job.

You implied in your earlier comment that higher worker productivity is always good — the only economists you've ever encountered who argue against it are we folks at GMU. I disagree with you that GMU economists are the only ones who insightfully understand that ultimately what matters is satisfying people's preferences, and that higher productivity does not always promote that end.

JohnDewey January 5, 2007 at 1:27 pm

Minimum wage advocates cite the Card-Krueger study as evidence that minimum wage hikes do not reduce jobs. But weren't it's conclusions based on short term observations of a single industry?

Would such a focused study be applicable to all industries? Fast food restaurants cannot really be outsourced, so the study must not take into consideration global competition in the labor market. Also, automating fast food worker functions may be more difficult than automating other jobs, such as industrial or commercial cleaning.

Does a lack of job reduction in the short term really say anything about longer term impacts? A marginally profitable retail business may be forced to pay higher wages in order to continue the enterprise and meet contractual obligations. But at lease renewal time, the sharply higher labor expense definitely influences the decision to continue the business.

python January 5, 2007 at 2:11 pm


You say:
1) Some workers will be displaced due to higher minimum wages.
2) It is not that difficult for the displaced worker to find new employment elsewhere.

But the new job must also pay at or above the new higher minimum wage.
So, why can't workers easily find those higher paying jobs now? If they can do it now, why do we need legislation?

So I think I would need an explanation from you as to why current minimum wage employees can't find slightly higher paying jobs now, but they will be able to when minimum wages go up.

Jennifer January 5, 2007 at 8:41 pm

One thing struck me as I read this thread was when someone argued that 1 to 3% of workers will be displaced and another argued that those 1 to 3% will find other jobs. It is my understanding from the articles I've read that the researches conducted state that an overall 1-3% of workers will be displaced, period. The research took into consideration that a portion of displaced workers would find employment elsewhere. Perhaps 10% would be initially displaced but 7% would find employment. The point of the research is that that 1-3% of workers would be displaced.

I don't think there is any hidden or undiscovered numbers in this.

SheetWise January 6, 2007 at 12:21 pm

More good comments on George Will at "A Stitch in Haste"

Scott January 7, 2007 at 12:58 pm

Aren't labor unions collective bargining agreements often indexed to the minimum age in some way? Isn't political support for this just a "payoff/back" to the unions?

Doesn't a higher minimum wage for non-union workers also tend to make union labor more relatively "competitive" ?

Let's not kid ourselves that this is any sort of "helping the disadvantaged" legislation. It's cynically delivering the spoils to one's political constituency.

Nobrainer January 16, 2007 at 5:27 pm

Russ & Don,

Something to consider when discussing the minimum wage is the Baptists & Bootleggers theory. About the minimum wage above, "the cost of financing it is paid by three groups—the workers (in the form of lower employment), employers (in the form of lower profits) and consumers (in the form of higher prices).

As demonstrated in the Econtalk Podcast from 1/15/07, some employers — probably the bigger ones — may find an increase in profits due to the loss of competition from those unable to afford higher wages. Employers as a whole group — and even prospective employers — may lose, but there's the real chance for many to gain.

Michael January 17, 2007 at 11:38 pm

I didn't see any mention of the positive benefits of minimum wage hikes to the worker and to the economy. They are paid more, spend more, pay more taxes, have better nutrition, etc.,etc.

Ron January 20, 2007 at 1:15 am

The wealth from a minimum wage eventually ends up in increased land value and increased taxation just as a supply side tax cut. If you rent you get no benefit.

Previous post:

Next post: