David R. Henderson  

The Role of Unions

PRINT
Robert Litan on Microeconomist... Predicting bubbles...

Every Labor Day in the last few years, we hear about the decline of unions and how that has been a bad thing. What I find striking is how economically uninformed most of this commentary is.

Start with the fact that what gave unions their big push in the 1930s was federal legislation allowing them to be the sole bargainer for employees, even for employees who had no wish to join or pay dues. What do we call an organization that is the sole seller? We call it a monopoly. And not the kind of monopoly that some people say Microsoft is or had been. Microsoft always has to compete with other software companies. No. Unions are the kind of monopoly that George Stigler wrote about in "Monopoly" in The Concise Encyclopedia of Economics. Stigler wrote:

Even today, most important enduring monopolies or near monopolies in the United States rest on government policies.

Interestingly, even two of the most prominent modern economist/defenders of unions admit that fact. In the article Labor Unions in The Concise Encyclopedia of Economics, Morgan Reynolds, formerly chief economist with the U.S. Department of Labor, writes:
According to Harvard economists Richard Freeman and James Medoff, who look favorably on unions, "Most, if not all, unions have monopoly power, which they can use to raise wages above competitive levels"

Unions, moreover, have a pretty ugly track record on race relations, which is why two prominent early 20th century black leaders, W.E.B. DuBois and Booker T. Washington, who agreed on little else, agreed that unions were bad for black workers. When people forcibly prevent you from competing and figure out ways to exclude you from working, you don't feel very good about them. I remember attending an antiwar rally in San Francisco in 1980 (when some people were worried that the U.S. government would go to war against the Soviets in Afghanistan--little did we know that the U.S. government was already at war against the Soviets in Afghanistan) and hearing some young black women sing a song in which some of the words were: "The union man, he decide whether I live or die." Incidentally, as Reynolds points out in Labor Unions, one of the economists who did the most to document the early negative effect of unions on black workers was President Carter's chief economist of the Labor Department, Ray Marshall.

You'll often hear people complain about unions today and say words to the effect: "At one point in our early history we needed them but today they're no longer needed." But this golden age of unions is a myth. That doesn't mean that we don't need them--possibly some people do. But what we don't need is government support of union monopolies.

By the way, over the last 40 years or so, unionization has changed dramatically. As Reynolds writes:

In the United States, union membership in the private sector peaked at 17 million in 1970 and had fallen by nearly half--to 8.8 million--by 2002. Barring new legislation, such as a congressional proposal to ban the hiring of nonunion replacement workers, private-sector membership will likely fall from 8.5 percent to 5-6 percent by 2010, no higher than the percentage a hundred years ago. While the unionization rate in government jobs may decline slightly from 37.5 percent, public-sector unions are on schedule to claim an absolute majority of union members within the next few years, thereby transforming a historically private-sector labor movement into a primarily government one. Asked in the 1920s what organized labor wanted, union leader Samuel Gompers allegedly answered, "More." Today's union leader would probably answer, "More government." That answer further exposes the deep, permanent conflict between union members and workers in general that inevitably arises when union-represented employees are paid monopoly prices for their services.

So our big challenge with unions nowadays is to rein in unions of government workers who are negotiating high wages and high pensions. That is what is wrecking state and local government budgets all over the United States.


Comments and Sharing


CATEGORIES: Labor Market , Regulation



COMMENTS (24 to date)
LD Bottorff writes:

Very good, but I think another important question to ask is, do unions really bargain for what is best for their members? In some cases, they do. But we know from Public Choice theory that just because a bureaucracy is supposed to promote a specific public policy does not mean that the incentives are in place to actually make that happen. Since unions are typically monopolies, just how well can we expect them to server their customers? I know that sometimes monopolies provide good service...and oftentimes they don't.

JKB writes:

Sure they're negotiating high wages and high pensions, but as long as they have the police and the court house employees, they can come and take the high taxes needed to pay those things. Monopoly on labor is one thing, but the monopoly on violence is where the money is.

mike shupp writes:

Bad unions! Evil monopolies! Got that.

Let's imagine a beautiful alternative, with no unions, no monopolies, and individuals acting as autonomous individuals under a rule of law. Suppose, for example, you're a big city school system and you employ say 10,000 teachers.

Of course, each of those teachers has a contract with you. Every year, the teachers come in, each with his or her own lawyer. and negotiate a new contract or a continuation of their contract. Their INDIVIDUAL contracts. Thus teacher A will inform you that she'll accept your wage offer but only if no more than 27 students are permitted to enroll in her classes. Teacher B will counter your offer by insisting on a higher contribution to his pension. Teacher C will take a 500 dollar per year pay cut if she is exempted from having a home room period in her work day. Teacher D wants Fridays off to tend to her ill child. Teacher E wants Fridays and Mondays off, and maybe wants to duck out of the second semester because he's trying to get a rock band running; also he wants a more generous termination scheme. Teacher F wants transferred to a school where he won't meet Teacher G. Teacher H wants ...

10,000 teachers, each with individual demands, each with clever meticulous lawyers, demanding very different contracts, all of which will be enforced to the letter by courts distinguished by their zealous enforcement of every iota of such contracts. And then there are the groundskeepers and janitors, the cafeteria workers, the various school principals, the teachers' aides, the school district's own administrative staff -- somebody has to keep track of these different contracts and monitor things, right? -- school nurses and guidance counselors and so on. There might actually be 20 to 30 thousand labor contracts to deal with. Every single year.

The beautiful part of such arrangements is that each teacher, groundskeeper, etc. would be responsible for performance. schedules, etc. that they had personally chosen as desirable. They'd remember each of their obligations to you, and be willing to adhere to such tasks, and they'd be sure to observe your performance of your duties to them. At no cost! -- it'd be something they just naturally did. Surely this would be a libertarian utopia!

Granted, it would cost the school board something to monitor each of those contracts. And there might be expensive penalties if those contracts got accidentally or purposefully violated. But that's how contracts are supposed to work. It would be naive to suggest that just maybe the school board would prefer to impose one schedule on all its employees, one schedule of salaries on all its teachers, one set of behavior standards. And it would be churlish to argue that in return for such convenience a school board might be motivated to offer somewhat better terms to a group of employees, to be accepted or rejected en masse, rather than pay the costs of negotiating individual contracts.

No, let's support liberty and law and the power of markets. No more ugly unions or morose, malevolent monopolies! Let's celebrate Labor Day properly by embracing perfection. What can libertarians do to create such ideal workplaces?

David R. Henderson writes:

@mike shupp,
I don’t know if you know the history of teachers’ unions. Your comment above causes me to suspect that you don’t. If I recall correctly, they became common only in the 1960s. Yet even large school districts managed before then to avoid the complexity of contracts that you fear. My father was a teacher in government schools from the early 1930s to 1971. Only in about the last 10 years was he in a union. I don’t think he ever consulted a lawyer, except to make a will, in his life.
Also, the vast majority of private schools, even large ones, are not unionized. Somehow that works out too without a lot of lawyers.

Mike Hammock writes:

David, there's also the fact that lots of other industries require uniformity in hours, standards, expectations, et cetera, and yet get by without labor unions. The same goes for private schools, which somehow are able to have similar contracts across all their (mostly non-unionized) employees. Mike Shupp's argument proves too much.

Massimo writes:

Most union supporters that I know of are K-12 education employees who want better employment terms for themselves. They aren't the types to think of this type of policy issue at an abstract level and consider what is best and most fair for society as a whole.

The types who read this blog like to daydream as philosopher kings about the most fair and elegant rules for society, and these criticisms of unions are old news.

mike shupp writes:

David Henderson:

In fact, you're wrong. I do know a bit about teachers' unions. My grandfather was a teacher and school superintendent up till the 1960's; my father was a teacher and school administrator into the 1980's. As a superintendent he did negotiate with individual teachers and I can recall various occasions on which he reflected (generally favorably) on their performance, their compensation, etc. Dad was an NEA member (a "professional association!) from the 1950's onward; he despised teacher unions in general and the AFT in particular, long after most of the world had concluded the NEA was effectively a union.

So. You might consider my rant is a bit more rhetorical than truly prescriptive. Conservatives and libertarians are apt to speak of a better world in which individuals and employers deal as equals. I presented a picture of such an example, and suggest that there would be costs associated with such a style of employment. In a world in which would-be employees were rare and precious, employers might find such costs onerous, and the mechanics of finding and satisfying employees could be cumbersome; I suspect one could make Coase-ian style arguments for dealing with employees through collective bargains of some sort with the idea of improving firm efficiency.

Of course, in the real world, there are almost always far more would-be employees than positions for actual employees, and employers generally find themselves in the position of dealing with employees on a "like it or leave it" basis. Thus, as you and others point out, many organizations get along quite happily without dealing with unions. But this is clearly NOT a condition in which employers and employees are dealing in complete equality. And of course, this is precisely the sort of thing that unions were devised to deal with.

Happy Labor Day!

vikingvista writes:

mike shupp,

Companies with tens or hundreds of thousands of employees don't have such trouble with it that they go looking to establish labor unions to improve efficiency, and they don't spend inordinate resources on court expenses regarding employment contracts.

Obviously not having a 3rd party establish uniform contracts is a complete non-issue, and always has been. It also has nothing to do with why labor unions were created or why they exist today.

Craig J. Bolton writes:

"Start with the fact that what gave unions their big push in the 1930s was federal legislation allowing them to be the sole bargainer for employees, even for employees who had no wish to join or pay dues. What we do call an organization that is the sole seller? We call it a monopoly."

Sounds like the way the medical industry has been organized since the 1880s.

vikingvista writes:

"Sounds like the way the medical industry has been organized since the 1880s."

To what are you referring?

Mike White writes:

A police force that yields unionized, monopoly power. True, scary, and maybe the root cause of the disregard for the rule of law (or legislation...not relevant to my point) in our society.

mike davis writes:

The really interesting question in all of this is simple to state: In a world without labor law (and antitrust laws) would there be labor unions?

For what it’s worth, here is the conjecture of someone who’s main qualifications are a full cup of coffee and a desire to avoid answering emails (aka, me).

Maybe. There are serious transactions costs that go into hiring and organizing a workforce. These costs are faced by both the buyers and sellers of labor. Labor unions might help reduce these transactions costs and so be of benefit to workers and management. (See some performing arts organizations for examples.)
Maybe. In some markets monopolizing the supply of labor could result in higher wages. However, if a labor union is to do this, it has to overcome (a) free rider problems (workers won’t want to pay dues) and (b) the problem of cheating on the cartel (other workers left unemployed by the higher wages will offer to work for less). Groups of people solve these kinds of problems all the time, sometimes with violence but usually in a peaceful way.
Maybe. Producer cartels will have an easier time agreeing to a price and preventing cheating if they all face the same input costs. Signing an industry CBA helps a lot. (See the U.S auto and mining industry in the 1950’s and 1960’s.)

This leads to the following (slight) criticism of David (and Morgan Reynolds and lots of other economists who write on this): There is a difference between a monopoly on labor and a monopoly on the bargaining process. U.S. labor law works to give labor unions a monopoly on the bargaining process. Of course the hope is that doing so will give unions a monopoly on labor, but they’re different things. The principled objection (ok, at least according to my principles) is to the government intervention in the market for agency, not the monopoly.

One last conjecture: I suspect labor law has had much less influence on labor unions in the U.S. than many people—progressives, conservatives and libertarians—would have you believe. But that’s for another day…. the coffee is nearly gone and the emails await.

Handle writes:

"Unions, moreover, have a pretty ugly track record on race relations..."

I think you're being a bit unfair and essentializing much more than is justifiable. It sounds a bit like an unwarranted smear. Are you trying to claim that there is something about labor unions that makes them especially likely to enact racist policies more than any other type or organization? It seems like you are making a claim that would warn us about the nature and attributes of unions today and tomorrow based on some incidents in the past, but it seems to me that claim is totally unsupported.

Unions are just one type of group association. Think how absurd it sounds to say the same sentence if you replace it with 'governments' or 'universities', which you could easily do with plenty of historical examples.

There were a few unions, in a few places, and a long time ago, which embraced policies that had a disparate impact on ethnic minorities. But most labor organizations around the globe and for most of history are not guilty of that charge.

triclops writes:

Handle,
Since some unions were created with racist objectives in mind, it's entirely fair. Maybe you could quibble that it should say "some" or "many" unions, but that's about it. He isn't saying that there is something essentially racist about unions, but there certainly was something essentially racist in unions that sought to prevent Blacks from plying certain trades.
It's no different than pointing out that many anti drug laws were enacted, in large part, over racist panics.

Charley Hooper writes:

@mike shupp

You create a scenario with thousands of individual employee contracts and say, "See, that can't possibly work."

But I've worked at three large organizations with thousands of employees each, none of whom were unionized. And these successful organizations certainly didn't buckle under the weight of this organizational or legal challenge.

So how do you explain that?

mike shupp writes:

Charley Hooper:

I spent 45 years, generally content, working at different places without being in a union (and a couple years at jobs with ineffectual company unions). Do you really think I was so stupid that I never noticed the fact?

Now. There's an oft heard conservative argument that unions are undesirable because they hinder equal dealings with companies and their employees. The point I mas making is that, (a) as a rule, employees and employers do not negotiate employment and working conditions as equals, (b) that unions are a device to improve the relative power of employees, and (c) from a company viewpoint, dealing with unions might be more efficient in some ways than dealing with employees as individuals.

This isn't rocket science, and I find it fascinating that no one who has responded here to my comment has been able to fathom it. I think you guys are reacting emotionally to a pro-union argument rather than logically.

Charley Hooper writes:

@mike shupp,

Your attacks are as unwarranted as they are unwelcome.

As to your first point, I have been both an employee and an employer. I agree that employees and employers do not negotiate as equals; each has advantages and disadvantages, strengths and weaknesses. If you look at most capitalistic acts among consensual adults, you'll find the same pattern.

I agree with your second point.

Regarding your third point, I can't imagine an employer ever wanting to deal with a union. I see huge downsides and either no or negligible upsides.

vikingvista writes:

mike shupp,

"This isn't rocket science"

Exactly. Which makes your perseveration with this efficiency argument hard to fathom. Let's step through it again:

1. If uniform contracts are a partially *mitigating* benefit to the cost of unions to employers, then the lack of employers seeking out unions clearly shows that the union costs to them still outweigh the benefits.

2. If uniform contracts *by themselves* are beneficial to employers, then you should see employers seeking out NON-union 3rd party contract negotiators in order to gain that benefit without the costs of unions.

3. If uniform contracts *by themselves* are beneficial to employers, they CAN easily establish them without unions. Therefore, unions offer NOTHING to employers in this regard that employers did not already have.

The arguments against your post are not general anti-union arguments. They are arguments against a specific and utterly hollow point that you made.


"as a rule, employees and employers do not negotiate employment and working conditions as equals,"

Sometimes an employee pushes an employer to accept the highest wage the employer can afford, and sometimes an employer pushes an employee to accept the lowest wage the employee can afford, and everything in between. But both parties are equal in their authority to veto the employee-employer relationship.

"that unions are a device to improve the relative power of employees"

...over the unemployed. Yes. That is what it means to prevent the unemployed from competing with the employed on wages, i.e. establishing above market clearing wages for members. Unions are generally not advocates for labor. They are at most advocates for their members against their nonmembers.

Steve writes:

"What we do call an organization that is the sole seller?"
Needs proofing.

David R. Henderson writes:

@Steve,
Thanks. Correction made.

David R. Henderson writes:

@mike shupp,
This isn't rocket science, and I find it fascinating that no one who has responded here to my comment has been able to fathom it. I think you guys are reacting emotionally to a pro-union argument rather than logically.
Actually, mike, Charley Hooper, Mike Hammock, vikingvista, and I were able to fathom it. You might want to check your original argument. It was that if a large school district were to face a work force that is not unionized, there would be a huge need for lawyers to negotiate contracts. That’s the argument you made and that’s the argument we responded to.
Now, you’re free, of course, to change your argument. But then for a reasonable discussion to ensue, you need to give some indication that you’re aware that you have changed your argument.

mike shupp writes:

David Henderson, et al:

Nah. The thing is, the notion that in the real world would-be employees and employers deal with each other as equals is hooey. Maybe a company has a need for a super-duper patent attorney or a chemist who's an expert on formulating aspirin or something and individuals who fit the bill can write their own ticket. But as a rule, companies have rather general requirements which can be filled by many potential employees. At the same time, there are usually more applicants for work than jobs, even in very good times for workers. So companies get to be more selective than workers and they don't deal as equals. That strikes me as straightforward.

There's another issue which I ignored, which is that companies can't actually be flexible negotiators in all circumstances. If you've got an assembly line which can be manned by 8000 employees, let's say, you really want 8000 people on the line whenever you're operating, not 7000 and not 9000, and that's not a negotiable point. I.e., a company really can't be a totally flexible negotiator in dealing with its workers and would-be employees have to accept that. (I don't see this as an evil thing, by the way. It's just something that is, like the law of gravity.)

Let me mumble to myself a few minutes ... I think what was obvious to me and what I failed to communicate, the thing that drives me up the fornicating wall, was not that "unions are a good thing" but that a common conservative/libertarian meme -- "unions prevent employers and their employees from dealing as equals" -- is complete BS because it ain't possible. Is that now clear?

As a whole 'nother issue, I advance the suggestion that in some circumstances companies might find it economically advantageous to deal with unions rather than dicker with employees as individuals. I don't insist on this; I just offer it as something a microeconomist might find worthy of a paper or two.

And as yet another issue, it's amusing that you and I and the rest have cleverly avoided mentioning the real reason historically for companies to deal with unions rather than individuals -- to ward off wildcat strikes.

Oh well. It's been an interesting discussion, and I wish my bests to you, to Charley Hooper, and to Vikingvista. And now the holiday is over and I guess we should all Get Back To Work.

Charley Hooper writes:

@mike shupp,

If I may, here's what I understand of your argument.

Employees and employers don't deal as equals right now. Therefore the argument that unions prevent employees and employers from dealing as equals is nonsense. In other words, you can't blame unions for preventing something that wouldn't happening anyway.

And because this perceived downside of unions is a myth, there might be some reasons for employers to choose to deal with unions.

Let me address your first point.

Because there are normally more applicants than positions, employers usually have a choice of who to hire. That's true. But employees also have a choice of which jobs to apply for. An employer might have 20 applicants for three positions. But those same applicants can choose to apply at 50 different companies. So they both have choices.

Are they equals? They both want to strike a deal with the right party.

vikingvista writes:

"As a whole 'nother issue, I advance the suggestion that in some circumstances companies might find it economically advantageous to deal with unions rather than dicker with employees as individuals. I don't insist on this; I just offer it as something a microeconomist might find worthy of a paper or two."


It is encouraging that you are now backing off of this wholly untenable argument even if only half way. You certainly deserve credit for that.

Comments for this entry have been closed
Return to top