Bryan Caplan  

Why Do Unions Oppose Merit Pay?

History of Fire Protection... The Real Foreclosure Scandal...
Gary Becker off-handedly remarks:
Not surprisingly, teachers unions fight hardest against reforms that change the way teachers are paid, especially when they introduce incentives for teachers to perform more effectively.
I don't doubt that unions tend to oppose merit pay, but the reasons are unclear.  Profit-maximizing monopolists still suffer financially if they cut quality; the same should hold for unionized workers.  Why not simply jack average wages 15% above the competitive level, and leave relative wages unchanged?

Or to put the puzzle another way: Once you've secured a raise for all the workers in your union, why prevent employers from offering additional compensation for exceptionally good workers?

Comments and Sharing

COMMENTS (25 to date)
John Thacker writes:

My guess would be:

  • A belief that this wouldn't be additional compensation, but would come out of the other money.

  • A belief that encouraging good workers would raise expectations of what is expected of a typical worker. Perhaps there's a preference for somewhat easier work instead more pay. There was that recent study claiming that visibly talented co-workers are unpopular because they raise the bar for others.

Fernando BriseƱo writes:

The more productive workers' loyalties would gradually shift toward the firm and there will be less incentives to stay within the union.

Milton Recht writes:

Might be due to union paranoia.

Merit raises and bonuses remove control of wage increase (merit pay) from union to worker and company. A worker deserving a merit increase is likely to believe the union is not necessary, that his/her pay is based on individual skills and not union power or union negotiations.

Also, increases chance that worker that does not get a merit award will leave company, increases union member turnover and over time increases number of retained employees who deserve merit pay and believe they do not need union.

Milton Recht writes:

Also, worker seniority is a big part of union member benefits and merit pay not based on seniority.

Hyena writes:

Most unions are run by elected boards, thus their positions on big issues like this simply reflect the position of the median. It's probably pointless to identify a cause beyond that other than "the median union member expects to lose out in a merit pay deal".

ajb writes:

The more flexible the pay structure, the harder it is to monitor changes in payment as the industry changes. For example, an industry that might normally see average wage growth of 2% net of macro (overall economy) effects could choose to freeze most workers pay while growing that of the best teachers. Thus the static raise in salaries limits company control over working conditions. The more the firm can vary conditions the easier it is to undermine union agreements.

Look at how weak tenure is at the university level. Compared to European systems, the fact that the US has no consistent professorial pay structure means that it's easier to punish low performers. Heck, the fact that profs can be paid differently both across and within departments means that tenure only guarantees employment not compensation. There are few consistent salary policies even within state systems like CA. In addition, the non-tenure track teachers give schools even more hiring flexibility. Hence university tenure rules aren't as damaging to labor flexibility as those for public schools. Unions are well aware of this.

Matt writes:

The excuse I hear from teachers about this is that it gives teachers an incentive to teach to the test and sort of cheat the system by making the lessons going over test answers. My solution would be to keep merit pay based on student performance and fix the tests so they accurately test for what students should be learning. If the test is good, then it shouldn't matter if the teacher is teaching to it. That might actually be a good thing.

HispanicPundit writes:

Megan McArdle answered that question here.

Matthew Gunn writes:

Several possibilities that might rationalize teacher unions' behavior:

(1) If merit pay induces better teachers to enter the workforce then the current set of unionized teachers may lose over the long run. Under merit pay, good teachers win and lousy teachers lose. Lousy teachers are likely to have lower than average wage increases and less job security. While the average compensation of teachers may increase under merit pay, the current set of teachers may lose if they are sufficiently lousy on average.

(1b) The core political/voting constituency of the union may be the worst teachers because they gain the most from the union's existence.

(2) Merit pay may undermine the teacher union monopoly. As AJB describes, merit pay may undermine the seniority system and other agreements. Merit pay inherently requires some teacher level measure of merit, and if we know who the worst teachers are, there will be increased pressure to let them go.

FC writes:

It is an article of faith held by many teachers that all teachers are excellent. I have actually heard people say this.

Eelco Hoogendoorn writes:

Why would this monopolist care about quality one way or the other? The consumption of their good is rationed in such a way that people couldnt possible consume more of it anyway.

And merit pay, with average wages remaining equal, for the average employee means working harder for the same salary.

michael writes:

Two fairly obvious reasons:

1) Unions exist solely to conquer market forces and put decisions in the hands of workers. Merit pay cedes power.

2) Once unions were created the dumb, surly and lazy hordes descended on (that is, self selected into) heavily unionized trades. You think these people want to be judged on merit? They just want a job where arson or rape are the only grounds for dismissal.

Charles R. Williams writes:

Teachers do not believe that performance can be fairly and objectively measured. Everybody in education knows who the good teachers are and everybody knows who the terrible teachers are. But administrators cannot be trusted with their subjective judgments because administrators' incentives are not aligned with student welfare. Any pay for performance system will be gamed by teachers and administrators.

Teacher pay is an enormous percent of school budgets so a raise for one teacher efffectively comes at the expense of his colleagues. Unions have an overriding interest in maintaining teacher solidarity.

Teacher pay is back-end loaded because of generous defined benefit pension plans based on final average earnings. So while teacher effectiveness plateaus at 10 years experience, teacher salaries continue to rise with the acquisition of worthless credentials and long experience that adds little to teacher effectiveness.

All of this explains teacher compensation and it will not change until the government monopoly is broken.

Jonathan writes:

Looking at the issue from both sides there seems to be two concerns.

1. How do you assess merit in a profession where the "merit" is not entirely based on the quality of an individuals work but judged on the output of students?

2. Without merit pay teachers have no incentive to work beyond the scope of the guidelines (textbooks) and quantitative measures (test scores). In the same way that a salaried employee may have less motivation than a person who works on commission only.

Other problems involve variables beyond an individuals control, such as a teacher who is given a class of the least motivated students, that teacher will have to work more diligently to achieve what other teachers attain through a selection of advanced or highly motivated students.

There is no issue with compensation for those who do more of their own initiative. The problem is in the measurement of merit and also, given the structure of unions, the danger of merit pay easily becoming a patronage system.

Abalone writes:

Your question stems from a lack of differentiation between the union and the people represented by the union. There are three parties in every union-management relationship: workers, management, and the union. The union has its own interests, not always entirely consistent with the best interests of members. You cannot expect a union to do what is not in its interests or in its essential nature. Of the three parties, the union has by far the least interest in the success of the company. The union's main interest is its own strength.

Regarding the inherent nature of unions, first of all, it is not the union's job to maximize performance of the workforce. That is management's job. Management lost control of performance when the union was formed. A union will not cede any power if it can avoid doing so. Allowing management to make performance determinations is ceding power, particularly when those performance determinations affect pay.

Second, the whole basis of unionization is mistrust of management. Unions are all about the perception of fairness and getting credit for that fairness. Fairness in any scenario is a matter of individual opinion based on individual values and individual perspective. If you have 100 people in a system, you will have 100 ideas about what is fair. Unions are all about minimizing perceptions of workers being treated unfairly by management. The most salient element of that is having negotiated objective standards for determining pay raises. Nothing is more objective than seniority. The further you get from that, the greater the fairness mire. Sure, paying good workers and poor workers with the same seniority the same wage may be perceived by better workers as unfair but it is simple, objective, clear, and evenly applied so the perception of unfairness by workers is minimal and manageable, the price they pay for union protection. That seniority is an imperfect proxy for performance does not matter to the union. It matters to management, whose job is performance.

Regarding merit pay, that is a figment. I have designed two performance management systems for two organizations, one merit pay and the other pass/fail. There is no way to pay for performance without a lot of overhead and discord. Usually the amount of money available to reward merit isn't worth it. People in a system know in their guts who are the best and worst performers but systems require criteria and criteria end up being either subjective or poor proxies. Since the system needs to appear to be objective, it goes with the poor proxies and ends up measuring what is measurable, not what really matters. Is class performance on standardized tests the salient measure of a teacher? It might be a suitable criterion for a pass/fail system, a minimum criterion for retaining one's job, but does it differentiate the best teachers? Not hardly.

Unless there is an environment where there is enough trust to tolerate subjective judgments by management or everyone is so overpaid that the differences in rewards don't trigger an unfairness backlash from workers, a merit pay system is unfeasible. Were the work environment to meet that standard, a union would never have formed in the first place so the question would be moot.

Charlie Quidnunc writes:

According to the union leadership, they oppose merit pay because it divides the membership, weakening the union:

From a labor perspective merit pay would also divide the work force and in the long run lessen our ability to fight collectively to improve public education. If salaries were not simply based on years of experience and number of college credits earned or additional services provided, the teaching force at any workplace would be more stratified (differentiated) and much less willing to stand together during a conflict with school site management or during a contract struggle. The role of the union would be seriously compromised.

Can't have that, now, can we?

Daublin writes:

Merit pay is not a pretty picture from an incumbent's point of view. Merit pay means years of working hard and competing against your coworkers. It means that when you do something well, the other people at your school are going to be unhappy about it. It means that fresh young minds can actually oust you from your job.

Seniority pay, meanwhile, means all you have to do is punch your clock. Just keep showing up from 8-4 every day, and your pay is guaranteed. You can go home every night and forget about work completely.

Merit pay might be worth it if, when you did better, your school made more money, or you could go to a higher-paying school. However, that's not how it works.

Brian, perhaps it would be more instructive to consider the question the other way around. Why do the workers in many fields actually desire to have merit pay? What is different for teachers and factory workers, versus musicians and athletes? Or even within the teaching world, why does the 8-4 stand-up teacher want seniority pay, while tutors go for merit pay?

mark writes:

Amen, Charles R. Williams. Most of the replies assume quality can be and will be measured fairly, an assumption the union correctly doesn't share. How is merit pay supposed to work? Do I know what other teachers are making? We do have merit pay in college, football and basketball thrive on it but they don't worry about who is the best instructor or how possibly you could test for it. The coaches who make the most money attract the best student-ATHLETES. They may also be great instructors but they sure aren't interested in being evaluated on how they do with median talent which is an irrelevant question anyway. Merit pay to my mind is really about making it easier for administrators. They don't have to spend their time deciding who are the good teachers or what makes them that way. They just plug in a formula.

david (not henderson) writes:

1) What's the point of having a union if your pay is going to be determined by your value? Bad precedent.

2) Merit pay introduces different opportunity costs of strike action. Specifically, it increases the opportunity cost of strkes for the most able workers. Undermines the whole solidarity thing that is the basis for union power.

3) Reduces perceived reliance/dependence on the union on the part of the best workers.

quadrupole writes:


You are precisely correct about 'dividing' the membership. It is not at all uncommon for a union to represent quite heterogenous groups of workers, sometimes of radically differing value. By capturing a valuable subcontingent, and holding their wages hostage to increases in wages for the greater population of unionized workers.

Consider an example I've actually seen at a University. Back in the dark mysts of time, IT workers got put in the same category in the HR system as secretaries (which wasn't so stupid back in the days of operators). Inertia being what it is, nobody bothered to change it. Enter a union, who sought to unionize that entire class of workers. Which they succeeded in doing, since it looked like a darned good idea to the secretaries who outnumbered the IT guys.

Now the University really really really needs to IT guys to function. And IT guys are expensive and get more expensive pretty rapidly (especially in the late 1990s when I was watching this happen). So the University desperately needs to retain them. So the conversation went like this

University: We really really really want to pay our IT guys 15% more next year because we need to retain them.
Union: We forbid you to do so unless you raise the salaries of everybody we represent by that 15%.

Literally, you had management being blocked from increasing the pay of workers by the union, because the union was using one segment of the workers as a bargaining chip for the benefit of another, larger, subset of their heterogenous bargaining block.

I suspect something similar is happening with the teachers union. Teachers are not a homogenous labor block. Some teachers teach math and science and are in high demand. Some teachers are exceptional. But most teachers are neither exceptional nor teaching high demand subject. By forcing the schools to raise overall pay for all teachers in order to be able to bid for high demand teachers, the union benefits the vast majority of their users by raising their wages far above what they would otherwise be, at the expense of suppressing the wages of their minority of high demand users.

anon writes:

well explained by Gambetta and Origgi 2009 on preferences for mediocrity.

ad nauseum writes:

There are probably a lot of distorted views about fairness and equality that too many union members ascribe to.

Also, it is true that teacher performance might be hard to gauge, but there are upstart ideas to work with. Like this one here . It may not be perfect, but its a start.

Easy guess: Large wage discrepancies reduce alignment between union member interests. Unions most willing to accept them didn't survive this. Unions most opposed to them survived.

Roger Sweeny writes:

Once you've secured a raise for all the workers in your union, why prevent employers from offering additional compensation for exceptionally good workers?

Something sorta kinda vaguely like that does exist in most school systems in the United States. There is a bonus system based on the number of graduate education courses a teacher has taken.

The system is objective and predictable. If you put in the time and do whatever busywork is assigned, you will get the credits. Accumulate enough and you go into a higher pay stream.

Alas, research seems to indicate that teachers with more credits are no better than teachers with less. Moreover, if you ask a teacher how useful his or her graduate courses were, the median response will be, "I got a few good ideas, but mostly they were a waste of time."

However, the system does help keep together the union/ed school coalition. Going over to a performance-based merit system would mean a significant diminishment of ed school employment and prestige. Unions support the present system and education professors learnedly tell everyone why performance-based systems are impossible and/or unfair--while studiously avoiding helping to come up with ones that would be fair and effective.

Stiles writes:

Probably for two reasons.

First, an expectation that the incentive funds would be taken from the overall compensation package and reduce compensation for some of the union membership.

Second, an expectation that merit pay would not be effective in improving schools and might even reduce their effectiveness. Teachers who philosophically support individual incentives have concerns about how accurately and effectively they can be applied given the typical spans of control in schools. Span of control ratios rarely come up in discussions of merit pay in schools, but they should.

Comments for this entry have been closed
Return to top