Merit Pay for Teachers

I talked to a Spanish teacher in a suburban Boston public high school last night. She expressed her disapproval of the idea of merit pay for teachers. “Some of my students hate their parents and some have been sexually abused,” she noted, “but my salary is supposed to be based on their performance?”

I responded that I thought the deeper question was how school administrators would evaluate merit. I asked “What incentive does a school administrator have to do a good job evaluating merit, or indeed, to do any work at all?” The teacher said “none, especially if they are a member of a protected minority group. They can’t be fired, no matter how little they do.” That made sense to me; the school is guaranteed to get nearly all of the students in the town regardless of whether or not the administrators do anything (few parents can afford to pay property taxes that fund public schools and then pay private school tuition as well). Wasn’t a teacher also guaranteed customers even if she didn’t work? “You have to try to sell the students at the beginning of the year that the class will be fun. Otherwise it is just unpleasant to sit with them for the remaining 8 months. If I’m incompetent and lazy, I will still get the same paycheck and the same number of students, but it will be tedious to share a classroom with unhappy students. That’s my incentive to work.”

Reflecting on this conversation, I was surprised that anyone thinks merit pay will work. Restaurants aren’t very important to our society or our future. Great empires have been built by countries with bad restaurants. Yet nobody would propose having restaurant compensation be determined by a government bureaucracy assigning “merit” to each restaurant. We allow citizens to choose which restaurant to visit and eventually the bad restaurants wither away and disappear due to lack of customers. Short of something like that in public education, how would we ever expect quality to improve? Wouldn’t parents and kids, simply by talking amongst themselves, quickly figure out who were the effective teachers and try to crowd into their classrooms, abandoning the ineffective teachers. An administrator looking for “merit” would simply need to count heads in the classroom and/or pay a teacher according to the number of kids who signed up (this is how education has worked for most of human history, actually; our current bureaucratized and tenured system is a relatively recent innovation; current “merit pay”).

I circled back later to the Spanish teacher and asked her if parents and kids knew who the good teachers were. “Of course. There is a huge amount of pushing by parents and kids to get into the best teachers’ classes. But at the end of the day the worst teachers still get a more or less full classroom and a full paycheck as well.”

[The complex merit pay schemes so far don’t seem to have worked, e.g., see this story on Nashville schools. Why isn’t anyone anywhere seriously considering the “parents/kids get to choose their school and teacher” approach? I don’t think it is fair to say that charter schools represent full choice because they don’t get anywhere near the funding of public schools (see this study on how the cost of public schools, if normal accounting measures are used, is much higher than quoted). Anyway, charter schools are hugely complex to set up. Assuming a normal distribution of effectiveness, the public schools already contain millions of above-average teachers. Without building any new buildings or setting up any new infrastructure, we could presumably improve outcomes simply by allowing those above-average teachers to attract more students.]

24 thoughts on “Merit Pay for Teachers

  1. Well, yeah. Any attempt to graft contrived incentives onto a broken organization is doomed to failure. People will always find ways to “game the system”. This stuff is proverbial in the software world: All known attempts at “objectively” measuring programmer productivity have ended up incentivizing counterproductive nonsense. If you count the bugs they fix, they check in dozens of easily-fixed bugs. If you count lines of code, they throw in a lot of extra newlines. If you count comments, they put in lots of long comments that don’t say anything. The list goes on and on. Old news.

    There’s no formula for fixing broken organizations, nor any formula for building new ones that aren’t broken. But there is an absolutely foolproof formula for identifying organizations which just *happen* for whatever reason (who cares?) not to be broken, and rerouting most of the resources to them. And it’s dirt-cheap and dead-simple to implement, too.

  2. The key line seems to be this:
    There is a huge amount of pushing by parents and kids to get into the best teachers’ classes.

    Perhaps that would be one of the more effective tests for merit pay: award pay based on the # of families that want into the class.

  3. Have you seen, “Waiting for Superman”?

    Having watched that movie, I don’t think we’d ever see merit pay implemented on a broad scale anyway, due to the teacher’s unions. My conclusion, having watched the movie, is that primary education in the US is hopeless, as long as teacher’s unions exist.

  4. Matt: I haven’t seen the movie. The teacher I was talking to was, I think, a union member, and she seemed dedicated to her job (though as she was not a native Spanish speaker, I suspect someone more qualified could be brought in from Latin America at a tiny fraction of the cost). So I think we could have quality education in the U.S. + unions. We would simply have to pay $100,000/year to ineffective teachers who had just two or three students (maybe their parents were recent arrivals to the town and didn’t realize that the teacher wasn’t competent) or maybe even zero students. The union can force the taxpayers to pay incompetent and lazy workers trillions of dollars, but they can’t force the school, I don’t think, to give incompetent and lazy teachers a full class of 20+ students.

  5. Sweden implemented a voucher system in the early 90s and it seems to have broad support today: At the micro-level, lecturers (Privatdozent) in Germany—including Immanuel Kant—historically were paid by individual lecture attendance: One could object that students might choose the easiest lecturer at the elementary level, but that is more a misalignment of incentives than a critique of the pay-by-lecture or voucher model: if you have everyone take a cumulative test at the end of the year (and link advancement to this test), then the students will choose the most effective lecturer.

  6. This one finally made me stop lurking on your blog and post, because charter schools do work, and people are working on making them happen in a much bigger way.

    We know they work because of cases like New Orleans after Katrina, where the only schools around were charter schools. There are still public schools here but nobody wants to go to them, and they are *vastly* improved for the competition.

    But, much cooler is that people have figured that out, and are applying private sector methodologies to getting charter schools off the ground. You say they are complex to set up, and they are. 4.0 schools is a non-profit in New Orleans which is setting up a program to help people start charter schools, giving them access to mentors, funding, tools, software, and a community to talk about these things.

    It’s still too complicated. There’s no reason for a dedicated person to need 4 years to start a school.

  7. I’ve been a public school teacher for almost a decade. One of my earlier students now works at a charter school here in Miami. The starting pay is $27k and they require the teachers to tutor after school.

    He calculated his pay to be $11 an hour. That’s teaching high school.

    Charter schools sound great if you don’t have to work in them.

  8. Don’t you think that the better teachers already have as many students as they can effectively teach? It would be nice to shrink the number of students in the ineffective teachers’ classrooms, and it might even enable some of them to teach well, but I don’t see any evidence that the good teachers can handle even more students.

  9. David: Thanks for the data. Tell your friend that he’s earning more than a lot of helicopter instructors! (including this one)

    Lawrence: Probably the best teachers could handle a few more, but you’re right that we would probably have to hire additional teachers if we want to allow the ineffective ones to have 0-3 students. Basically we would have to accept that government jobs serve as a welfare system for the incompetent and lazy.

  10. I don’t know if parents would make the best choices selecting teachers. In medicine and learning often the users are not the best judges of quality.
    Teaching has two main components: knowledge and ability to teach. The first can be easily measure with tests, the second may be more difficult to ascertain because I believe it is an art. What we can do as a society is to insure that Spanish teachers do speak Spanish and math teachers do know how to solve a quadratic equation.
    I have some experience with NYC public schools, and I can assure you that some of the teachers my son had before this year (he is in seventh grade in the Hunter College High School) could not have passed a simple knowledge test on the subjects they taught. Why aren’t teachers better screened? That way we would have to deal only with those who “do not know how to teach.”

  11. “Some of my students hate their parents and some have been sexually abused,” she noted, “but my salary is supposed to be based on their performance?”

    Actually, yes I think it should be. As you said, it depends on how one defines performance. And I love how the only incentive the she seems to think applies is her own desire not to be bored.

    In San Francisco one isn’t bound purely to geography for school assignment, so we spent the winter touring multiple schools and speaking with principals and teachers. These were largely K-5 schools. A contrast: at two of the schools the principals were clearly hard workers and performers, having brought their schools from being on the closure lists to having active populations, vibrant Programs, and very active PTAs bringing in enough additional funds to hire extra teaching staff. However, one really stood out as awful — her great achievement was figuring out how to make sure the kids lunchroom use was staggered by grade. She was only at the school because the previous low-performance school she was a principal at was changed to all Spanish and she didn’t have the language skills.

    My mother is retired from one of the rougher school districts in California (West Contra Costa, formerly Ruchmond). She relates stories of teachers she used to work with in the district who had been in the district during WW2. They used to work teaching all day, then work swing-shift at the Richmond yards welding Victory ships, which paid far,far better than teaching did. Clearly their motivation to teach was something more than either money or a desire not to be bored. I have a friend who teaches only emotionally disturbed kids in a criminal justice diversion program. Clearly she does it for more than the money (has to as she’s being screwed around on those idiotic “teaching credential requirements” when she used to teach graduate students in University).

    For anyone I recommend a recent talk at Cato called “Cloning Superman”. One of the panelists was a key player behind the Swedish voucher movement. Well worth an hours listen.

  12. I’ve been lurking a while as well…but this one got me riled.

    What causes success of industries? When the bad participants can fail, and be driven out of business by more successful competitors, and when new entrants can adopt new practices and try them (and usually fail).

    If you don’t have a situation where the bad teachers and the schools which adopt bad approaches lose funding, and the ones that deliver better education for cheaper get more customers…you don’t have a path to getting better.

    Very simply, there is a conflict between the desire of the public to have schools that are getting better…and the desire of the teachers to have a safe, stable teaching job. Pick ONLY one.

  13. I’m not an expert on this, but from what I’ve read Weighted Student Funding (WSF) seems to be a very promising model for public school reform. The concept behind WSF is that per-student funding follows the student. Under this concept, the school district ties its funding for each student with the school the student chooses to attend, (so when a student moves between schools his or her funding moves with him or her). This in effect enables students and families to “vote with their feet” and thus draws educational resources to the more effective schools.

    A WSF approach can be applied within public school systems. It doesn’t require charter schools to be created. (see for example ).

    In a WSF system, if students begin to move from School A to School B, their funding moves with them from School A to School B. So if School B begins losing students, it also begins losing funding. Thus School B has an immediate incentive to improve while funding resources are being drawn to more effective educational institutions.

    One of the things I like about WSF is that it seems to have the potential to use our inherent human irrationality to allow better schools to spontaneously emerge. In theory this would allow for good schools to emerge within systems as more or less “self fulfilling prophecies” of evolving communities perceptions.

    For example, let’s say there were a public school system of 12 more or less similar and mediocre schools. Parents would wonder which of the 12 schools they would most like their kids to attend, and would begin to look for cues to distinguish which were better. Perceptions would somehow arise within the community of which particular schools were better, perhaps for irrational or unpredictable reasons. Parents would begin to transfer their kids to the perceived-as-better schools, and those schools would see their budgets begin to grow, enabling them to do more. Everyone within the perceived-as-better schools would begin to think they were in fact enrolled in a better school, and would begin to act accordingly. Gradually perception would become reality.

    So I think WSF has a lot of potential for getting us out of our current educational quagmire.

  14. How would you squeeze more students into a good teacher’s class if their classroom is already full? You have to build and heat a twice bigger school for that.
    Where I went to school the classes were mostly full.
    If you aim for the situation when half of the classes are close to empty, then you aim at building and heating a school that is half-empty. And that is hardly efficient.
    I am not saying that your idea is not a good one, but there are all kinds of subtle details. For example, there are two math teachers, one good and one bad. So the good one gets 100% students and the bad one gets 0%. So the bad one walks and then, who takes his place? Another teacher. But let’s assume that the first teacher is a genius and the best teacher in town, point. Then the new second teacher will still get 0% students, and so on, so what is the evolution of this situation?
    Also, good teachers are not always the more popular ones. An easy teacher may be more popular than a good but “hard” one. I have seen some exceptional professors in the university who are not popular because something about their personality does not appeal to students. Heck, I attend a university course right now by an exceptional professor, one of a kind, and the subject is one of a kind. However, the attendance is very lean, just because of 2 factors: the exam is easy to pass even without attending, and this is the earliest lecture in the day, at 8:15 am.

  15. Philg, you should be happy to earn such a paltry amount to contribute towards the common good and train the next generation of chopper pilots. To an unemployed person that’s a lot of money! At least that’s the logic I’m hearing from Fox News, courtesy of Jon Stewart:

    So what’s your pass rate and what the system do to make you teach better? 😉

    >Basically we would have to accept that government jobs serve as a welfare system for the incompetent and lazy.

    That’s a bit extreme and is lumping the good in with the bad, and you appear be blaming the workers, in this case the teachers, for the problems caused by a bureaucracy that’s mainly out of their control as individuals.

    I’ve seen exactly the same thing in big corporations; where the incompetent and lazy hang around for years and get paid a lot more than teachers. Shouldn’t this be a bigger concern as they’re wasting more of peoples money as they are paid more than the public-sector?

  16. Mick: Shouldn’t it be a bigger concern that an incompetent or lazy person works for, say, Intel, than as a schoolteacher? Despite my selfish interest as an Intel shareholder, I would have to say “no”. The incompetence of a Intel employee will not result in a child being denied an education and an opportunity to enjoy a self-sufficient life. You may reasonably believe that some or all government employees have earned the right to draw a paycheck despite not doing any work (because there are guys like Bob Nardelli getting paid $500 million for wrecking Home Depot?). But I don’t see how it is reasonable to believe that a 10-year-old child deserves to be parked for an entire school year in the classroom of an incompetent teacher.

    Separately, and much less importantly, the money to pay that incompetent Intel employee was voluntarily exchanged with Intel by people who preferred Intel products to competing products such as AMD. The money to pay for an incompetent schoolteacher is forcibly extracted, at gunpoint if necessary, from Americans who, on average, earn less per hour than the teacher and who may be working 10-30 years longer than a schoolteacher who retires in his or her 50s.

    [What is my pass rate? So far the students whom I’ve signed off for FAA checkrides have all passed (i.e., 100%). What’s my incentive to do a good job? There are competing helicopter schools at the Lawrence and Norwood airports. There are six other instructors at our helicopter school. Unlike a typical American schoolchild, our students enjoy the freedom to choose their instructor and their school so that they can find the most effective learning environment.]

  17. Depressing subject – about to get more depressing.

    I raised two kids in a nice neighborhood with “good” schools.

    Thought experiment: “You’re an administrator, responsible for hiring teachers in a good neighborhood, higher than average salaries for teachers.” Do you:

    a) hire nice looking teachers who seem dynamic and hardworking even though they don’t particularly like children and don’t know a lot about their subject.


    b) hire cranky people who don’t get along with adults but maybe have a passion for their subject and care about kids.

    The results are as you’d predict. The “invisible hand” of the market works much better for restaurants than it does for determining a quality education. Perhaps people are better at evaluating the quality of a taco then they are the quality of their kid’s geometry teacher.

    American schools are extremely sensitive to community input and do a superb job of balancing community demands. Educational quality is not very high on the list of things communities demand of schools.

  18. I have been waffling on this issue for some time. I am not sure the schools are completely to blame for the lack of quality education. But our Education Institution does seem terrible.

    I feel as though our cultural beliefs regarding education may share the majority of the blame. Yet, cultural values are such a nebulous subject that it’s hard to pin anything down. Especially if the critiques might not be particularly PC.

    The following article from the NY Review serves as a contrary viewpoint to some of the arguments made on this comment thread.

    The Myth of Charter Schools

  19. Phil, are your students very good at selecting instructors? Or many times they go with the first one they meet? Are the most popular instructors the best ones?

  20. Anonymous: Are students very good at selecting instructors? I doubt it, since they selected me! But many of do extensive research via online communities or telephoning experts. Here’s a video interview with a guy who talked to an FAA inspector and then flew up from Maryland to fly with us in Massachusetts:

    On average, I would say that our students do less research than a tiger mother would for her precious darlings (the Tiger Mom’s kids went to private school, so she had a lot of choice), but they do some research and typically fly with several instructors and/or several schools before settling on a plan.

    Are the most popular instructors the best? I would say so. A flight student does not need to be very sophisticated to distinguish between someone who gives clear and helpful explanations and someone who does not.

  21. When I went to high school, the students knew which teachers were good and which teachers sucked. The other teachers knew it, too.

    None of our output incentives are aligned, though. As a parent, it seems much more important for my children to get a College Degree than to get an actual education, and this seeps back a lot into high school.

    Also, a lot of the community views schools as day cares keeping the little vandals locked up for 6 hours a day. If I had a magic teaching machine that could zap a day of learning into a kid’s head in 5 minutes, a lot of adults would demand the perps be locked away so they couldn’t mess with stuff for the other 5 hours 55 minutes.

  22. This recent article about the private-sector test scoring industry suggests that you probably don’t want to put too much weight on test results.

    “The money to pay for an incompetent schoolteacher is forcibly extracted, at gunpoint if necessary, from Americans who, on average, earn less per hour than the teacher and who may be working 10-30 years longer than a schoolteacher who retires in his or her 50s.”

    I’m curious about this point. Would you describe yourself as agreeing with the libertarian view that taxation is theft?

    It seems to me that libertarians fail to recognize the distinction between authority and force. Generally speaking, the two are inversely related: the less legitimate authority a government has, the more it has to resort to force (as in Libya right now). Conversely, I can’t think of any recent examples of the US government having to force people to pay their taxes at gunpoint.

    Louis Halle, writing in The Cold War as History, describes the relationship between power and consent:

    … real power is always something far greater than military power alone. A balance of power is not a balance of military power alone: it is, rather, a balance in which military power is one element. Even in its crudest aspect, power represents a subtle and intimate combination of force and consent. No stable government has ever existed, and no empire has ever become established, except with an immensely preponderant measure of consent on the part of those who were its subjects. That consent may be a half-grudging consent; it may be a consent based in part on awe of superior force; it may represent love, or respect, or fear, or a combination of the three. Consent, in any case, is the essential ingredient in stable power–more so than physical force, of which the most efficient and economical use is to increase consent. By using physical force in such a way as alienates consent one constantly increases the requirements of physical force to replace the consent that has been alienated. A vicious spiral develops that, continued, ends in the collapse of power.

    The libertarian view that taxation is theft–that is, that taxation is illegitimate–is quite a radical one. It means that the legitimate sphere of government is drastically narrowed to the “Night Watchman” role of national defense and policing. Any other sort of problem where the government might step in, claiming that collective action is required, is an illegitimate attempt to overstep the bounds of its authority. (Alexander Volokh on protecting the Earth from an asteroid. The Onion.)

    This is a very different view from that of the Founders (although they’re unlikely to have had asteroids in mind!). In particular, the Federalist Papers, arguing for the ratification of the Constitution, explicitly argued that the United States needed a stronger government than that provided by the Continental Congress.

    There’s a number of major challenges facing the US which would arguably be best dealt with through collective action, i.e. through the government. In Canada, single-payer health care provides better outcomes at lower cost than the US system: in 2006, Canada spent 10% of GDP on health care (public and private), while the US spent 15%. Climate change is another obvious one.

    If the libertarian view prevails, it also leaves the field clear for corporate power. (I don’t know if you’re aware that the Cato Institute, one of the foremost exponents of the libertarian view, was founded by the billionaire Charles Koch in 1977.) Given the size of corporations today, one of the important functions of government is to provide a balance of power (another concept emphasized by the Federalist Papers).

Comments are closed.