Valuing teachers

Good teachers are valuable, writes Eric Hanushek in Education Next. A teacher at the 69th percentile in effectiveness will increase lifetime earnings by $212,000 for a class of 20 students.  A very low-performing teachers at the 16th percentile will decrease earnings by $400,000. Compared to the average teacher’s salary — $52,000 in 2008 — the effects are large.

U.S achievement could reach the level of Canada and Finland, if we fired the least effective 5 to 12 percent of teachers and replaced them with average teachers, Hanushek estimates.

That would lead to a huge increase in economic output: $112 trillion over the lifetime of someone born today.

How do we do it? Hanushek has little faith that efforts to improve recruitment or teacher training will be effective. Trying to change poor teachers into average teachers also has proven difficult, he writes. “While such efforts undoubtedly help some teachers, there is no substantial evidence that certification, in-service training, master’s degrees, or mentoring programs systematically make a difference in whether teachers are in fact effective at driving student achievement,” he writes.

He proposes “a clearer evaluation and retention strategy for teachers” to “deselect” the least effective 5 to 10 percent. “At a minimum, the current dysfunctional teacher-evaluation systems would need to be overhauled so that effectiveness in the classroom is clearly identified.”

The teachers who are excellent would have to be paid much more, both to compensate for the new riskiness of the profession and to increase the chances of retaining these individuals in teaching. Those who are ineffective would have to be identified and replaced. Both steps would be politically challenging in a heavily unionized environment such as the one in place today.

With all teachers paid the same, regardless of performance, salaries now lag pay in other professions, he writes.

In the 1940s, the salaries of male teachers were slightly above the average pay for all male college graduates, and female teachers had higher salaries than 70 percent of other female college graduates. Today, despite the collective bargaining process, the salaries of male teachers are at the 30th percentile of the distribution of all college graduates, and women who teach are at the 40th percentile of their college-educated peers.

“Salaries several times higher than those paid teachers today would be economically justified if teachers were compensated according to their effectiveness,” Hanushek concludes.

About Joanne


  1. Maybe I am missing something, but there appear to be some dots that are not connected.

    A great teacher increases the lifetime earnings of a class of 20 kids by $212,000. So …. each child’s lifetime earnings were increased by $10,600 (212,000 / 20). Assuming a lowball number of years worked of 35, each child income will have been increased by approximately $302 / year ($10,600 / 35).

    For this small amount, it is suggested that effective teacher be paid several times higher than the average $52,000 salary. Why would you pay a teacher $150,000 a year so that one of their students would earn an additional $302 / year?

  2. I love how ‘source’ is “Author’s calculations” (in *very* small print). It’s made to look like it has a source, but in fact, it was just made up by the author from unnamed and/or non-existent sources.

  3. Roger Sweeny says:

    MN Mom,

    A teacher has a class for one year. If a teacher increases that class’s lifetime earnings by $212,000, he or she has in effect increased the wealth of that class by 212,000. In some sense, he or she has performed a service worth $212,000 to the class.

    Next year, he or she will have another class and will increase their lifetime earnings by $212,000. Once again, he or she has performed a service worth $212,000 to the class.

    If you assume that every good teacher could increase a class’s lifetime earnings $212,000 every year–so that 12 years of good teachers would increase the class’s lifetime earnings by $2,544,000– the math works.

    I don’t doubt Hanushek’s statistics but I think his conclusions are wrong.

    One: a very good teacher amid average teachers is probably going to be able to make a big difference because his students will be well below their potential. A very good teacher amid other very good teachers will encounter classes that are already working at potential. The bump you see in statistics now will not persist if all teachers are very good.

    Two: much of schooling is signaling. “I have a high school diploma so I will almost certainly make a better worker than the other applicants who don’t.” That makes a high school diploma worth a lot–and a teacher who keeps his students on track to graduate increases their potential earnings significantly. When everyone graduates high school because all teachers are good, the diploma no longer serves the purpose of distinguishing the graduate from the non-graduate. It is no longer a valuable credential.

    Of course, if what is learned in high school makes the student a more productive worker, the student should eventually earn more. However, as we all know, very little of what is taught in high school makes anyone more productive.

  4. Stephen Downes: try here:

    Google is a great tool, isn’t it?

    MN Mom, the idea is that the teacher is valuable not just because one student will earn more in one year, but because ALL the students will earn more in ALL years (on average). So the potential gains are really large. This really isn’t that hard to understand.

    (FWIW, I’m personally pretty skeptical that these gains are correctly measured.)

  5. I’m not sure what the point of this is if the states are unable to pay teachers as it is.

  6. Big flaw in Hanushek’s argument: there is no indication (as far as I know) that the students in the 84th percentile of high school achievement are the ones who made the greatest GAINS on standardized tests over the years. It is quite possible that they did well on the tests all along. It is also possible that those with the greatest gains did not end up near the 84th percentile.

  7. Peace Corps says:

    I think the point might be that if schools would get rid of the lowest five percent of teachers, taxpayers would be more willing to pay the remaining teachers more. As long as people can point to one teacher they know that is doing either nothing or is a totally crappy teacher, they are going to be unwilling to pay teachers more.

  8. Sean Mays says:

    Great, let’s get rid of the bottom 5%! But, what metrics are we to use? If I give low, but honest, grades does that make me crappy? We’ve got kids nationwide getting A’s and B’s but being placed in remedial education classes left and right once they get to college. Suppose we use some general standardized test of math ability. In order not to be crappy, should I resort to re-teaching kids what they didn’t learn before they got to my class and ignore the supposed content of my course; I’ll look “good”, but will have failed to deliver what my course promised. And on and on …

    The idea is a good one, granted. But the actual implementation is fraught with all kinds of perils, unintended consequences and moral hazard / agency problems. People point to the private sector for merit/incentive plans, but many of those don’t work so well either. Plenty of companies have been wrecked by senior managers (and regular folks) who got great compensation for “hitting their metrics”. If the private sector with more money and experience and motivation for getting this right hasn’t managed it well, I fear public education is likely to fall even further short of the mark.

  9. Peace Corps says:

    This isn’t really my area, but if you are a teacher like me, you are very lucky in your situation if you can’t think of one teacher in your school that is not doing his/her job to a minimal level of competency. I can think of 4 in my school, 2 at my son’s school, and 2 at a school previously worked for. Parents know about these teachers. Administrators know these teachers. Yet nothing is done to get rid of them.

  10. Sometimes it is more complicated than that, PC. For example, let’s say there’s a certain teacher I have nicknamed Captain Nemo for his innovative lesson planning with technology. Everyone knows he does this — yet he’s coached several teams to state championships. Who could get rid of him? Another teacher, just horrible, yet he does the yearbook — something most sane people do not touch.

    In any case, I find that the weak teachers in my own department move on pretty quickly — lots of peer pressure. I don’t have hire/fire power, but I make sure I am in those classrooms quite a bit.

  11. Roger Sweeny says:

    If the private sector with more money and experience and motivation for getting this right hasn’t managed it well,

    My God, what a damning statement. During the last hundred years, public schools have enrolled every American for at least eight years of their childhood and they don’t have enough experience? Hundreds of billions are spent on public education every year and not enough has been spent on figuring out how to keep good teachers and get rid of bad ones? There is an entire industry, the schools of education, which are supposed to have the motivation to figure out how to make things better and they haven’t done the job?

    What the ed schools seem to have said is, “There’s no way to determine who’s a better or worse teacher–except by how many courses they’ve taken from us.”

    Ah, but they are all honorable men and women.

  12. Sean Mays says:


    I think we’re talking at cross purposes. The schools have massive experience educating students, but their experience with performance based incentive systems is considerably less, if we’re talking about identifying and removing teachers we need a reasonable system for that. This has NOT been to my knowledge the focus of school administration, they’re relatively inexperienced at it. I was trying to argue that the business world has focused on it for much longer, with more people and more money at stake and they still haven’t gotten the kinks worked out. The latest numbers I could easily find were from 2006. US expenditures (state and federal) for K-12 education were about $562 billion (National Center for Education Statistics) versus total GDP of 13 trillion (Bureau for Economic Analysis) that’s about 5% of the total economy.

    I’m not even going to touch the ed school angle. If math wars and whole language versus phonics and social promotion and self-esteem, etc don’t make one question the value proposition of ed school; I don’t know what will.

  13. Changing the subject slightly, the present value of an additional $212,000 spread over a lifetime is a lot less than $212,000, especially if it doesn’t even begin for 10 years or more.  For instance, at a 5% discount rate, the first 30 years of an additional $212,000 spread over a 40-year working life ($5300/yr) starting 10 years ahead is only worth about $50,000.

  14. Roger Sweeny says:


    I completely agree. “The schools have massive experience educating students” but until now have hardly tried to identify better and worse teachers–and have affirmatively refused to try to pay better teachers better. When I tell my students that, they are incredulous. That can’t really be right, they naively think. Of course, you try to find out who’s better and worse. Of course, you pay the better performers more. But–once past those first three years–most schools don’t try at all. Schools haven’t gotten to the point where they have to get “the kinks worked out” because they have no system at all.

    To be fair, they do have a system. You are paid more the more ed courses you have taken–and in some places, you will be fired if you don’t accumulate a sufficient number of credits in a certain number of years. But just about everyone in the business knows that courses taken have nothing to do with how good a teacher someone is. Yet nobody really challenges the system because it is easier for everyone. Administrators don’t have to try to evaluate people, and teachers who want more money know they can take courses (which they are pretty much guaranteed to pass) and move up a pay grade.

  15. It’s not that hard to determine a teacher’s value added, at least in areas like reading, math. language arts and science. Just give the students a computer adaptive test such as MAPS in the fall and again at the end of the school year and use it to determine whether students have made more or less than a year’s growth. Because it’s done at the beginning and end of the school year, the teacher is not penalized for summer drops which tend to be higher for low SES and English language learners. Because the test is adaptive, it permits the students who are above grade level to show their actual growth over the year.

    I prefer this to administrator observations which tend to degenerate into a check off list that stifles innovation. For example, last week I had my second grade students writing similes. They were so excited and engaged that students were shouting out their favorite similes, laughing, and running to their friends to share what they had written. An administrator might well have concluded that I have problems with classroom management, but my value added scores show that my methods pay off.

    Lightly Seasoned, it’s unclear what you are accusing “Captain Nemo” of doing. Are you suggesting that he “pirates” lesson plans off the internet? As long as he’s not breaking any copyright laws, I don’t see what the problem is. It’s better to download an excellent lesson plan off the web than to waste time writing a mediocre plan. Why reinvent the wheel? Again, student growth should be the goal.

  16. Ray: “Finding Nemo”

  17. Lightly Seasoned, I know that you’re being witty and clever, and it’s a shame that a trog like me doesn’t get it. But could you just describe what this teacher is doing wrong? “Finding Nemo” doesn’t help me at all.

  18. Steve Jones says:


    I was just browsing these comments but feel inclined to respond to your suggestion about testing a teacher based on a MAPS test. And since you’re a teacher I must say I’m surprised. There are so many factors that could influence such a test. If a teacher spends the year trying to get unruly students settled down for lessons, that may be a huge success and there would be no progress on the tests you can see. Perhaps there are learning disabled kids in your class and they may show some beginnings of understanding, however they show no test gains. In my class, I have 32 kids and most of them are ESL (or former ESL). Throughout the year I get many new kids coming straight from China. The kids just learning English (who speak only Chinese at home) may make progress in a base level on understanding of English, but would still show no progress on a standardized ELA test (when we prepare for the state test they essentially guess and leave writing responses blank — they’re just not ready). This is the problem with all of this testing/assessment. There’s one and ONLY one way to evaluate a teacher — do the work. You can’t pick a test and use that since there are so many ways a teacher has an impact on his/her students (you should know that). But nobody wants to do the work, instead they want to find a single standard. This author, for example, has created from thin air a calculation on how a teacher impacts income, never really highlighting how her percentile or income numbers were found.

    I’ll use a similar example to yours. I had a publishing party recently for a “fantasy” writing unit. My students were very excited and parents (as well as administrators) were thrilled at the progress shown by my students. Now, I can’t speak for the makeup of your class, but my class is full of kids on food stamps, with IEPs, and as I said before kids with little or no English. Yet there were kids with little English who were able to get a few sentences down on their own, kids who go home and speak Chinese who were able to show character development they had never before displayed, and IEP students who were able to get down a full story rather than a sentence or two. NONE of this shows up on the standardized scores. This level of progress gets simplified to 1, 2, 3 or 4. My “value-add” is therefore low, even though peers, parents and admin can see the progress. But somebody crunches numbers somewhere and therefore I’m “ineffective.”

    One last note: Be careful on touting your “value-add”. Many formulas tend to show a lot of fluctuation. I saw mine go from “above average” to “ineffective” in one year. When you’re around a few years and have a bad score come from thin air you may not be so thrilled.

  19. Computer adaptive tests are very different than the standardized tests that you are describing. The difficulty level of the test is adjusted as the student is taking the test. This means that a student can test way below grade level at the beginning of the year, test somewhat below grade level at the end of the year, and still show that the student has made more than a year’s growth over that time. Also the MAPS test uses a RIT score which is an equal interval scale that goes over 200 points. This is much more fine grained and allows the test to pick up student progress that would be missed on a 1, 2, 3, or 4 scale. I agree that the use of end of the year state tests is inappropriate for determining value-added. It is a shame that this is the most common way that value-added scores are currently determined.

    For what it’s worth, I teach in a Title I school with an ELL population of about 20% and my classroom size is 25 students. It’s challenging, but your situation sounds much more difficult. I have enormous respect for what you are achieving.

  20. Roger Sweeny says:


    It would certainly be unjust to judge you based on your students’ scores on a 1-2-3-4 scale.

    It seems equally unjust to ask you to teach students who can’t write English, in the same room and at the same time, as students who read and write at grade level. Unless you are one of the exceptional teachers who can successfully differentiate instruction across such disparate levels, the situation cheats both those at the low end (who can’t follow) and those at the high end (who proceed slowly lest the rest of the class fall behind too far).