Obama Wants to Pay Teachers What They’re Worth

It sounds as if Barack Obama has been listening to some economists (maybe even Austan Goolsbee): he has come out in favor of merit pay for schoolteachers. From an A.P. article:

Democratic presidential candidate Barack Obama told the largest teachers union Thursday that performance-based merit pay ought to be considered in public schools.

Teachers at the National Education Association’s annual convention have expressed concerns about merit pay, which is gaining favor with lawmakers, including those currently rewriting the No Child Left Behind law.

Teachers say they worry that linking their pay to their students’ test scores would be unfair to teachers who have students from disadvantaged backgrounds. Teachers also say it isn’t fair to offer merit pay only to people who teach courses that are tested, like reading and math, but not to those who teach subjects like music or art.

Obama said teachers’ salaries should be increased across the board, but he also said there should be fair ways of measuring teacher performance and compensating teachers accordingly.


One obvious disadvantage is that you are assuming that test scores at different ages accurately reflect the quality of the teaching. I should think that all it tells you is how well the teacher prepares you for the test.

If it does happen though, a large component of it should consist of improvements on last year, not absolute scores.

Matt Nicholson

Judging by experiences on this side of the pond (i.e. the UK), this is a bad idea. Performance-related pay means the introduction of some method for measuring teacher performance, which inevitably means that the teacher starts doing whatever is necessary to increase the measurement, rather than what is best for the child. This could have all kinds of unintended side effects.


No no no. Test scores are an inherently flawed way of measuring student intelligence (they're great at measuring regurgitation skills though!). Basing teacher performance on how well our children can recite facts is only going to make things worse. Education needs to be completely restructured to teach children how to learn and the joy of knowledge rather than beating creativity and curiosity out of them by forced memorization exercises and busy work.


I agree with #1 and #2.

There are two basic ways to do performance based compensation (in more or less any profession):

1) The boss judges performance.
2) Performance is evaluated by numeric goals.

#1 allows for a great deal of subtlety in evaluation - the boss can pick up many small things that the employee does well (or poorly) that a numeric system would miss. But I suspect that this would not work very well in the educational system, as personal and political factors might override true evaluation (or at least, those who got the short end of the stick would THINK that the evaluation was unfair, which is almost as bad.

#2 also seems impractical to me. My memories of my best teachers are that they taught a little bit about a lot of things - teaching not only the subjects at hand, but other academic and life skills along the way. Implementing #2 would create a very strong 'teach to the test' incentive, that I suspect would do more harm than good.



wow - you people are all crazy. how can testing a group of students NOT give you some better idea as to how those students are performing? How can some data be better than no data?

You all sound like the tired old mantras from teachers unions who reject any sort of attempt to objectively measure how well teachers do on their jobs.

I read a study once that concluded that average class size has NO CORRELATION to how well students do. You know what does strongly correlate to student achievement? parental involvement and good teachers. Yes some teachers can perform better than others - and it can be measured.

Teachers unions don't want this and instead tout small class sizes - to increase the number of teacher jobs! when what we should be doing is firing crappy teachers and giving good teachers larger class sizes.

Bravo to Obama, I say - this is just the type of stuff that might eventually fix the crappy public school system that we have. Higher pay, merit-bases increases, firing of bad teachers, and some sort of voucher system that will bankrupt crappy schools is just the kind of forward thinking that is needed to fix the educational system.



i'm a teacher.

i support the idea of merit pay.

i also (as an engineer) know that in order to make merit pay effective we must develop functional metrics to evaluate a students growth. the simple multiple choice tests, administered once a year with a 5 month feedback lag are not going to be effective at this.

developing a metric that both provides good feedback, and isn't capable of being gamed by the subjects (either students or teachers) is non trivial.

the largest obstacle to effective reform, however, is the overwhelming attitude of those who know nothing of what it's like in the classroom (as evidenced by badger99's post) who think they know the answers to very intractable problems, and who will throw a wrench at every valid argument made in favor of positive school reform.


Why does it have to be one or the other? Why not combine some standardized metrics with management evaluations? Isn't that how it works in most other jobs? Why is education so different? Some sort of algorithm based on test scores, average grades, percentage of kids passed to the next grade, classroom attendance, etc., then a final set of variables based on the qualitative evaluation? To get around grade inflation and whatnot, there could be some sort of auditing of the assignments, tests, and grading, measuring it to the district's curriculum. As for variables impacted by disadvantaged districts, these can be normalized to some sort of national average.


wow - you people are all crazy.

Boy, is that ever right. And you're crazy too.

Why does everyone ignore the statistical studies? It's not about teachers. It's not about reading to them, or taking them to the library, or that ridiculous No Child Left Behind.

Only 10% of students at select colleges are from divorced parents. But no politician is going to look out on a group of voters and say, "the problem with education in America IS PARENTING!"

It's about students from families of the lower socio-economic classes. If you go to the inner city and pay teachers by test scores, then you are going to be paying them the least money. The students who supposedly need the most help have the poorest paid teachers. THAT'S WHAT WE HAVE NOW!

Because energy is cheap in this society, we self-select our communities. The lower classes are seperated from the upper. The car is our escape from living with them. So poor students have no role model other than poor parents with poor parenting skills.

It's the parents...stupid! What are you going to do or say about that?



Well, to me merit pay is great, but as everyone already notes it is very difficult to define a metric. We bump into the fact that we need (a) experts in each area, that (b) are functionally universally accepted.

Just as importantly, merit pay operates on a market economy, and that means there will be losers. In other words, some students will HAVE to lose, because some teachers will be losing (and thus those students lose by proxy). Some teachers will be worse than others, and by definition the students they taught will have suffered. Those students are the 'costs' of this market, and we need to either accept that or look for a different option.


I'd also like to know when we will stop paying so much to our oldest teachers (who simply bide time for their last few years of work waiting for retirement) and retired teachers. What kind of messed up compensation system is it that pays more to teachers who don't teach (retired) than those who do (particularly young teachers)?

Every time a teacher complains about his or her pay, I remind them that they've chosen to belong to the unions that set up their ridiculous compensation system.


If teachers starting getting paid on merit, it is a guarantee that an increase in cheating will happen. Who is going to monitor that? How much will that cost? Personally I think the principals are better judges of who the good teachers are and the principals should be the ones who choose how much a teachers is paid whether to fire a teacher or not. Eliminate tenure and you won't have these issues.


As a public school teacher (high school math) this is a topic that while makes for a good discussion, the problem/situation contains far too many variables to ever really get a handle on any one of them.

While I am a propoenent of concept of incentive based pay in most professions, it would seem to make more sense in professions in which perfomance is more easily quantified. Take the cashier at Home Depot... couldn't we pay him more if he sped up the line?

On the other hand hand certain professions are very hard to "incentivise." Doctors and lawyers deal with far too complex situations. These professions have held controll of their public perception and resist having their performance "quantified" because of the complexity of their work.

Educating a child is just as complex and drawn out a process as any service a lawyer or doctor performs. Yet because the educational community has accepted (not embraced) testing, we have opened up ourselves to public scrutiny.

I would have to agree with commenter #3, the biggest problem we face as educators is that we are being forced to teach a lot of irrelevant information. As a math teacher, I find it difficult each year to teach certain topics (such as conic sections) that have very little practical use, but must be covered because the "test makers" determine them to be part of the core curriculum.



Every time a teacher complains about his or her pay, I remind them that they've chosen to belong to the unions that set up their ridiculous compensation system.

Oh yeah, that will make every young teacher happy. Guaranteed. No retirement. Ok, problem solved.

Why do conseratives so readily blame teachers? And liberals so readily throw even more money after bad?


Comment #3 is spot on. The test-driven curriculum and "lazy curriculum" of busywork is a very big problem, and I think it actually exacerbates the "poor parent" part of the equation.


As someone with an engineering background, I learned that most most products have an useful lifetime. Cars and software for example are updated frequently, but every so often because of complexity issues they are rebuilt from the ground up. Why should societal issues be any different?

Merit based pay and testing a are just like adding brand new high tech headlights to a beat-up 1970's Ford Gran Torino. It may help you see the raod better but the car is still a peice of crap and going to break down at any minute.

Many societal systems how outlived their useful lifetime, yet politicians only look to fix problems rather than re-engineer completely new systems. Merit pay for teachers will fix certain issues but it will create additional problems as well.

When companies want to enigeneer a new system they rarely ask engineers to multi-task, that is work on a new system while trying to update/support/fix the old system. Usually they assign these tasks to completely separate groups. Unfortunately we expect politicians to be able to do both... maintain the status quo, while providing forward thinking leadership. Is it any wonder that we get societal systems that are as complex and broken they cuurently are?



Why do conseratives so readily blame teachers? And liberals so readily throw even more money after bad?

First things first -- let's replace "conservatives" with "Republicans" and "liberals" with "Democrats." Like many other issues, each party has dug into their required positions with education. In fact, I find Obama courting the NEA the same as Romney wooing a crowd at Bob Jones. This wouldn't be a story had Obama not actually said something uncertain and unsettling about teacher compensation right in front of many politicized union leaders.

The big problem with education other than a decrease in parental involvement is the bureaucracy of it all. Before we start trying to tear down or reform teachers' unions, we should first look at the politics of running schools in the first place. We should look at systems that allow crooks to run districts into the ground while pocketing coffers. We should look at the favoritism continued to be placed on athletics programs that should be subsidized by money directly from parents' wallets or from (gasp) corporate sponsorship. We should question why test scores are the de facto measurement for performance.

Everyone wants to cry, "Teachers! Teachers! Teachers!" on either side of the debate, but no one dares question whether the entire system needs an overhaul. And why not? Because this is America, and in America, when we establish a bureaucracy, no matter how broken it is, we keep it running at all costs, even if that cost is the long-term economic health of America.



I think that parents, students, and teachers all need positive incentives. Static test scores for teachers' success is not as good a signal as dynamic scoring. To obviate the temptation to tank initial test scores, students need to be incented to score well. And I believe that parents should be able to write off what their children pay in taxes.


A voucher system, where parents and students opt in and out of schools, taking their money with them, will also make schools more competitive.


egretman - nice post. If you re-read my post, I said that parenting and good teachers are the highest correlators to a student's success. Your rant about parenting, while nice, has absolutely nothing to do with evaluating and paying teachers. Stay on subject here.

the largest obstacle to effective reform, however, is the overwhelming attitude of those who know nothing of what it's like in the classroom (as evidenced by badger99's post) who think they know the answers to very intractable problems, and who will throw a wrench at every valid argument made in favor of positive school reform.

I would argue that the largest obstacle to effective reform are teacher's unions that, by and large, resist any and all attempts to change the current system. If we all agree there is a problem, then why won't teacher's unions allow more change? I know the mass. teachers union resisted the state-wide MCAS testing system and fought it tooth and nail. When it finally was put in place, it allowed massachusetts to see where students were failing, what teachers were bad, what teachers were good - and test results have gone up. It might not be the answer, but it is a step in the right direction.

I can't believe how many commentors here seem to be resistant to using basic free-market economic principles to improve the system. Isn't this the freakonomics blog?



I agree entirely with #16, it is easy to blame teachers, but what about the politics of local school systems. The fact that an educational non-expert can get elected to the school board and make decisions that may negatively impact test scores (i.e. the politics of science eduacation and don't even get me started on how corrupt textbook adoption is) is ridiculuous. Would doctors and lawyers ever allow themselves to be controlled by your local band of elected ignorrant individuals (outside the profession) and then allow them to make meaningful changes to policy?