A bill linking teacher pay to student performance passed the Indiana legislature today, just one of many such measures wending their way through Republican controlled statehouses nationwide. “Merit pay” is a perennial proposal here in Washington state too, and a longtime favorite of conservative “reformers” everywhere.
But will merit pay measurably improve student performance? I don’t think so, and to understand why, we need merely examine the Seattle Public Schools, starting with the following two assumptions regarding student outcomes:
- The better the teacher, the better the performance; and…
- There is substantial disparity between Seattle public schools.
These assumptions seem pretty self-evident, and thus all else being equal, it should follow that the better schools must have the better teachers. But of course, all else isn’t equal.
Educational performance within the Seattle school district largely tracks along socio-economic (and thus racial) lines, with the best performing schools generally found in the whiter, more affluent northern neighborhoods, and the worst performing schools clustered in the poorer, more diverse Southend. So if the Northend schools owe a substantial amount of their advantage to teacher merit, there must be some mechanism by which they are attracting and retaining better teachers.
But there isn’t. HR often assigns teachers to schools with little input from principals, and whatever say teachers have in their own assignment is based on seniority. And even if the district wanted to use assignment to favor one school over another, the metrics for evaluating teacher performance are iffy at best. Thus lacking a mechanism, market or otherwise, to account for an unequal distribution of good, bad, and average teachers along the north/south divide, we can only assume that there isn’t any.
I’m not saying that teacher performance isn’t a factor in student outcomes. I stand by assumption number one. Merit matters. But it just doesn’t seem to matter enough to even begin to explain the outcome disparities between Seattle schools.
Thus even if merit pay could work to incrementally increase teacher performance (a big “if” considering the lack of credible supporting data, and the simplistic/deterministic assumptions it makes about what motivates teachers), incrementally increasing performance across a district or across a state would have much less impact on overall student outcomes than closing the yawning performance gap between schools and between districts. For the bigger problem facing educators isn’t that our good schools aren’t good enough, but that there aren’t enough of them, a disparity that merit pay simply cannot solve.
So it shouldn’t be done?
Just as health insurance companies compete by getting only getting healthy customers if this comes to pass schools will compete by getting high performance students. The problem is the way insurance companies avoid sick people is by giving bad service to people who file claims. Schools will cancel remedial programs and it will go back to the bad old days for special needs children.
But if Seattle can’t concentrate educational resources North of the Ship Canal, how will we fight off the non-white majorities outside?
Seriously, though, it’s not state or district funding that impacts education results – it’s the ability of rich parents to invest their excess capital in their kids, show up and demand resources at city, district, county, and state public meetings, dump tons of money into after-school programs (both private and public, the latter thru PSTA and other fundraising), and pay for private tutors when their kids fall behind. Most of which we then DEDUCT FROM OUR TAXES.
Poorer neighborhoods can’t do that, overall.
The system is fixed. Just plot where the private school kids come from and you’ll see it’s even more fixed than you think it is.
Why does Goldy’s daughter attend school on Mercer Island then? He just had a post how he is from the South End, you cannot complain about the situation if you have no skin in the game.
Do you feel the same about increasing teacher pay for receiving advanced degrees and/or getting national certification? I thought the rationale for increasing pay for doing those things was that it meant better teachers.
@4 cause he’s planning for the light-rail-enabled future.
there are fundamental issues w/ merit based pay outside of the problem of identifying whether the benefit was produced by an individual’s merit. below is an excellent talk given around merit based rewards systems & how & why they fail to produce the best results. anyone who has worked at a large bureaucracy (& picked their head up occasionally to see what was going on around them) has seen the failures of merit based pay in action. also, the studies that disputed the value of merit based pay were funded by the federal reserve & carried out by economists from mit, harvard, u-chicago etc. they were all sure that merit based pay was the solution they were looking for & the experiments proved just the opposite.
http://www.youtube.com/watch?v=u6XAPnuFj…
no quick fix, but just a thought.
Professors are a professorial teaching population. Ostensibly published professors, or ones who are members of the Academy of Sciences (or other organizations) are more worthy profs, because they’re involved in their field/currently engaged in new findings.
Can we create professional peer organizations for teachers? I think profs have more time on their hands to get published than a k-12 teacher has to be involved in a teacher organization. BUT materially it has to be easier than putting a man on the moon. Good veteran teachers recognizing good up-and-coming teachers, teacher halls of fame for commitment/innovation, other-things, etc. I know there is the golden apple award but I’m not sure of much else of this ilk exists. Teachers should be able to “level up” as it were in a manner that would be beneficial to their students.
You seem to be making an assumption that “best performing” = high levels of student proficiency. But, that’s probably the last way I would define good performance for schools. I’d be more inclined to look at growth on the student level (what do you do throughout the year with the kids you’re given, not, do you have a bunch of high SES kids with high proficiency in your school).
Also, merit pay doesn’t have to be about just providing an incentive to perform better. I’m a proponent of merit pay, but I don’t favor it because I think teachers will perform better if they just had a monetary incentive to do so. It’s at least as much about fairness for me. Is it fair that an excellent teacher gets paid less than a teacher with more seniority who hasn’t written a new lesson plan in 10 years and shows videos 2 days a week (a real example from my own experience in school)? Ultimately, a fairer way of compensating teachers (along with higher pay levels) may attract & retain more/better teachers to the profession (indirectly raising performance).
Additionally, if the merit pay structure includes an evaluation of teaching methods (which it should — multiples observations by multiple people in a year), then it would normalize the process of coaching and giving performance feedback to teachers (something that doesn’t happen often today, at least after teachers obtain tenure). If there’s money on the table, then observations/evaluations matter, and can’t just be a throwaway (as they are in many schools today). So, that may also help improve overall performance.
Merit pay is not an easy thing to get right — it’s tough to design a fair system that effectively measures performance (both student learning results and teaching methods) and deals with all the devilish details. So, I’m cautious about implementing it whole-hog on a state-wide level. But there are schools out there doing it well, and I think it’s something worth pursuing.
@5 Most of the research shows that teachers with advanced degrees are not better teachers than teachers without advanced degrees (unless the subject of the degree is matched to the subject taught — e.g., a masters in Math for a math teacher). So, personally, I’m opposed to paying for degrees.
But, national board certification has been shown to be a good way to identify high-performing teachers. So, I think paying more for teachers who are NBPTS certified is a great idea. I know they pay more for it in Chicago (and the union – or the district? can’t remember — pays for the certification process).
Poverty stricken white kids do better than poverty stricken black kids. Take some poverty stricken Chinese peasants and bring them to America and they’d do better than poverty stricken white folk. Poverty has very little to do with it, evolution has almost everything to do with it.
Godly, you dirty commie.
Does paying more for seniority measurably improve student performance?
The longer the teacher has been punching the timeclock the better the student performance?
Really?
How about a compromise, instead of merit pay we give out merit badges.
People talk about this concept of “improved student performance,” as if everyone knows what that means. I’ve never heard a satisfying definition of it, and if we can’t define it, we can’t measure it. If we can’t measure it, we can’t bribe for it. If it’s all about test scores, you can certainly spend money to improve the training for tests, but the skill of test-taking has limited utility outside of school.
The Seniority Myth
Merit pay is certainly flawed, but I’m unclear how it could be more flawed than seniority pay. The concept of merit at least attempts to measure outcomes; basing pay on seniority is just throwing up our hands and saying it is impossible so we’re not even going to try.
I would like to see more effort put into improving the metrics for merit pay, absolutely. But in order to get that effort, we need people to agree that it is a fundamentally good idea, even if difficult to implement.
And in the meantime, I’d like to see people acknowledge that just because seniority is easier to measure than performance, that doesn’t mean we should base pay (entirely) on it. Height is also easy to measure, and nobody’s suggesting we use that.
@4,
First of all, my daughter spent 7 years at Graham Hill Elementary. But she has a mother too, who lives on Mercer Island, and given the choice between Aki Kurose and Islander, I think we made the right one.
But you raise an interesting point. If my daughter isn’t in the Seattle schools, I don’t have a right to write about. I’ve also been told that if I’m not working at a union paper, I shouldn’t write about labor issues, and since I’m not a nuclear physicist I shouldn’t write about Fukushima.
It seems the only think I’m qualified to write about is myself, and then I get criticized for too many first person references. Guess I should just quit.
18
don’t let the door hit your fat ass on the way out
If a merit based system were implemented, the job of evaluating the teachers’ merit would go to the lowest bidder, and the end result would be a shitty, half-assed merit based system that would barely work and that everyone would complain about.
Money has everything to do with it.
The second sentence says merit pay is a proposal in Washington state, and then the whole rest of the post, and the discussion, is deep in the weeds of the Seattle school district. McGinn’s conceit that he could run Seattle schools is spreading.
If the state is proposing merit pay, why not discuss what it would mean for the state, which actually is diverse? Rather than just whitey white Seattle?
@19,
Anybody who knows me knows that I have a boney ass.
Performance pay is all about control, not performance.
It’s not that the theory is wrong. Yeah, we’d like to have great teachers. The problem is that teaching is such an enormously complex task that *no-one knows what a great teacher is*, let alone how to measure it.
Case in point: a recent study found that kindergarten interventions (in class size) had a small effect, which wore away by the time the children were 12. So not much point, right? But then a follow-up study finds that the children who had smaller kindergarten classes got better-paying jobs, despite getting equal graduation grades to the other group. So is kindergarten class size important? Who knows!
Given that we cannot at present measure teacher quality, we should stick to doing the few things we do know to work: giving respect and resources to the teachers we have, rather than grading them into meaningless strata.
@8 there certainly are professional peer organizations for teachers- the National Council for Teachers of English is just one for example, but there are many. They publish journals, host conferences, and offer professional development throughout the country.
@9 students who come from low SES backgrounds experience smaller growth in skills and knowledge than students from higher SES backgrounds. It has to do with the messages and preparation they receive at home so they can come to school ready to learn. The gap gets bigger every year between students from high SES and low SES backgrounds. Additionally, high poverty schools have high levels of transience- you don’t finish the year with the same kids you started with.
@13 and 17- you seem to be buying into the Fox news narrative that experienced teachers are just punching the clock before they can begin collecting their sweet pensions. Public school teaching is actually a profession that requires constant professional development in order to maintain certification. Studies show teachers with five years of classroom experience improve student learning more than first and second year teachers. Experience in the classroom is actually a good measure of teacher success. Fifty percent of new teachers leave the profession in their first five years, before they ever become really successful teachers. It makes sense to pay teachers more to stick with the profession, improve, and take on leadership roles in their school communities.
If there’s a Nobel Prize equivalent for “Most ironic invocation of the Argument from Ignorance”, I nominate this post.
Horrible idea!
@will in seattle
As I’ve been getting older, the idea of parents paying a little extra in taxes for their children’s education seems like a good idea. There was a bill in the house last year to equalize the bank accounts of the school districts so we could get rid of the poor/rich situation, but it met a silent death.
@8 – Those organizations already exist. Like the National Council of Teachers of English. The NCTE membership goes from PreK all the way through the college level, and has special cohorts for the different grade ranges.
I know such organizations exist at least for Math, Science and Reading as well, and I would assume for other subjects, too. I know for sure the NCTE has a journal – The English Journal – and publishes mainly practicing teachers on the PreK-12 level.
Those organizations do make a difference if you join them and read the publications, which not all teachers do.
The fact of the matter is that teacher quality isn’t the best predictor of student success. Bright, motivated kids with stable families (and a single parent home can be a stable family, as can a home with two dads or two moms or whatever) with a decent income are going to succeeed in spite of lousy teachers. But poor kids without a strong example at home, who don’t always have enough good food to eat, and who don’t see a future for themselves are going to struggle no matter who you put in the classroom with them.
I agree with the posters above who said that pay shouldn’t be based solely on seniority. But until there is a good way to measure teacher effectiveness, beyond standardized test scores, I don’t know of another method that will work.
Ken Mehlman- where do you get the idea that schools just teach our students to read and write American English? We are not living in the 1800’s. Schools actually teach a variety of subjects including math, science, and technology skills as well as preparing our young people to be critical thinkers and global citizens. Jesus Christ- it’s not a factory of reading and writing instruction and testing.
For the last few weeks I’ve been listening to a grade Z Seattle actress read an advertisement about this topic on the liberal AM talk station. I admit to a little bit of conflict: As a union member myself, I’ve often thought that seniority is a double-edged sword, as it can, under poor management, allow mediocre people to coast – and this country is currently experiencing an epidemic of poor managers. It seems to be something they are teaching in business schools.
However, I’m quite wary of this concept when it comes to education. What would happen to a dedicated, long-term teacher who chooses to work at an under-performing school? Would it mean that all the good teachers with seniority would flock to the well-funded, parentaly engaged schools like Laurelhurst Elementary, while schools in poor neighborhoods would get a revolving door of new teachers who are paid the lowest salaries?
Despite what the failed actress on the commercial implies, just because you are new to a job, that doesn’t make you better. It just makes you cheaper.
It’s really hard to know what the goal is or what to measure since now even prestigious 4 year college degrees are coming under scrutiny for their lack of return on investment.
I’d almost rather do away with compulsory education entirely and make all schools more like a combination community college and library where people/kids can come to socialize, play, do athletics and gather as much book learnin’ as they feel up to…
Catalina @30, your second paragraph bears repeating every time this nonsense comes up. Thank you.
And I often pick on without remembering to praise, so wanted to offer this late shout-out to Goldy for kicking off a fine discussion of all this with your good, good post.
There is some legitimate merit in seniority pay. It is pretty easy to demonstrate that generally an experienced teacher is more effective than a new inexperienced teacher. Thus, it makes sense to pay an experienced teacher more. This is true of almost any occupation. It shouldn’t be the only measure of teacher pay, but it is certainly legitimate for experience and/or seniority to be part of he equation.
@34 – CVDR’s graf 2 acutely nails the dynamics of the status quo. New teachers disproportionately staff “bad” schools, because that’s where the openings are. After a little experience, they bail out … or clamber up to better schools, or better districts, or private systems … leaving a new round of openings for a new round of new teachers in the “bad” schools.
I also urge extreme skepticism when considering most “reform” ideas offered by republicans. Many republicans, particularly in the south, hate public education entirely, and view it as a handout to poor black kids (many white kids were sent to private school when bussing started in the 1960s). Half of what they propose is intended to gut or sabotage public education.
Merit pay makes a good sound bite, and it sounds like a good idea if you don’t think about it much. But the big problem is trying to effectively evaluate teachers. Test scores are a poor indicator. Kid’s test scores can be effected by poverty, shitty home life, hormones, how much sleep they had the night before, what they ate for breakfast (if anything), if they just broke up with their boyfriend, what drugs they are on, etc… all of which is completely outside the control of teachers, no matter how effective they are. Teachers work solo 99% of the time, with no observation by principals or other administrators more than a few hours a year. So there isn’t really anyone who can judge their performance by observation. And observational measuring like that is highly subjective, and can fall victim to the popularity and personal prejudices of the principal or administrator. Yes, in an ideal world, it would be great if we can pay excellent teachers more, and get rid of shitty teachers. But as a practical matter, it is really, really hard to fairly and effectively measure teacher performance.
The one thing that has proven again and again to have a measurable difference in outcome is to have smaller class sizes. That means more teachers. That means more money. But we’d rather pay less taxes and gut our schools, and blame it on a lack of merit pay than for the one thing we know would actually help.
There are many problems with the merit pay proposals on the table right now, but the one Goldy concentrates on isn’t one of them. All the concrete merit pay proposals I have ever seen measure your merit by the change in standardized test scores between students going in and students going out of your classroom. So if you are a south Seattle teacher who gets a bunch of dunces and elevates them to being merely poor students, you will be better rewarded than a north Seattle teacher who gets a bunch of decent students and they exit your classroom as still just decent students.
That said, a better proposal than trying to legally mandate any particular objective quantification of teacher merit is to just say “the principal is the boss; he can hire and fire whomever he likes and pay them what he wants; he is accountable to the administration and the parents”.
There are only three reasons the Republicans are pushing “merit pay” for teachers: the primary one, is to break the teachers unions, of course.
No longer will the union be able to set standards for pay raises (for advanced degrees, for instance), and hiring and firing.
Then, if Republicans allow their children to attend public schools, those will be in the lily-white, moneyed suburbs where students will score high, and they’ll get the “better” teachers because they’ll be able to pay more for them.
Fuck the poor neighborhoods.
Third, down the line: follow the money — I’ve read that those “standardized teach to the tests” and the lesson plans to prepare for them was a Bush family member’s business.
Whenever Republicans are pushing a plan of any sort, it’s because one of their own has dreamed up or has a business that will benefit from that “privatization.”
All privatization boils down to “one of my friends, relatives, campaign contributors, needs to make a big pile of government cash.”
How does the Seattle data inconsistent with merit having a big impact? Imagine that merit has a big impact. And imagine that south-end schools get bad students and north-end schools get good students. And imagine that good and bad teachers are equally distributed between north-end and south-end schools. Then you would exepct that some south-end classrooms would see lots of improvement and some would see none, and some north-end classrooms would see lots of improvement and some would see none, so that overall the south-end students would still lag the north-end students. Which is, you know, what we see.
Here is a numerical version of the model. Measure student performance in “points”. Suppose half of teachers are good and half are bad. Suppose a good teacher increases the performance of his students by 10 points, and a bad teacher increases the performance of his students by 0 points. Supose, as Goldy does, that good and bad teachers are equally distributed between north and south Seattle. So the average student, whether in the north or the south, has his performance increased by 5 points per year. If south-end students enter school performing at 50, and north-end students enter school performing at 80, then the 30-point gap will remain between north and south students, despite the fact that good and bad teachers have a substantially different impact.
Uh, David @ 39, do you realize what you just said? If you put everything in the hands of the principals, you are essentially putting everything in the hands of the School Superintendent or whatever that job is called where we just got rid of ours in Seattle because she was incompetent overseeing money and a misanthropic asshole on top of that. The teacher union system may seem very political (example my son had a xxxxxx teacher last year who kids were sure was smoking dope and who couldn’t or wouldn’t teach well and who only this year was forced out after 1000 years in the system), but in fact the current system distributes power in a way that mitigates the politics and eventual corruption and nepotism that comes with dictatorial systems. The “big daddy save us!” strong-man fantasy isn’t the answer.
Oh, that accountability to the parents thing? That is a joke. The one thing the school board and Goodloe refused to do was to account to parents. Goodloe would snear condescendingly at parents when she visited a school and you tried to talk to her. I get much better accountability being a pest with individual teachers for my kid.
Ah sheesh. Why waste time talking to people whose thoughts derive from ideology rather than fact?
@42 You are not incorporating the truth that the substandard base of the students in the under-performing schools/economic/social classes means that the gap will grow overtime all things being equal. This is why it is so hard to help these schools – they don’t just need more resources – they need way more resources.
Also, you are not incorporating in your model the impact of there being multiple teachers each day for each kid after 5th grade in Seattle. My son has had very bad teachers take up huge amounts of time and energy for my kid in order to meet the the bad teacher’s standard, thereby having less energy to put into other classes with a better TRUE learning return on effort.
@23 is right: “pay-for-performance” is really about “pay-for-compliance”. The GOP has always hated secular public education. First they weaken the institution by discouraging attendance and degrading the unions, then they will control the remainder through market pressure, in particular the de-secularization of curricula.
As a high school German teacher, here’s a question that I’ve never seen asked, let alone answered. How are they going to determine my merit? No one tests my students on what they learn in my classroom. Are all high school electives teachers just left out of possible pay raises from here on out in a merit pay system? I really would like to know what states that have adopted merit pay have done for those teachers.
Cracked @43:
The managers-evaluate-their-reports model works in thousands of successful large companies across the U.S. While lots of companies do try to put a semi-objective verneer on the process, I’ve never seen any company do it purely on the basis of test scores; fundamentally it comes down to trusting managers to evaluate well. If they don’t, either the business goes south or they are eventually replaced by better managers.
I agree that the fact that school districts face few competitive pressures can cause problems for this model. That’s why I’d actually go further and advocate competetion between schoools. Essentially, I’d propose that school districts disappear; each school becomes its own business. The state pays it a certain amount for each student whose parents choose to put him there. The principal is essentially the manager of that business. If the business is poorly run, it attracts few students and goes out of business. If a school-owner attracts many students and turns a good profit, he may choose to open new schools that duplicate the successful model. If not enough decent schools are open to satisfy voters, the state increases the per student voucher.
Cracked @44: I absolutely agree those are real effects that are not incorporated by the model. I don’t claim that the model is complete or even fundamentally correct. I merely claim that it shows that the persistence of a north-south difference is not inconsistent with teacher quality mattering, which is what Goldy claimed in his original post. As far as I can see, none of the additional effects you propose incorporating would mean that a north-south difference could not persist if teacher quality matters.
Marstorr @46: Gute Frage. Leider habe ich keine Antwort. Posten Sie mal, wenn Sie eine finden.
@36 – Absolutely this is true. Virtually all studies show that a 2nd year teacher is better than a 1st year teacher, and a 3rd year teacher better than a 2nd year, etc. But, most studies also show that the benefit of seniority drops off fairly quickly . There’s some debate here about the exact point, but generally, after 3-6 years, there is no additional benefit to an additional year of experience. Current seniority pay systems reward for anywhere between 10 and 30 years of experience (depends on the district – 15-20 is pretty typical).
David Wright – I think the problem with your argument about “just let the Principal do it” is some combination of (A) school Principals do not have the kind of training/experience in evaluation that business managers do (B) quality evaluations and performance feedback are not the “norm” in schools (though, to be fair, this is hard to do, and many businesses struggle with it as well) and (C) there tends to be more of a culture of favoritism in schools. Not that these things can’t be overcome… But, anyways, I think the Principal should play a strong role in evaluation, but others (APs, Deans, even Master Teachers if a school has a career ladder) should also observe teachers and have input into the evaluation process. Kind of like at many professional services firms — as a consultant at a big firm, I had a direct manager, but the entire practice leadership came together to determine final performance / pay decisions.
@46 – there are a few ways to handle this issue, none of them perfect. Denver has language/music/art/etc. teachers set their own student learning goals at the beginning of the year (what are the concepts student need to master, how will the teacher measure this, what are the specific goals — e.g., 90% will master concept X). This requires a rigorous review process to make sure goals are consistent & fair.
Some schools/districts don’t use student learning results at all in their evaluations, it’s all about observations and evaluation against specified teaching behaviors (e.g., Charlotte Danielson’s Framework for Teaching).
It’s one of the more difficult issues with performance pay, if a school wants to use student learning results in their system.
The Merit Pay idea is a copout. An attempt to get something for (almost) nothing.
One of the most effective, proven, ways to really improve education is to have smaller class sizes. But that costs money, and between Tim Eyeman being a greedy self centered asshole, and the Democrats being too chicken to make any attempt to increase revenue, we are slashing education funding, so we can’t hire more teachers. Instead, we want to try to give an incentive to better teachers and get rid of shitty teachers. A wonderful idea, but the proposal is mostly smoke and mirrors, and won’t work.
I have a good friend who used to be a math teacher. And a really good one. But she left to become a computer programmer, which paid 3 or 4 times what she made as a teacher. Offering to pay her a couple thousand dollars a year more under some sort of merit pay system wouldn’t have made the slightest difference in her decision to leave teaching. If you want to give excellent teachers a monitory incentive to stay, then you’re gonna have to come up with something a lot better than this shitty merit pay system. You’ll have to pay them a competitive wage to what they could make elsewhere.