Merit Pay, continued

Avatar

Will

Will writes from Washington, D.C. (well, Arlington, Virginia). You can reach him at willblogcorrespondence at gmail dot com.

Related Post Roulette

No Responses

  1. Avatar E.D. Kain says:

    Merit pay alone will probably not do much in the short term, but in the long term it promises to bring in and keep more skilled teachers. Of course, if we spend too much time quantifying and teaching to tests, skilled teachers will go work somewhere else where the merit pay is as good or better but they have more room to breathe. So autonomy is easily as important as pay, though the two together make the most sense.Report

  2. Avatar Freddie says:

    Another argument that would be laughable in any other profession is that teacher achievement is just too hard to quantify (see here or here), making any merit-based pay scale unfair and vulnerable to favoritism.

    You may find it laughable; it is also true, as I have written at length about before. I don’t find a handful of links to information generated by people with an inherent interest in advancing certain readings of educational data to be persuasive.

    There are many things to point out undermining the notion that there is a simple or clear connection between student input and student output, but they deserve a space of their own to be explained. What I will say just briefly is that one fact that is understood perfectly well within teaching and not well outside of teaching is that individual teacher’s student performance varies wildly from one year to the next, both in terms of changes to grades and improvement to standardized test scores. One year a teacher’s class will show the best improvement in a school or even district in a given standardized test; the next year that same teacher will be among the worst or even the very worst. It happens all the time. Do you suppose that these teachers are suddenly becoming markedly worse at their jobs than they were the year before? And then, as if by magic, suddenly swing in a positive direction again the next year? Or is the fact that individual students have a myriad of factors that contribute to their own academic success or failure (not least of which is their effort, which can vary wildly given life circumstances)? Further, there is often no correlation at all between the teachers that are considered the best among administrators, parents and colleagues and those who score the best on “objective” means like changes in standardized tests. Are all of those people wrong? Are parents worng if they prefer one teacher who is engaged and helpful and caring over one who seems bored, disinterested and disillusioned, if it’s revealed that the former saw worse standardized test improvement and less improvement in student retention of comparable material?

    We embrace this system – warts and all – because it works; employees who get rewarded for performance are, on balance, more successful and productive.

    I believe that this, too, is untrue, and certainly not true in any kind of pleasant, uncomplicated way.Report

    • Avatar Will in reply to Freddie says:

      As a general rule, I find data more persuasive than anecdotes.

      Look, I buy that classroom performance varies wildly from year to year. But I don’t think this means that it’s impossible to identify good teachers for the purposes of compensation. It’s also worth noting that there are very teacher-friendly options for implementing merit pay out there, like the two track program I mentioned earlier that preserves a tenure-based option.

      Finally, we’re talking about trying merit pay in failing school systems like DC. Is experimenting with performance-based pay such a risk when the status quo simply isn’t working anyway? Your cost-benefit analysis just doesn’t make sense to me.Report

    • Avatar Ryan in reply to Freddie says:

      This seems like a pretty dangerous position. If classroom results are uncorrelated from year to year, that’s an awfully good argument that teachers simply don’t matter. Let’s just put all the kids in an empty room and see what happens. We’ll save the taxpayer a lot of money.Report

  3. Avatar Dan Miller says:

    Freddie–if improvement in test scores isn’t a good way to measure teacher quality, it seems like we could come up with some way of measuring who actually is a good teacher. Maybe increased collaboration amongst teachers, coupled with some form of peer review; maybe evaluation by principals, or other professionals who monitor teaching; maybe even videotaped teaching evaluated by off-site experts. These are just a few ideas I came up with off the top of my head, and I’m not an education expert by any means, but there should be some way to approach the problem.

    At the end of the day, teachers aren’t widgets–they’re not interchangeable, any more than employees in an office are. We shouldn’t treat them as such; instead, we should reward the best so that they continue doing what they’re best at, and attract more people who would be good at it.Report

  4. Avatar Jim says:

    “But that doesn’t stop every other organization on the planet from at least trying to measure employee performance for the purposes of assigning compensation. ”

    What a ridiculous, uniformed claim. No part of the military in any country measures “employee” performance for the purposes of pay; at best they do it to select people for greater and more demanding levels of responsibility. The whorishness of the marketplace is as out of place in the military or for that matter any other part of government as the miltary ethic of duty and self-sacrifice would be in a corporation. Bottom line: the marketplace is not the Unifed Field Theory of society. It’s good for what it’s good for, and useless for everything else.

    Effective teraching, where the student actually learns, is a form of leadership, and leadership is a relationship between leader and led. Teaching is a relationship, like parenting, like the relationship a sergeant or a petty officer has with her subordinates. You don’t pay for performance in the context of a relationship – that’s the difference bewtween marriage and prostitution.Report

    • Avatar Will in reply to Jim says:

      Really, Jim? Military promotions aren’t based on merit?Report

      • Avatar david in reply to Will says:

        Jim is wrong about the military. Fuckups can’t make E-5. Yearly reviews make sure of it. Promotion to non-commissioned officer from the enlisted ranks is not automatic.Report

        • Avatar Jim in reply to david says:

          “Fuckups can’t make E-5. ”

          Yes indeed, and one such kind of a fuck-up is someone who is in it for the money. People who are in it for the money sell drugs – and thought hye may get promoted, that’s not why they get promoted – or get married. Getting married, “to each according to his needs”, is the easiest and fastest way to increase your income in the American military.Report

      • Avatar Jim in reply to Will says:

        “….do it to select people for greater and more demanding levels of responsibility”

        You obviously missed that part. Read more closely next time.Report

        • Avatar david in reply to Jim says:

          Whatever floats your boat dude. I was in the Air force. I made E-4 early to make more money not to get more responsibility. I saw E-4’s sitting around unable to make E-5 for ten years and were involuntarily discharged. Those E-4’s were already doing the job they would have as E-5’s and even for a year at E-6. They were testing and making reviews for MONEY. We were not noblemen serving for honor. No one thought of drugs apparently.

          Teachers don’t have a rank structure or any metric like testing or yearly reviews that matter. This is simple. If the military had this simple seniority system it would be a disaster.Report

          • Avatar Jim in reply to david says:

            “… I made E-4 early to make more money not to get more responsibility. I saw E-4’s sitting around unable to make E-5 for ten years and were involuntarily discharged. Those E-4’s were already doing the job they would have as E-5’s and even for a year at E-6. They were testing and making reviews for MONEY. ”

            So were they getting more as E-4 s or not?

            “We were not noblemen serving for honor.”

            You alredy said Air Force, so you’re repaeting yourself here. Just kidding.

            ” No one thought of drugs apparently.”

            So apparently the money motive was not all that strong after all.Report

  5. Avatar valuethinker says:

    I assume everyone knows what organizational science gurus Robert Sutton and Jeffrey Pfeffer (Stanford dept of industrial engineering, Stanford Graduate School of Business) have said about merit pay?

    It was tried before in California (in the 1920s I think) and it failed.

    The evidence is largely that individual bonuses and merit pay are demotivating, divisive and reduce employee effectiveness.

    America’s most successful steel company, the Southwest Airlines of steel, was a company called Nucor. Under Ken Iverson it grew whilst conventional wisdom was that the US Steel industry was doomed– Bethlehem and US Steel both went broke eventually.

    Nucor had a 25 year record of steadily growing shareholder value in one of the world’s most cyclical industries.

    Iverson had a rule. He did not pay bonuses on individual performance. on nothing less than plant performance, or company performance.

    Pfeffer, Jeffrey and Sutton, Robert

    ‘Hard Facts and Dangerous Half truths and total nonsense: profiting from evidence-based management’
    Pfeffer says that individual bonuses only work for areas of performance that are ‘over determined’. I took this to mean things like beating your own time on the marathon.Report

    • Avatar Will in reply to valuethinker says:

      A collective rewards program might be worth trying. As I’ve written earlier, I think we should encourage local experimentation. However, there are some recent studies that suggest better student performance is correlated with merit pay for teachers:

      http://bear.cba.ufl.edu/kenny/documents/Teach_incent_pap.pdfReport

      • Avatar Jim in reply to Will says:

        One way that it would influence student performance is that it would tend to keep good teachers in teaching, rather than motivating marginal teachers to improve. This is because low salaries in general are keeping good teachers out, or causing good teachers to leave, and merit pay is almost certainly the best way to improve salaries.

        It may sound like I am arguing from both sides; I’m not. It’s just think that the belief that better pay is a motivator in all kinds of work is nothing more than a belief.Report

    • Avatar Ryan in reply to valuethinker says:

      I think the danger with merit pay in this kind of context – which we see in all the “pay for performance” schemes in government these days – is that providing management leeway in evaluating performance allows for favoritism, which really leads to a lot of schism in the workplace. Tying bonuses to performance in a different way (maybe the collective workplace bonus, for instance) might point the way forward.Report

      • Avatar Jim in reply to Ryan says:

        In government especially that is only one of the dangers. The real danger is that it gives managers a tool to construct coteries of loyal underlings. These little empires don’t just set employees against each other; they are in fact separate, unaccoutnable little satrapies unresponsive to direction from higher authorities. They are a mechanism of mutiny.Report

  6. Avatar North says:

    I like merit pay and think that in some form it should be implemented but I think you’re glossing over some of the serious issues that teachers face as compared to say mechanics or accountants. I think a lot of it boils down to children and parents. Parents can be absolute blights on teachers, especially today where so many parents seem to be expecting teachers to fill in the growing gaps that parents are just no longer filling. A badly parented kid is not just a horrible deadweight in of him/herself in the class but will actively drag down the entire classroom in which he/she is placed. Parents are horrifically biased when it comes to their misbehaving children and inevitably blame the teacher. Additionally the teaching and (especially) discipline methods in schools are massively straight jacketed. If teachers are going to have merit pay built into their compensation then I feel that it is imperative that many of these restrictions on their methods must be removed/relaxed and they must be provided with backup for dealing with disruptive students in the form of discipline with some teeth in it. I’m not saying the switch needs to be brought out of retirement but they need something more than the options they currently have available.
    We forget at our peril that children come into this world as little better that screaming excrement spewing little balls of sound and fury. It takes a mountain of parental effort and work from teachers to turn the little animals into someone you’d want to sit next to on a bus. If teacher compensation is going to be made into something based on performance than the teachers must absolutely be given the freedom to perform. Otherwise it’s nothing but parent pleasing pap.Report

  7. Avatar Jaybird says:

    Exactly how do we want the system to be gamed? The way it’s been gamed for the last X years (the conservative approach) or do we want to boldly move forward and game it in a new way (the progressive approach)?Report

  8. Avatar T. Hunt says:

    I don’t reject the idea of merit pay out of hand, but the question is how to implement it.

    Implementing merit pay in the classroom would be similar to implementing the same in a garage if there were other mechanics who came behind you and loosened bolts, removed pieces and misplaced parts ordered for specific jobs. Or who just stood in your way. Then you could be paid on the basis of a successfully rebuilt engine or brake job after random events which might or might not have happened to your particular repair. In that situation, I think merit pay would be scrapped after about a day and a half.

    If teachers had sole control, this might work but in the present situation where the child is returned to the parent at the end of each day, true merit pay as determined by test scores or progress is impossible; or at least severely limited. In no other profession is one’s work actively undone with the regularity that it is in schools. And that’s only taking into account the little things like failure to return weekly reports on the child’s schoolwork signed by the parent, the total abandonment of the child to TV for as many hours as spent in the classroom and a general lack of respect and cooperation from most parents. It doesn’t even begin to address things like Dover, PA, teachers burning crosses in students arms or the Texas school board’s all out assault on science and social studies.

    And how are teachers supposed to be monitored and mentored if the management of the school changes every year or 2? Where one school might have 6 different administrators in 8 years and almost as many vice principles? With a staff of over 60 in many schools and only one principle and one vice principle, the simple factor of time prohibits anything but the most cursory oversight. Many schools require that teachers only be observed for one hour or one period each year. And many principles find it hard to work this into their schedules AND give an in depth evaluation of the teacher’s performance, along with their many other duties. And how do you effectively counsel a teacher when the observation was in November and the grade level stats come out in May?

    Not to mention that teachers are hardly backed up in disciplinary situations anymore and have to deal with idiotic ‘zero-tolerance’ policies.

    I think looking at teaching in light of any sort of business model is a mistake. In no other profession that I can think of does the outcome of work depend on so many factors outside the worker’s control. Teaching is a cooperative effort, one that should involve the teacher, school administrators, the community and, most importantly, the parents. Compare the education a child will receive between the classroom where no one will even volunteer to be a room mother and the classroom where one individual parent spends close to 8 hours per day involved in her child’s education and the contrast is stark.

    Also, we are losing science and math teachers right and left. Why? Because they can make more money and have a shorter, less stressful workday in the private sector. And many of those who are good in math and science never even consider teaching because of the low pay and lousy working conditions. How do we attract people to the profession in the first place if there is little or no visible upside? To think that pay is the only factor is silly but to think that it doesn’t play a role in life decisions is also a bit daft. Sure, we all want to do what we like or feel called to do but we’d all like to live comfortably while doing it.

    Until there is a way to make parents accountable and involve them in a meaningful way in their children’s education and to get the community to value teaching and the rearing of their children as a shared effort, merit pay will never be a factor in the hiring and retaining of teachers.

    T. HuntReport

  9. Avatar Dan Miller says:

    It seems that one problem might be the relatively flat organizational structure of teaching. It’s not like the military, where if you do well you can get continually promoted; or like a corporation, where not only promotions exist but also starting your own firm etc. There’s teaching, being a principal/administrator, and that’s about it (and the skills from one area don’t necessarily lend themselves to another). As a result, merit pay leads to people having the same responsibility but getting paid very different amounts, which might be problematic. I’m not sure if there’s a way around this, although the group-pay notion is interesting.Report

  10. Avatar Kyle says:

    If the current system fails to distinguish between teachers who work 14 hour days, tutor their kids on the side, habitually help students achieve two and three grade level advancements in a year and teachers who leave before the bell and show movies each day and whose kids regularly show little to no improvement.

    Isn’t that more likely to encourage the latter than the former? If you had to choose between that system which doomed thousands of poor, minority, and immigrant children to futures of illiteracy and jail time and a system that placed more pressure on teachers to teach and was less hospitable to bad teachers while increasing retention rates, literacy, and numeracy, which one would you choose?

    As for merit pay itself, value added testing, eliminates the huge annual disparities Freddie is talking about and allowing teachers and administrators to control for individual circumstances (long-term teacher illness, maternity leave, student flux, etc…)

    If bonuses are automatically set – say 15,000 for grade level improvement over a year and 5,000 for every sixth of a grade level over based on class averages, that doesn’t leave much room for administrative meddling.

    If bonuses aren’t zero-sum, it wouldn’t make any sense to actually compete with other teachers or sabotage them. Basically, you’re rewarded for just doing your job and then you get something extra if you happen to do a particularly good job.

    Moreover, performance incentives might actually make more teachers interested in professional development, team teaching, coaching, and improving their own skills. Presently it varies enormously by district but currently the only incentives to try and be a better teacher is if you’re already bad or out of the goodness of your heart. You’ll have to excuse me if I think we can do better.

    Fines and criminal penalties for testing fraud for both administrators and teachers would help deter adult-aided cheating. There’s no reason you couldn’t design a system flexible enough to allow for peer review and in-class observation at a teacher’s request, with caveats to prevent abuse of that system.

    Interestingly enough, some schools and districts have piloted a sort of merit pay for students by paying for good testing scores. Food for thought.Report

    • Avatar Travis in reply to Kyle says:

      How do teachers control for the effects of the parents and the household?

      Merit pay only makes sense where someone has complete or near-complete control over the outcome. That is not remotely true when it comes to teaching in public schools.Report

  11. Avatar TFT says:

    Merit pay will lead teachers to argue over which kids will be in their class because nobody is going to want low-scoring Johnny.

    Merit Pay determined by principal evaluations are problematic because many principals have either never taught, or have never taught the grade they are evaluating, have been out of the classroom for too long, or are just there because they want more money and less screaming.

    Why not let students, parents and teachers do the evaluating? They know the school more intimately than the principal!

    Value -added measures bring the same problems of year-to-year variability; some students are so negatively impacted by their out-of-school circumstances that they ebb and flow with their hunger pangs.

    Do doctors get raises based on outcomes? Police? Firefighters? Meter maids? Government representatives? Presidents of the United States? I don’t think so.

    Freddie, you da man.Report

    • Avatar Kyle in reply to TFT says:

      This is illogical.

      Teachers can argue all they want but it’s not at all common practice for teachers to determine classroom selection. This is also highly speculative criticism of a nonexistent plan/problem on the order of death panel hysteria.

      Principal evaluations – which nobody is really talking about – are bad because you say that they’re wildly incompetent, under-qualified, and have bad motives. The same could be said of teachers but on that issue…silence.

      So let everyone else assess teachers because even though it’s a problem to use student test data or for principals who aren’t qualified teachers to do so, parents and children are somehow better arbiters? Also, somehow parents who aren’t even at school and often have limited contact with teachers know schools more intimately than principals…really?

      So that’s a reason to feed children not scrap value added testing, which is a better if not perfect solution.

      Doctors sort of get raises on outcomes in private practice. Police, firefighters, and meter maids, are promoted based on skill, performance, records, and tests, not just longevity and you’ll find that the unions are even more strongly opposed to anything approaching those measures than they are dual track systems with merit pay.

      Finally, all of the occupations you list don’t have tenure and these days it’s easier and quicker to fire a President you don’t like than an incompetent teacher with tenure.Report

      • Avatar Travis in reply to Kyle says:

        No, your argument is illogical because the analogy fails.

        Those police officers have complete control over the outcome of the tests because THEY TAKE THE TESTS. They learn, they study, how well they do is dependent on their efforts.

        Teachers cannot follow a child home and make sure they do their homework. Teachers cannot keep a child from staring at video games all day instead of doing their reading. Teachers cannot repair a broken home. Teachers cannot make a child learn who has no support from their family in trying to learn.Report

        • Avatar Kyle in reply to Travis says:

          That’s simply not true. First, tests aren’t the only metric used for promotion, someone who tests well but has 15 investigations from IA under their belt isn’t going to be climbing the ladder of success in a hurry.

          I’m not proposing, nor is anyone for that matter, that merit pay constitute the only way to judge performance, the only way to compensate teachers, or that the methods used the only ones to be used in making retention decisions.

          I’d also like to point out that in the wake of Ricci, it’s silly to claim that test takers have complete control over the result of their efforts.

          My point, which is not a grand, sweeping claim is that it doesn’t make sense to say say well we don’t have merit pay for firefighters as a reason to say we shouldn’t have it for teachers without looking at compensation and promotion schemes more broadly. You and TFT are isolating things to compare while ignoring some crucial contextual differences and I stand by calling it an illogical comparison.

          Teachers cannot follow a child home and make sure they do their homework. Teachers cannot keep a child from staring at video games all day instead of doing their reading. Teachers cannot repair a broken home. Teachers cannot make a child learn who has no support from their family in trying to learn.

          Then why have teachers at all? If they can’t overcome difficulty, lets just shut down all the innercity schools because apparently, we’re wasting our time.

          Look, teachers succeed all the time. They succeed in Newark, New Jersey. They succeed in Compton, California. They succeed on reservations. They succeed in Appalachia. What we’re talking about when we’re talking merit pay, is allowing schools and districts to induce more people to become teachers, to continue developing their skills to become better teachers, and to reward harder than average work with better than average pay. Not, the reverse. That’s a low, low bar and the ferocity of opposition to that idea – which just might make a difference – isn’t just unfathomable, it’s unconscionable.Report

          • Avatar TFT in reply to Kyle says:

            My point, which is not a grand, sweeping claim is that it doesn’t make sense to say say well we don’t have merit pay for firefighters as a reason to say we shouldn’t have it for teachers without looking at compensation and promotion schemes more broadly. You and TFT are isolating things to compare while ignoring some crucial contextual differences and I stand by calling it an illogical comparison.

            That makes no sense. Nobody said that because firefighters don’t receive merit pay then nobody else should. That is just silly.

            Merit pay will not ameliorate the problems that create the achievement gap; and it is the gap that we are all up in arms about, and it is the reason for the reform movement.

            Poor student performance is a symptom of something. You seem to have decided that it is a symptom of bad teaching. I think it is a symptom of poverty; schools cannot end poverty.

            So, your solution is to bash teachers and unions into submission in the hope that your top down management style will force teachers to suddenly have the ability to ameliorate all that ails a child. Well, I call bullshit.

            Schools are not like a typical business. Teachers are not widgets, nor are the students. They are people with issues, like all of us, and those issues sometimes take up all your time (the parenting part of teaching). Until you find a way to evaluate teachers while correcting for all the variables, which I say is impossible due to the variability of the variables, you wiil be left to either continue to blame teachers, or, you can blame a society that is more interested in mere survival than their children’s education.

            Education, when you are impoverished, is not a gift. It is another place where your “class” is marginalized.

            Schools need to get back to educating the whole child, not prepping them for a test that many can’t pass because they live in hell.

            Oh, my son, who goes to a school along with many impoverished students, is doing incredibly well. Not because of his teachers, or his school, but because of me and his mother. We have an advantage many of the other families don’t–money. Not much, but we aren’t hungry or shoeless, like millions of others are.Report

            • Avatar Kyle in reply to TFT says:

              “Do doctors get raises based on outcomes? Police? Firefighters? Meter maids? Government representatives? Presidents of the United States? I don’t think so.”

              “Nobody said that because firefighters don’t receive merit pay then nobody else should.”

              Really, because that seems be quite directly what you’re implying. Perhaps you’d care to elaborate on the point of those examples, if it’s something else.

              For the record, I happen to think that many different things contribute to poor educational outcomes and we get nowhere soon by looking for a magic silver bullet or hoping things get less difficult. I think poverty has a strong impact, so does bad teaching, underfunded schools, terrible work environments that don’t support teachers or students, poor management, poor governance, fraud and systemic abuse of power and resources. This is also why I’m not of the mindset of opposing reforms because they’re imperfect, let’s make them better then instead of just waiting for poverty to magically disappear.

              And here’s a question if teaching is so impossible to evaluate, why have any teaching standards at all? Why not let anyone teach if they’re so inclined? Why have tests? Why have classes? If there’s no possible way to tell good teachers from bad, why even credential them? That isn’t a challenge…it’s a question. I’m really curious how something that’s skillful enough to require professional schooling can also be wholly unmeasurable.Report

              • Avatar Travis in reply to Kyle says:

                Because what a student learns is entirely up to that student. You can help them learn. You can present information to them. You can give them opportunities to demonstrate and use that knowledge. You can tell them why it’s so important to learn what they’re learning.

                But ultimately, there are some students who don’t want to learn, who aren’t interested in schooling and are simply there to kill time until they drop out. Compulsory education requires that they sit in front of a teacher for 6 or 8 hours a day. You cannot force them to learn, or even to care.

                There are many more students who would like to learn, who would like to value their education, but whose family and life situations don’t allow them to. Maybe they’re working the midnight shift at McDonald’s to help their mother pay the rent. Maybe their parent(s) ignore them, leave them to sit on their Xbox or their PS3 all night rather than talk to them about their classes or make sure they do their homework.

                You’re proposing that teachers would be paid, in part, based on how many of these uncaring or struggling students randomly happened to be in their class. They have no control over whether a student cares to learn. They have no control over whether a student’s life situation is destroying their ability to learn.

                You’re talking about subjecting teacher pay to the roulette wheel – and the game is rigged to favor rich, white suburbs.Report

              • Avatar Kyle in reply to Travis says:

                So means test merit pay. Merit pay for teachers who teach at schools with X% of kids on free lunch programs or use other markers of socio-economic need.

                I mean this is what I’m talking about.

                I don’t disagree with any of your descriptions of how difficult teaching is and various other difficulties but I still don’t see why we shouldn’t try and test difficult solutions to see if they’ll improve a system that fails so many, so regularly, and continues to do so after decades of cries of educational crises.

                Under the current system. Teacher X works at PS. 134 and makes $56K a year. They have unmotivated students and Teacher X doesn’t really try very hard because they’re demoralized.

                Teacher Y also teaches at P.S. 134. Teacher Y makes $37K a year but hasn’t been demoralized yet so he puts in 60 hours a week and manages to inspire 30% of his class to become motivated and improve from a 3rd grade reading level to a 6th grade reading level.

                Under the current system, that’s what they make and that’s the end of that story. With merit pay, Teacher X still makes the same $56K and Teacher Y gets a $7k bonus at the end of year as recognition for extraordinary effort.

                Problem. Where? I mean if merit pay were the only solution, I think you could legitimately say it’s not a cost-effective solution. However, as part of a solution, I don’t see what the big deal is.

                If the concern is politicized evaluations, let’s address that. If the concern is a competitive work atmosphere, let’s investigate instead of speculating wildly. If the concern is cost-effectiveness, let’s study it.Report

              • Avatar TFT in reply to Kyle says:

                I still don’t see why we shouldn’t try and test difficult solutions to see if they’ll improve a system that fails so many, so regularly

                Your premise is faulty. First, the system isn’t failing anyone. Society is failing and low scores are a symptom.

                Second, what solutions? Merit pay? As a solution to poverty? Or a solution to your false problem? Either way, won’t work.

                The issue is larger than teachers and schools, therefore the answer must be larger than just simply paying teachers a bit more, no? Isn’t that just postponing the inevitable–privatization of education?

                I think you should stop spouting nonsense about how we could conceivably find a way to fairly measure the unmeasurable, and start advocating for those with the least among us; it is they who are in trouble, and they don’t need experiments in productivity that treats education as a business.Report

              • Avatar Travis in reply to Kyle says:

                What about Teacher Z, who puts in the same amount of effort as Teacher Y, and gets no measurable improvement for any number of reasons unrelated to his effort or lack thereof?

                Teacher Z put in just as much effort as Teacher Y and gets not one dime more to show for it. Why? Who knows – maybe his best student was out sick for the test, maybe his students don’t take fill-in-the-bubble standardized tests well, maybe his class is even more difficult to deal with. All he knows is his extra effort and time were valued at zero, whereas Teacher Y’s extra time was worth money.

                You don’t think this is going to instantly lead to dissension, frustration and demoralization?

                And the “means testing” idea, while interesting, is entirely academic — none of the “merit pay” proposals bandied around have actually included such an feature.Report

              • Avatar Kyle in reply to Travis says:

                That sucks but basically under the current regime, no matter how much effort, Teacher Z makes he still makes the exact same amount. Even assuming a poorly designed system for determining bonus pay, at least with bonuses that raises the chance of getting more pay for harder work from 0% to you know a positive number.

                So on one hand you have no chance of getting bonus pay for doing a particularly good job. On the other you have some chance.

                As for the means testing. The objection – you raised – was that merit pay would favor the suburbs. Which might be useful when talking about statewide programs – doesn’t hold as an objection to merit pay systems in cities like Chicago, DC, New York, LA, etc… Though they have some very good public schools they also have perennial teacher shortages and would probably benefit from bonuses that keep teachers from fleeing to the suburbs as well as inducing more people to teach.

                It just seems to me that we’re making the perfect the enemy of the good, here, and while that wouldn’t fly in a discussion on say environmental conservation, climate change, or health care reform, it’s somehow ok in education because the children who lose out aren’t our own and often we don’t know them because none of us live in the Bronx, East LA, or SE DC.Report

              • Avatar Travis in reply to Kyle says:

                It should be noted that I’m a huge fan of the idea that standardized tests are generally meaningless.

                I scored a perfect 800 on the verbal section of my SAT. When it comes to actual classroom performance, I scrape along the 3.0 line at a far-flung flagship land-grant school.

                I’m really, really good at taking standardized tests and filling in bubbles. Other things? Not so much.Report

              • Avatar TFT in reply to Travis says:

                Kyle,

                Reducing motivation to money is bound to keep us right where we are. Not all teachers want or need a couple thousand extra bucks for the privilege of being evaluated in a way that doesn’t (can’t) include all the variables that matter.

                Get your law degree and make millions. Just stay out of this debate.Report

          • Avatar Travis in reply to Kyle says:

            _Then why have teachers at all? If they can’t overcome difficulty, lets just shut down all the innercity schools because apparently, we’re wasting our time._

            That’s a ridiculous strawman.

            What you’re refusing to admit is that the world is not simple and easy. Teachers can have a huge effect on how well a child learns. But all their efforts are likely to go for naught if the children they teach go home to a crack-riddled household, or a barren pantry, or an empty living room with TV, the Internet and video games replacing parental guidance.

            The fact is, poverty-stricken inner-city communities are far more difficult to teach in than affluent suburbs. If teachers who take on the challenge of teaching in poor places are punished in the wallet for doing so, who’s going to want to teach there at all?Report

            • Avatar Kyle in reply to Travis says:

              I’ve refused to admit no such thing. I recognize entirely what you’re saying and I the thing is I agree. I’ve never said we should only have merit pay, or that merit pay is the last, best hope of education reform. What I’ve done here is simply argue that it has a place as part of a package of reforms meant to address different problems that all affect public education in this country.

              I think it makes very little sense at all to say teachers are important but we shouldn’t try to figure out if teachers are doing a good job. Relatedly, I think this nation does a poor job of supporting teachers, particularly those who teach in the worst circumstances, but this wasn’t a post on teaching, this wasn’t a post on education reform writ large, this was a post on merit pay and I happen to agree with Will’s point that opposition to merit pay relies mostly on a bizarre form of professional exceptionalism that remains virtually taboo to examine because to do so is to be anti-teacher.

              The point of my comments hasn’t been to deny anyone money but to argue for paying some teachers more money for doing a good job and to give some direction to how we might measure a good job. None of that precludes paying all teachers more. None of that precludes looking at more comprehensive reforms to help education. None of that is anti-anyone. It’s not even particularly anti-union.Report

      • Avatar TFT in reply to Kyle says:

        Teachers can argue all they want but it’s not at all common practice for teachers to determine classroom selection.

        Yes, it is. Are you a teacher?

        Principal evaluations – which nobody is really talking about

        Well, that’s who evaluates me! And most! What are you talking about?

        It sounds like you enjoy your intelligence a bit more than it deserves!

        Easier to fire a president? Ok.Report

        • Avatar TFT in reply to TFT says:

          The “ok” was sarcastic.Report

        • Avatar Kyle in reply to TFT says:

          I’m willing to admit I might be wrong on the classroom selection but the district where I substitute teach and none of the districts I’ve studied over the years have had anything that looks like incoming class selection. Student reassignment sure. Tracking and teacher selection for particular programs, also sure. However, in general education, heavy-handed teacher selection of incoming classes is just something I’ve never seen. Teachers complaining about lack of that control, being saddled with students halfway through the year, having counselors and administrators abuse the transfer system…sure. Which I admit is an issue of concern.

          Principal evaluations are common now but in the realm of education reform, most merit pay proposals center on testing regimes not principal evaluations…which is what we’re talking about.

          So, maybe I’m misunderstanding you when you say “Merit Pay determined by principal evaluations,” but nothing I’ve read leads me to believe that’s any more a credibly debated solution than so-called death panels. Which isn’t to say there isn’t discussion of a role for principal and peer evaluations but a proposal where that’s the whole of merit-based determinations I haven’t encountered. So, hey if that’s out there, point me in that direction and I’ll probably be right there with you. Sole merit based determinations made by principals is a terrible idea and anything remotely approaching that has been vociferously opposed by the unions for decades for good reason.

          Easier in the sense that Presidents are term limited to 8 years, and can be voted out after the first four. Stories of incompetent tenured teachers who have taken years to get rid of or who successfully kept their jobs while not exactly common, are certainly common enough to warrant skepticism that expedited firings are a realistic option.

          It’s one thing for teachers to bring their experiences to the table/discussions to help craft solutions. It’s necessary to spot problems, craft better solutions, and improve. It’s quite another to deny the same courtesy to other stakeholders in education. Teachers in aggregate are far too willing to blame everything but bad teaching and far too unwilling to support a harder line on maintaining professional standards.Report

          • Avatar TFT in reply to Kyle says:

            And most outside of education are willing to blame teachers without any idea what they are talking about. Where is their courtesy, not to mention their knowledge of what they are talking about?

            You have caved on your 2 big points. You are not even really a stakeholder yet, being childless (I assume). But, then again, we are all stakeholders, right? I mean, the education of our nation’s children is important to everyone, right? Except that we can’t really educate people who aren’t ready.

            We need universal health care, early childhood programs, to raise teacher salaries to attract the best and brightest, lose the putative NCLB sanctions, and bring back PE, music, art, shop, and recess!

            All the energy being put into how we blame everyone but ourselves (society at large) could be better spent participating in government, making our leaders accountable, and voting in our best interest as opposed to the oligarchs best interest.

            I suppose a Yalee hopes to grow into an oligarch?Report

  12. I have to take issue with a couple things in this post.

    > Another argument that would be laughable in any other profession is
    > that teacher achievement is just too hard to quantify (see here or here),
    > making any merit-based pay scale unfair and vulnerable to favoritism.

    As I said over here: http://www.ordinary-gentlemen.com/2009/09/why-not-give-merit-pay-a-shot/comment-page-1/#comment-21177, the fact that “it’s not impossible” != “it’s easy” or even that it’s plausible.

    You cite (http://www2.edtrust.org/NR/rdonlyres/5E91FD34-C8AA-462B-8796-59A6F0101F81/0/AlamedaCOE5906part2.ppt), but I’m going to have to call you out for a “science fail”, Will. That powerpoint presentation references this study:

    Heather Jordan, Robert Mendro, and Dash Weerasinghe, The Effects of Teachers on Longitudinal Student Achievement, 1997. The actual article is here: http://www.dallasisd.org/eval/research/articles/Jordan-Teacher-Effects-on-Longitudinal-Student-Achievement-1997.pdf

    This is a preliminary study done over a decade ago, published as a conference paper. Where’s the final results? The conclusions in that paper don’t spell out the limitations of the method. They don’t talk about generalizability. (here’s a better source from a published journal: http://www.sas.com/govedu/edu/teacher_eval.pdf) They do generally show that teachers can be measured as providing an impact on standardized exams, but that’s only one measurement of teacher quality, and a debated one at that.

    > Leaving aside the fact that it is possible assess teacher quality despite
    > demographic and environmental factors,

    I don’t think you can leave that aside, but if you’re going to give up on that line of reasoning, that’s okay with me 🙂

    > I think anyone who has ever held down a job recognizes that
    > favoritism, personal connections, and outside circumstances
    > play a role in determining workplace success.

    The question is, is this role hugely more important than any measurement methodology? Or are the equal partners, or what?

    > But that doesn’t stop every other organization on the planet from
    > at least trying to measure employee performance for the purposes
    > of assigning compensation.

    This is a wildly out of bounds claim. You must provide a *mountain* of evidence to back up a claim of this magnitude.

    The converse, in my opinion, is *orders of magnitude more likely to be true*: virtually every organization in the world either does not bother to try to measure its workers’ efficacy, or it does so in a completely inconsistent manner with results that have no grounding in any real measurement of efficacy whatsoever, and simply exist to provide a framework of justification for those “favoritism, personal connections, and outside circumstances” that play a role in determining workplace success. With the exception of line manufacturing and sales, where workplace measurement is performed in units per time, this just does not happen.

    I’ve been working since before it was legal for me to do so, albeit in an ad-hoc fashion prior to 15. I’ve worked office jobs, retail, food service, data entry, systems administration, mowed lawns, housecleaning, telecommunications work, automotive service… hell, I spent 3 months working in a slaughterhouse (in no particular chronological order).

    I have *never*, *ever* been systematically evaluated for performance at *any* of those jobs, even the slaughterhouse where production of unit per time was actually measurable.

    When I have been evaluated for performance, the evaluation has existed to put paperwork in an HR folder, and has been completely unrelated to my salary history (in fact, in one instance, the existence of an exemplary evaluation did nothing to save my job when one of those “political favoritism” moments came into play). I know of no small or medium sized business which credibly measures its workers efficacy using any repeatable method (I will grant that there undoubtedly are some, but again you’d have to provide me with a friggin’ Everest sized mountain of evidence to make me believe that this was commonplace). Large corporations certainly do follow measurement processes, but the number of large corporations that do so in anything resembling a scientific manner must be laughably small; Likart scale evaluations mean nothing unless they’re properly constructed, and every damn evaluation form I’ve ever seen anywhere is built in such a useless fashion that it measures nothing at all.

    All of that is of course anecdotal, and doesn’t qualify as real hard evidence. But I’m not making this claim “everybody measures performance”, Will is.

    Show me evidence to support this, or I reject it out of hand as unsupported.Report

    • Avatar Will in reply to Pat Cahalan says:

      I think you need to distinguish between systematically measuring performance and measuring performance, period. I’m not suggesting that *every* teacher needs to be evaluated using the same methodology; simply that schools experiment with different methods (some subjective, others more statistically rigorous) to determine teacher performance. My broader point is that every other business/government agency makes at least some effort to figure out how well employees are performing, and I’m not sure why schools don’t take a similar approach.Report

      • Avatar Will in reply to Will says:

        Although you’re right that I should have used more qualifiers when describing that study on teacher performance.Report

      • Avatar Pat Cahalan in reply to Will says:

        > I think you need to distinguish between systematically
        > measuring performance and measuring performance,
        > period.

        Er… why? To me, bad methods of measurement are actually *worse* than no method at all, they produce false results (generally, false results that are easily manipulated to justify those political and social factors).

        If anything, if you were counting up {organizations that don’t measure}, {organizations that measure badly}, and {organizations that measure well}, an actual worthwhile analysis of measurement would be something like…

        If measurement is a proper method for improving results, {organizations that measure well}*{the cost of measurement} > N*({organizations that measure badly}*{the cost of measurement})+{organizations that don’t measure at all}, for some N.

        … particularly since you’re proposing that we just cut loose with “go measure” and let several different methods of measurement crop up in the educational venue, you’re pretty much guaranteed that you’re going to get some cruddy methods of measurement that will be misused.

        You can’t just say “yay, we might get a good method of measurement out of it”, because you’re ignoring the cost of doing it in the first place.

        > My broader point is that every other business/government
        > agency makes at least some effort to figure out how well
        > employees are performing, and I’m not sure why schools
        > don’t take a similar approach.

        I still don’t believe that this is true, even if you toss “crappy measurement” in the bucket with good metrics. Again, it’s anecdotal, but even in places where I’ve had reviews, they have not come close to equaling the number of places where I wasn’t. I just don’t buy that most places do this. I will gladly revise this if you can defend it.

        You might want to start here: http://www.amazon.com/Abolishing-Performance-Appraisals-Backfire-Instead/dp/1576752003/

        In addition, if you’re not disputing my assertion that by and large, across the board the vast majority of these measurement methods is useless… then why does it matter that everyone else does it?

        Wouldn’t it make more sense for schools to say, “Gee it looks like almost everybody else either doesn’t do this, or they do it really, really badly, so it’s largely a complete and utter waste of time?”

        Like I said, I do think it’s possible to measure teacher performance in different ways. I don’t think it’s easy (and by extension, doing it properly domain-wide would be expensive), and I think there is a huge problem that most of the even-somewhat-acceptable methods of measurement that I’ve seen (such as the one cited here) don’t necessarily measure what we actually want teachers to do, particularly in multiple disciplines.

        People like to measure stuff. It makes them feel like their decision making is based entirely upon a rational analysis. The fact of the matter is that unless you’re using a good measurement, you’re getting results that are worse than useless, because they give you bad results you can interpret any which way you like, *and* you have to pay to do all the measurement (which again is all overhead).

        People talk about making various things more measurable, it’s a big push among people who talk about government reform (generally conservatives) just like people who like to make things more regulated (generally liberals). From a meta standpoint, both groups are trying to accomplish the same thing.

        “I don’t believe that (this system) produces (optimal results), ergo I want to audit the process so that I get the results I want”. In the case of conservatives, (this system) is usually the government and (optimal results) is usually “less waste”. In the case of liberals (this system) is usually business, and (optimal results) is usually “fewer externalities”. Both groups like to say, “Whoo, this is going to solve my problem!”

        In both cases, they generally are discounting (a) the cost of what they’re doing (b) the efficacy of what they’re doing and (c) the unwanted consequences of what they’re doing.

        I’m not saying that there isn’t a time and place to audit things. Some industries need regulation, some government processes need less waste. But the more complex the thing is that you’re trying to measure (and teacher efficacy is right up there with “stuff that’s REALLY hard to measure”) the less likely it is that you’re going to be doing anything real and useful.

        In the particular case of teaching, *it’s completely a moot point*, because we can’t fire the bad teachers and replace them with good ones. There aren’t enough teachers (good or bad) to go around, period.

        If you want good teachers, you’re going to have to pay through the nose to get them; in most places you just don’t have a competitive salary structure for teaching, at all. The demand is so high that you’ll have to really hike up the compensation to get anywhere near what you need.

        Before we go auditing (using a difficult measurement), trying to find out who the bad teachers are, let’s figure out what we would do to replace them.Report

Leave a Reply

Your email address will not be published. Required fields are marked *