Why not give merit pay a shot?



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

Related Post Roulette

20 Responses

  1. “Forty years of psychological research demonstrates that when someone is faced with a complex, creative task — like teaching — money is an ineffective motivational tool, and may even delay progress. Professionals engaged in creative work are more likely to be motivated by autonomy, and by the feeling that they are part of a larger, socially important enterprise.”


    I haven’t had a laugh that good in awhile. I’ll read this quote to my employees the next time they ask for a raise.Report

  2. Avatar North says:

    Gotta agree. Merit pay makes sense. Inasmuch as there should be teachers unions for public sector teachers they should be focussing on making sure the criteria for the merit pay is fair and objective. Not subject to the whims of the absolutely madhouse parents who plague teachers.Report

  3. Avatar E.D. Kain says:

    Merit pay may very well work wonders. I know lots of people who won’t become teachers because they know they won’t be able to pay the bills on a teacher’s salary. I would have to take a significant pay cut to go into teaching even though it’s something I think I’d like to do.Report

  4. Avatar Pat Cahalan says:

    Kevin Carey’s piece (linked at the end) has a good point, but your link to the Michelle Rhee story doesn’t exactly bolster the idea that teacher quality is measurable. In fact, from that actual article:

    > “[The] data largely confirm that idea [that family background is the leading
    > cause of student performance], demonstrating clearly that the best
    > predictors of a school’s achievement scores are the race and wealth of its
    > student body. A public school that enrolls mostly well-off white kids has
    > a 1 in 4 chance of earning consistently high test scores . . . a school with
    > mostly poor minority kids has a 1 in 300 chance,” Tough reported in the
    > New York Times after examining a host of studies”

    Paying teachers better will indeed keep the best teachers in the system (I know a number of great educators who left teaching because they simply couldn’t afford to raise their own kids in their own home in the same neighborhoods where they were teaching). Measuring teaching skill itself is a wicked problem, though… it requires a lot of qualitative measurements in situ by experts in education, and it doesn’t scale very well.

    You’d have to be very, very careful to make sure that any merit-pay based program actually provided incentives for people to continue to work at those bad schools, or you’d quickly see all the good teachers go to the non-disadvantaged schools so that they could be paid competitively, and the educational gap would just widen.Report

    • Avatar Will in reply to Pat Cahalan says:

      Pat Calahan –

      Agree that teacher quality is difficult to measure, but it’s definitely not impossible, and making the effort is certainly better than sticking with the status quo. From the HuffPo piece I link to:

      “Rhee’s argument in favor of aggressive weeding out of teachers who fail to achieve performance benchmarks has received strong support from studies conducted for the Consortium for Research on Educational Accountability and Teacher Evaluation (CREATE) of students in Texas and Tennessee.

      The CREATE studies showed remarkable differences in student achievement and performance depending on the quality of the teachers. For example, the first graph shows the percentile math ranking of students with ‘effective’ teachers rising 21 points, from 55 to 76, from the start of the third grade to the end of the fifth grade. In contrast, those with ‘poor’ or ‘ineffective’ teachers saw their percentile rankings drop by 20 points, from 57 to 27.”

      Moreover, it is possible to control for adverse demographic factors when evaluating teacher performance:


      • Avatar Pat Cahalan in reply to Will says:

        It may not be impossible, but it’s certainly a much, much, much larger problem than many people think it is. Will’s post here (http://www.ordinary-gentlemen.com/2009/09/merit-pay-continued/) doesn’t do justice to the problem.

        From the research article you cite:

        > The most credible way to identify teacher effects
        > is to regress test scores on teacher dummy variables
        > when teachers are observed with many classrooms
        > (netting out idiosyncratic variation in classroom
        > performance) and controlling for variation in student
        > characteristics and other classroom specific variables.
        > The data I collected for this study links teachers with
        > their students for a period of up to twelve years, and
        > contains up to five years of annual test scores
        > for all elementary students in a number of schools

        While I agree that it is in fact possible to measure teacher quality, this particular method has a number of severe generalizability problems. It’s time consuming, for one. You can’t measure teachers that haven’t been in the system long enough to provide the appropriate measurements.

        I understand the desire to measure teacher performance, and I’m all for it conceptually. It’s going to cost money, which we would otherwise be spending on teaching (I’m a security wonk; audit is always expensive, and it’s almost always entirely overhead). If we go with a method like the above, it’s going to have edge cases (young teachers, new teachers). Implementing it has an unintended consequence: verifiable (using the method) “average” teachers will probably be preferable to un-verifiable teachers who might be of higher quality.

        Since the standard anti-teacher’s union argument is that “mediocre and poor teachers with seniority (and, correspondingly higher salaries) stay in the system and younger, possibly better teachers don’t get in”, it’s notable that this method is of unknown value to resolve that particular problem, since it will only remove poor teachers… and I suspect (admittedly without evidence) that the number of mediocre teachers vastly outnumbers the number of actual bad ones.

        It’s also completely useless to measure teacher quality in subject matters that aren’t subject to standardized testing (art, music) and would need to be evaluated particularly for those subjects that are tested largely on qualitative rather than quantitative measurements (literature, for example).

        Also from the article you cite: “However, efforts to improve the quality of public school teachers face some difficult hurdles, the most daunting of which is the growing shortage of teachers. Hussar (1998) estimated the demand for newly hired teachers between 1998 and 2008 at 2.4 million–a staggering figure, given that there were only about 2.8 million teachers in the U.S. during the 1999-2000 school year”.

        This is obviously dated, but as I recall it’s more or less accurate (at least in California, hiring of non-certified teachers is pretty commonplace because there simply aren’t enough certified teachers to go around).

        Measuring teacher quality is largely a pointless exercise if you have no options to replace the underperforming teachers. If you’re short 2 million teachers, getting rid of 1 million teachers you *have* isn’t exactly a great solution. Not to say that we shouldn’t measure the efficacy of teachers, but this seems to be a solution to a secondary problem, no?Report

        • Avatar Will in reply to Pat Cahalan says:

          Pat Calahan –

          I agree that there are a lot of conceptual and structural barriers to adopting merits pay; that’s why I’m in favor of experimentation on the state and local level. Let a hundred merit pay-inspired flowers bloom, I say.

          As for your concerns about the shortage of teachers, I think that boosting teacher pay would do a lot to make up for that shortfall.Report

  5. Avatar Scott says:

    The teachers unions will never give up any of their control to allow a program such as this and anyone who thinks so is naive.Report

  6. Avatar Kyle says:

    You know, I’ve been trying to wrap my head around the quote from Jackson for a bit now because quite frankly the tepid support for school based bonuses is odd.

    Teachers have long been wary of giving principals and administrators sticks/carrots to crack down on teachers or to professionally punish or reward people for workplace politics. Those concerns are pretty central to fights for tenure and rigorous due process for teachers in danger of being fired. So the idea of mitigating the concerns of individual accountability for bonuses/merit pay by simply making them school based seems more likely to foster enmity and damage cohesion than the original idea Jackson’s expressing skepticism about.

    Does anybody else find that odd or am I missing something here ?

    Either way, I’m woefully unsympathetic to the cottage industry of education reform opponents whose arguments boil down to suspect data and cries of how “difficult” or “hard” or “impossible” it is to do difficult things or how “unreasonable” it is to hold people to standards that look gold-plated compared to other industries.

    What’s really hard is finding a job, not going to jail, and leading a life where you can take advantage of the opportunities around you when you can’t read because you weren’t successfully taught how in 13 years of schooling.

    What’s really unreasonable is talking about talking about accountability for job performance as if it were a bad thing. This country (and I suspect many parents, teachers, etc..) would be livid if pediatricians blocked health care reform while habitually protecting malpracticing doctors who left children injured or crippled for life.Report

  7. Avatar Mark says:

    I think the parallel non-tenure/higher-paid track is a better idea than having administrators “measure” performance and assign bonuses based on the results. The problem with teaching is that it is seen as a profession of last resort and attracts weaker college students as a result. Programs like Teach for America have shown that there are few barriers to entry in teaching, which is why established teachers rally around seniority and oppose merit pay in a way that might seem more germane to supermarket baggers than creative professionals.

    But I do also see their point: on-the-job performance evaluation is so colored by your boss’ prejudices that it’s impossible to figure out anything other than who the absolute worst teachers are. And merit pay could easily turn into a version of the Gold Gloves in baseball, which are so reputational that it’s easier to win ten years in a row than it is to win once. Not to mention the difficulty of deciding whether a teacher should be rewarded for his or her impact on the performance of honors students, average students or special-needs ones. And, to make matters worse, it would penalize teachers like my high school chemistry teacher, who told us that what we got out of his class was up to us – fostering self-reliance for college, rather than hand-holding.

    Bottom line: we need to pay (some? many?) new teachers more so that we attract smarter teachers. Tenure seems irrelevant (especially when all my friends who were new teachers got laid off every year.) Old teachers can stay in the old system if they want to.Report

  8. Avatar Steve says:

    I am a relatively new teacher (5th year), and I have been the union building rep for two years now. Whenever merit pay comes up there is a predictable reflexive urge to hind behind the “but how can we fairly measure performance” stance. No surprises here. Personally, however, I think the greater fear among some teachers is that eventually our profession will be carved into a hierarchical system where core subject teachers are compensated more than say, the art teacher or the Spanish teacher – the same teachers that start applying for bar tending jobs whenever talks of program cuts come up. For them, the logical extension of merit-pay is that school districts and state departments of education would marginalize their contribution and content areas (as they already do by not including them in state standards testing) by creating compensation structures that reward teachers in areas with a perceived higher rate of return for society. Even though an argument could be made for such a system, I would not want to work in a building wherein teachers who put in the same amount of hours and dedication and make the same lasting connections with students get paid using different structures. One of the main motivations for me to stay at the school where I currently teach is the dynamic of cooperation and professionalism between the staff. I am not sure it would exist in an environment wherein certain teachers got paid more than others based on their content area. What if the English teacher were also the student government teacher? Would he or she get a two-tiered contract because their aren’t any state standards from which to evaluate the growth of students in leadership? In my state I support efforts that are currently being made to explore merit pay, but I think everyone involved needs to buckle up for what is sure to be a bumpy ride.Report

    • Avatar Will in reply to Steve says:

      Steve –

      Interesting point, and I can see why you’d be concerned. Here’s my counter-argument: shouldn’t we place more value in educators who teach “core” subjects like math and science? I’m not sure merit pay would automatically make certain teachers second-class citizens, but I wonder if there’s something to the idea that some subjects deserve more and better paid educators.Report

      • Avatar Steve in reply to Will says:

        Well, I think that most education policy departments would probably spontaneously combust in outrage if such an idea were floated seriously. I happen to have personal knowledge of this because I was practically laughed out of a grad seminar on ed. policy and the developing world when I referenced scholarship supporting the notion that countries with scarce resources subsidize university math and science education at a relatively higher rate than the humanities. They seemed to be concerned, in some ways justifiably, that the end result would be an automaton society with no one left to fight for issues of the conveniently ubiquitous concept of “social justice.” The practical problem, I believe, is: who will be the bureaucrats that will decide the criteria for selecting these core content areas. I happen to teach an international relations course to juniors and seniors that I feel provides them with important skills and perspectives. It certainly is not a traditionally core subject, but may serve as a foundation for students who go on to pursue similar studies in college while all the while honing traditional reading, writing, and analytical skills. Should such a course yield to Math and Science because those who would teach the course decide to be Ph.D. policy wonks where the money is? It is a good question for debate, but I think that there is a powerful argument to be made, probably by those same grad school peers of mine, that math and science education while important should not be taught in a vacuum void of social context and perspective.Report

    • Avatar Kyle in reply to Steve says:

      As what may come to a shock to readers of my comments in Will’s second post, I actually think this is very good point and, Steve, I’m glad you made it.

      I think there’s a lot to be said for support for and the greater inclusion of the more elective subjects. I think certainly, when looking at merit pay, the focus is on performance incentives for teachers at under-performing sites and as inducements to teach in places with teaching shortages (urban and rural environments). That discussion, however important, is incomplete in the ways that you highlight.

      I think the uneasy point where we are at a societal level is that with renewed focus on the achievement gap and the pronounced failures of public education, it’s hard to say well let’s spend 1 million dollars on music teacher salaries in Westchester when we could spend that million on English and Math teachers in the Bronx. Not least of which because without basic literacy and numeracy skills, other subjects not to mention opportunities in life are all but inaccessible.

      I guess I’m glad to see cautious exploration of how this idea might work.

      I’m curious to know – I’ve heard comments on both sides of the fence on this – what your thoughts are about how the dynamic of cooperation and professionalism is affected by teachers who aren’t cooperative or for that matter professional. My experience has shown that the site work environment has a huge effect on how teachers see themselves professionally in relation to other teachers.Report

      • Avatar Steve in reply to Kyle says:

        I currently work at a charter school that shares a campus (not building) with a traditional comprehensive high school where I worked previously. The work environments that I personally experienced were very different. I think it would be very hard to over-state the importance of having a unified, collegial working environment like we have at our charter school, and I think benefits absolutely trickle down to the students. In my experience, if you can establish that culture of collaboration and mutual support then an occasional bad apple either conforms or gets out of the way. We were lucky, because a committed administrator and a core group of teachers who shared the same vision started the school from scratch. I have been in other situations where faculty members with personal agendas make life very difficult and division is hard to deal with effectively if archaic constraints on personal mobility are in place.
        I guess my main point is that I place a very high premium on having a work place where I want to show up everyday and colleagues that I can count on for support. I am 100% convinced it happens to be good for kids as well.Report