Saturday, May 17, 2008


Interesting juxtaposition of Ed Week articles this week—a commentary from James Starkey, 40-year classroom veteran from Colorado, Please Don’t Do Me Any Favors, and an article in the news section, Principals’ Group Seeks Influence on Incentive Pay. Subtext question in both pieces: What motivates people to improve their performance?

Starkey’s observations about hard-working and undervalued teachers rang true for me. Like Starkey and his wife, I have been prejudged and patronized dozens of times by those who pigeonhole teachers as noble, but benevolent and underpaid missionaries to the unschooled young. His points about the recurring churn in K-12 standards development, and the insulting assumption that mere teachers would have no idea what critical content and solid achievement markers look like in their subject disciplines, are spot-on. But then he says,

“ProComp [Denver’s performance pay plan] and similar programs arise from the assumption that teachers could work harder, and I simply reject that notion.”

Meanwhile, over at the National Association of Secondary School Principals, Executive Director Gerald Tirozzi declares that the NASSP doesn’t endorse performance pay for school administrators, then disingenuously offers a carefully constructed template for how to pay principals for their, umm, performance. Tirozzi does note, correctly, that it’s wise, when an issue is politically hot, for an organization to get out ahead of the rolling train and develop some policy recommendations of its own. I have no idea if the list of administrative performance indicators that the NASSP offers is complete and/or valid, but it’s certainly a good starting point for discussion:

The organization suggests looking at other variables, such as graduation rates and promotion rates, student enrollment in rigorous coursework like that developed by the Advanced Placement and International Baccalaureate programs, college-attendance rates, school climate data, parent-participation data, and teacher-retention and -transfer rates.

It’s really too bad that the term “merit pay” has been so thoroughly co-opted by the aggressive conglomerate of those who believe that testing and punishment are the only effective policy levers for improving schools. It’s impossible to start a discussion on how we pay educators in America without two knee-jerk notions popping up: a) merit pay means paying teachers or principals more for raising test scores and b) educators should be offended by the idea that increasing compensation will lead to better results. Indeed—the first on-line commenter on the NASSP article says she’s insulted by the implication that she could work any harder than she does.

Directly linking standardized test scores to compensation bonuses is an intellectually dishonest, even dangerous, idea with fairly predictable “unintended” consequences—some teachers will behave just as B.F. Skinner might predict, conditioned by rewards, and others will struggle with their personal moral compass around what it means be a measurably good teacher. Meritorious teaching is a real thing, however—most parents could give you a quick sketch of its characteristics and benchmarks, without needing a standard deviation or statistic. And meritorious teaching should be recognized, deconstructed, modeled, emulated, studied, nurtured—and rewarded. The same goes for exemplary school leadership.

I understand and appreciate the origins of the single-salary schedule for teachers, a half-century ago: equal pay for equivalent work, providing strong incentives for teachers to complete academic degrees and strengthen their professional educational attainment. And I don’t buy the specious argument that educational institutions should follow a “business model” and pay teachers based on productivity. Teaching is not piece work, and there is also very little evidence that the highest-paid employees in most businesses are necessarily the most efficient or resourceful. Businesses don’t have the compensation-incentive problem knocked, either.

The difficulty here is that lockstep teacher pay is no longer moving us toward best use of available resources to reach our educational goals. The single salary schedule penalizes exceptional talent, innovation and effort, and rewards staying put and accruing credits.

There is nothing morally wrong with figuring out what meritorious teaching looks like, then actively pursuing and rewarding it. I would respectfully suggest that James Starkey has the causal direction of pay for performance backwards. We should not offer performance pay to teachers under the assumption that they will work harder for more money. We should offer performance pay to outstanding teachers and educational leaders because they deserve it.


RDT said...

We should not offer performance pay to teachers under the assumption that they will work harder for more money. We should offer performance pay to outstanding teachers and educational leaders because they deserve it.

Actually, economist's daughter that I am, I'd argue that the best reason for rewarding outstanding teachers financially is that it makes teaching a more attractive career for outstanding teachers. The fact that we're short of outstanding teachers suggests we might not be paying enough.

Nancy Flanagan said...

Thanks for reading and commenting, Rachel--and for stipulating that it's outstanding teaching we're short of, not potentially good teachers to fill our classrooms. I am interested in the quality of a teacher's performance once on the job--and that's not guaranteed by a higher starting salary.

I live in a state where teachers are traditionally well-paid, perennially among the top five, although we have slipped a bit in an economic downturn. Although I know there is a general correlation between higher pay and higher student achievement data, mainly in collective bargaining states, I am hard-pressed to describe the teaching force in my state as also representing the top five in teaching talent or motivation. Nor do teachers in my state ever say "well, gee--we're so well-paid, relatively, that we'd better do some great work to demonstrate our determination to earn or maintain those higher salaries."

While you might increase the talent pool by offering higher salaries, once people are drawn into teaching, rewards should be based on strong performance. I think it's a chicken-egg dilemma: the outstanding teaching comes first, from the intrinsic desire to be a great practitioner and servant (plus skills, knowledge, reflection and effort). If the money comes first, as incentive to raise performance, it isn't linked to that internal desire for excellence and service.

And, yeah, I do agree that figuring out what constitutes outstanding performance, excluding standardized test scores, is the proverbial devil in the details. Economists use the data sets they have on hand--standardized test scores. It should be up to educators to envision and flesh out rich data portraits of teacher performance. The NASSP took a stab at it. I found it interesting that they did not suggest that data to assess principal performance include evaluations by staff.

j m holland said...

I would argue from an economic standpoint that we should always reward what we want more of and more great teachers seems obvious.

Where i find a bone in the soup is that as soon as I say merit based pay many policy folks say value-added measures. I think we really need to rethink why we want to base teachers' pay on what students do instead of on what teachers do. I am fully aware and comfortable with high standards. That should be one of the things I am held accountable for but I don't feel like because I do a good job and Nancy next door does a stellar job, I shouldn't be rewarded for meeting standards. One of the problems I discovered recently with value-added measures is the problem of fade. This is when a kid has a great teacher in 3rd grade then two average teachers. That kid's learning that the 3rd grade teacher would be rewarded for has not been maintained. Then what do we do?

Let the madness stop. Here is an idea, just pay everybody who teaches fair to fabulous better and send the rest back to teacher school.

Finally, consider basing pay on "what teachers do with kids" (Pianta, EdWeek 2008) instead of the results of what teachers do with kids. We need to think about Karl Popper's theory of falsification, "How could this child's achievement not be due to how well the teacher did or didn't teach?" How many factors are there that could affect how we answer that question? Could we ever really "know"? This, to me, is the fundamental problem of basing teachers rewards on student testing. I can't see a valid causal relationship. What can be seen, with a number of tools,(NBPTS, CLASS, etc.) is what a teacher "knows and is able to do" in a classroom with kids.

Nancy Flanagan said...

Hey, John.

You raise some interesting points about measurement and causality--and I freely admit that the answers to those questions are incredibly complex and nuanced. I don't have answers, and even if I thought I did, I would stress that my firm belief is that these issues should be determined locally. Measuring teacher performance should be done in context and be needs-based, I think.

My point was that even though there are all kinds of confounding variables, and hard questions around measurement models, it doesn't mean we should be happy with the inadequate model we have. There is this subtle pressure on teachers to accept a compensation model that rewards the status quo, with undertones of martyrdom: I didn't become a teacher to make money, I'm only in it for the kids, who really need me, blah blah. This leaves teachers with only one response when a more entrepreneurial way of paying teachers is suggested: we're doing this because it's vital work--we don't need to be paid well to do a good job.

This is known as valuing your heroic self-image above your own economic interests. And it doesn't lead to any kind of professionalism or self-determination for the teaching profession--or, as Rachel points out, to a high-quality pool of teaching applicants. I agree that the emergence of value-added modeling has changed the conversation. I'm not sure the best strategy for teachers is resistance. In fact, what bothers me most is that teachers are often not part of the conversation in determining what does matter in teacher performance.

Thanks for reading and posting.

Anonymous said...

While most people have no problem with compensating outstanding teachers, there is a lot of ambiguity about what comstitutes "outstanding" and how to measure "outstanding".

Test scores, even when adjusted for value added, only measure the short term retention of content. Observation by one person is total dependent on that individual's concept of good instruction. If school is supportive and creative that really enough without rigor and relevance? We still seem to be struggling with defining "outstanding teaching practice" and until we do, it is unlikely that there will be limited support of performance based pay.

While some process such as NBPTS have done a good job of defining standards and determining if teachers can teach well, stakeholders want assurance that the daily practice of teachers aligns with what "they know and are able to do." Until we develop more effective and rigorous teacher evaluation methodologies, this will be difficult to impossible to determine. A compbination of tools such as peer observation, student and parent input, and portfolios that could do this. What I wonder is if we have the committment to develop and employ these kinds of tools.

Teaching has a long tradition of practicing in isolation. Single salary scale needs to change if we are to attract and keep good teachers and it's not just the money, it's the mindset of professionalism that acknowledges quality work. Would teachers be willing to submit to a more rigorous assessment of practice? Could we make the transition by allowing teachers to opt into a "outstanding practice" assessment that held the potential for additional compensation?

Nancy Flanagan said...

Susan, I agree with absolutely everything you said about problems in measuring teacher performance. It would be a tremendous challenge to devise fair and workable models of assessing teacher performance.

However--until we start acknowledging that some teachers do better work than others, we're stuck with a lockstep compensation plan that sometimes functions as an entitlement program for a subset of teachers. When you work in a publicly funded and monitored enterprise, those low-performing teachers become the flip side of your comment that most people have no problem with rewarding outstanding teaching --the public says they're happy to pay the good ones more, but they want the unproductive teachers gone.

If we are going to measure teacher performance, I would rather muck around trying to find a workable model, than spend the millions of dollars on statewide VAM data-link systems that some policymakers have now decided are The Definitive Answer to improving teaching.

Thanks for commenting.

philip said...

Why can't we "measure" good teaching by asking teachers, their peers, their students, their administrators, and invested members of the community to do the measuring?

I turn in a dossier every year until I earn (hopefully) tenure. I don't see why teachers couldn't do the same thing...

Nancy Flanagan said...

Actually, I think observed evidence from the folks you mention--peers, students, community, parents-- should be part of an evaluation portfolio on all active practitioners, tenured or not. Teachers often worry about the single-shot administrative review, fearing that a capricious (or jealous, or threatened) principal would give an excellent teacher a kiss-of-death review. Having a panel of qualified reviewers and some kind of rubric would provide much better data--not only to determine whether the teacher is performing well and should be compensated as such, but to inform the teacher in question about performance concerns from a wide audience. Kind of an adaptation of the "360" performance review that businesses do.

Again, there would be time and expense involved in developing comprehensive reviews--but there is a lot of expense involved in building those VAM data bases and analysis tools. And they don't provide any constructive feedback, with the possible exception of teaching policy-makers what a "quintile" is.

RDT said...

Nancy Flanagan wrote:
I'm not sure the best strategy for teachers is resistance. In fact, what bothers me most is that teachers are often not part of the conversation in determining what does matter in teacher performance.

I'd agree with that. However, my impression is that one of the major reasons for teacher resistance is a conviction that if merit pay comes, teachers will not have input into either what is important, or who is performing well.

My cynical side also wonders if the increased interest in merit pay that has come with value-added models has something to do with the fact that it would allow teachers to be evaluated in an entirely top-down manner.

j m holland said...

"This is known as valuing your heroic self-image above your own economic interests."

I totally agree with your post Nancy.

And the comment above is totally on the mark too.

Mind if I quote you the next time I talk to a policy maker?

Nancy Flanagan said...

Rachel, I think you've gone right to the nub of the argument. It isn't about methodology, in the end--it's about power. If teachers were to structure complex evaluation systems and submit themselves voluntarily to regular assessment of their work, it moves them much closer toward full professionalism. A profession with three million practitioners would be a politically powerful force.

I actually think that much of the pushback against National Board Certification comes from the worry that a rapid increase in the number of teachers who have demonstrated accomplished practice via a "scientifically based" assessment would give them undue power in educational decision-making. So, researchers and policymakers go looking for holes in the evaluation system. There are holes in every evaluation system in the social sciences, of course.

I actually think that--over time--teachers would be harder on themselves than administrators or ed bureaucrats in setting admission standards and practice standards. A lot of the dynamics of the current organizational power structure depend on teachers being not only compliant, but taught to value camaraderie over excellence.

Thanks for some great thinking.

Nancy Flanagan said...

Thanks, John.

And you don't have to quote me. You're very articulate, all by yourself. (smiling)