Obama's school agenda - Merit pay v 'performance-based pay'

Obama gave a major education speech at the U.S. Hispanic Chamber of Commerce yesterday. The money quote:

For decades, Washington has been trapped in the same stale debates that have paralyzed progress and perpetuated our educational decline. ... It's more money versus more reform, vouchers versus the status quo. There has been partisanship and petty bickering, but little recognition that we need to move beyond the worn fights of the 20th century if we are going to succeed in the 21st Century. Well, the time for finger-pointing is over.

Been much buzz about this talk -- the main points of which are at the bottom of this post, or visit the White House for the speech itself -- but on a key point, regarding teacher pay, much of the coverage was unintentionally misleading. The headline in my local paper, for instance, and in many others as well, was "Obama expresses support for merit pay," and the AP story is getting heds like "Teacher merit pay essential, Obama says."

Actually he didn't say that, and this is a mistake that's going to frighten a lot of teachers even more than what he did say; for what Obama came out for was performance-based pay, which -- functionally and especially politically -- is a different animal and a looser, still-fluid concept that might or might not include merit pay, and which offers a more comprehensive but arguably more compatible challenge to current ties between performance and pay. It's important not to mix them up.

Performance-based pay is the looser of the two terms, and it essentially means that you'll pay teachers more if they do their jobs particularly well, and you pay them even more if they do well the hardest jobs, like teaching in tough schools, subjects, or other conditions. The idea can also encompass -- and to some people implies -- the sort of thing Michelle Rhee is proposing in Washington, where teachers can opt for a much higher pay scale (into the six figures) if they take an untenured track that makes them vulnerable to job loss if they perform poorly. (Tenure, which is virtually permanent, is virtually automatic in most school systems after 1 to 3 years, and is often won as long as a teacher doesn't paw any students or buy them drinks. A few years ago my own district gave tenure to one of the worst math teachers we've ever had.) This performance might be measured by a number of factors such as student scores; peer, parent, and outside-expert evaluation; certification by third-party outfits like the National Board for Professional Teaching Standards; advanced study; and so on - a list to be determined, in a sense. Performance-based pay is a work still in progress, in short, a set of principles with emerging guidelines for practice.

Merit pay, on the other hand -- which might or might not be part of a performance-based pay system -- is understood to mean a system that moves teachers' pay up or possibly down with their students' scores on standardized tests. This can be problematic, as it can punish teachers that teach in the toughest schools or take on the most challenging students and can create a "teach to the test" approach. It's generally unpopular with teachers and teacher unions.

"Performance-based pay" is what Obama's proposing, along lines still being worked out. But many stories conveyed the impression he's proposing the more tainted and limited merit-pay approach. This has already created some confusion. Witness these tweets from this morning, for instance:

CraigKarasky: Great move by Barack to propose merit pay for teachers and fight the status quo/teacher's union. http://tiny.cc/lGjC1

jaahsten: merit pay for teachers isn't going to work. try again obama.

justwright: and how come the teachers unions are suddenly warming up to the idea? Merit-based pay has been suggested before.

But Obama's not talking about just merit pay, he's talking about a more flexible set of incentives and rewards based on a broader set of criteria and measurements. "What you want to do is really identify the best and brightest by a range of metrics, including student achievement," Education Secretary Arne Duncan said in an AP interview.

This is an extremely crucial and extremely touchy aspect of education reform. Many teachers object to the idea that you can effectively measure teacher performance. And indeed the data on that is a little thin, primarily -- just as in medicine -- because we lack national standards and measures that would allow us to see what works and who's doing things well. But people are starting to figure out how to do this, and the weight of the evidence so far suggests that you can accurately measure teacher performance, especially you use a combination of metrics, including classroom observation, to do so. That approach gets you that good old converging evidence as well as a way to check one metric (parent or expert evaluations, for instance) against another, like movement in test scores or student engagement. Most studies have found performance-based pay (in the broadest sense) improves education, and that it (along with other rmeasures) improves education more effectively than if you just raise teacher pay across the board, which is, alas, "expensive and ineffective."

The most convincing case for both the importance of teachers and the value of rewarding the good ones while firing the bad probably comes from Eric Hanushek, who was featured in Gladwell's story. Hanushek's findings about education quality bears strong similarities to the data that Obama's budget director Peter Orzsag emphasizes about health care quality: Quality varies immensely across the nation, even in areas demographically similar, but the variations in quality of education and medicine do not correspond to variations in money spent. (The other parallel between education and health care here is that the US spends far more per student (and per patient) than many other developed countries who get markedly better results . Another sign something is badly amiss.) As the Hanushek Wikipedia page puts it,

[Hanushek's] research shows the overwhelming importance of teacher quality, although teacher quality is not closely related to the salaries, education, or experience of teachers..

Buyt as Gladwell's story shows vividly (go here and search for "Picture a young preschool teacher"), you can measure teacher effectiveness both by tracking how the teacher changes the students' achievement and skiil levels over the year and by things like videotaped observation that reveals how well teachers engage students.

If the argument that there's no way to systematically evaluate teachers -- that only the teacher's fellow teachers really know -- ever held water, it doesn't now. You can watch the teachers. You can ask parents and kids (filtering with needed salt). You can track how much improvement a teacher consistently brings in her students' achievement and skills compared to their achievement and skills the preceding year. Teaching efffectiveness can be measured. The idea that it can't be, and that we shouldn't try, is unacceptable and unworkable.

There are signs that teachers are beginning to understand this -- and that the push to get higher returns on teacher pay is something to join rather than fight, especially if it can mean higher pay for good teachers, as Michelle Rhee is proposing in DC. There are signs the other way, too, of course -- a resistance to any change, a somewhat natural tendency to defend one's profession/tribe and a security that in today's economy is increasingly valuable.

The key -- and Obama and Duncan are proving very good at this so far -- will be making it clear that this isn't an anti-teacher move, or a move to cut teachers' pay, but a move to reward good teachers more richly while raising standards, accountability, and the overall level of teaching and learning across the board. Every study, as well as common sense, makes it clear that this would mean systematically firing teachers who just can't teach; the data shows clearly that peers, supervisors, student achievement, and outside observers -- and I would suggest most parents, too -- can all tell good teachers from bad in a teacher's first three years. In a sensible system, the bad ones would be let go, but at present they are almost always given tenure.

A rational system would send those people to other pastures, where they'd probably be happier, and reward those that remain more richly. This would vastly improve our schools, send our children out much more ready to take on and enjoy the world -- and would seem to make teaching more enjoyable and more rewarding, in every sense of the term, even if less secure for those who don't do it well. It's not clear to me how huge an issue this security/tenure issue is; but I suspect it looms large, as naturally it would. (This is why many eduwonks suggest either a) leaving tenure in place for most present teachers and using non-tenured performance-based tracks for new hires or b) as Rhee proposes, giving both new and older teachers a choice between the two, confident that younger and new teachers will go for the higher-paying nontenured track, and that that -- and some defections, as it were -- will gradually move faculties toward the newer scheme.)

Security is nice (or so this freelancer has heard). But as everyone else's security shrinks, and our children emerge from school increasingly overmatched in an increasingly competitive world marketplace, it's getting harder and harder to justify security for teachers who aren't good at it.

Well. This could make things rather warm for me here in town. And to think I started this post really just to give you
Ezra Klein's nice clean wrap-up of Obama's main points yesterday. Here they are:

What does [Obama's] "compromise" approach mean in terms of concrete policy goals? Here are some highlights:

* Charters: In the biggest concession to reformers, Obama said he supported every state lifting caps on the number of charter schools allowed to open in a year, provided that states also have accountability guidelines for assessing charters and closing down ineffective schools. To put this in perspective, just two years ago, Randi Weingarten, now the president of the American Federation of Teachers, was aggressively lobbying in Albany to prevent New York from raising its charter school cap.

* Curriculum: National standards are emerging as a consensus point between teachers' unions and free market education reformers. Obama also supported higher standards today, saying, "Our curriculum for eighth graders is two full years behind top performing countries. That is a prescription for economic decline." But his agenda stops short of pursuing national curriculum guidelines or tests, promising only "to promote efforts to enhance the rigor of state-level curriculum."

* Teacher pay: Obama promised a federal investment in developing "performance pay" plans in 150 school districts. The language here is key. "Performance pay" is supported by teachers' unions, and awards salary bonuses to teachers based on a variety of factors, including classroom observations, teaching in hard-to-staff subjects and schools, and improving student achievement. "Merit pay," on the other hand, is understood as directly aligning teacher salaries to student test scores.

* Higher-ed: Obama promised to cut out middle-men in federal student lending, simplify the FAFSA form, and invest in community college efforts to better prepare students for the job market.

* Early childhood: Obama's budget will include "incentive grants" for states to develop uniform quality standards and target care and education to the most disadvantaged children.

Ezra Klein notes that, as he puts it,

Obama hates summer. "We can no longer afford an academic calendar designed when America was a nation of farmers who needed their children at home plowing the land at the end of each day," he says. "That calendar may have once made sense, but today, it puts us at a competitive disadvantage. Our children spend over a month less in school than children in South Korea. That is no way to prepare them for a 21st century economy. That is why I'm calling for us not only to expand effective after-school programs, but to rethink the school day to incorporate more time - whether during the summer or through expanded-day programs for children who need it."

I love summer, but I'm with Obama on this, as in much else on his education agenda. We are not preparing our children to compete it the world they'l graduate into -- yet when you point this out or urge change in schools at all, when you point out the problems, the most common response is "Why don't you focus on the good things about our schools?" This will seem a lame response when it becomes clear how far behind their international peers they are -- and when that gap really starts to hurt.

More like this

Last month I drew notice to an Atlantic story about (and an interview with) Michelle Rhee, the Washington, D.C., school chancellor who is aggressively pressing reforms in that district, most notably an effort to replace tenured teacher tracks with a system emphasizing higher salaries but more…
A while ago, I listed what I thought made Massachusetts' educational system work: The recent educational regression reform plan in Massachusetts and the Obama Administration's educational proposals both seem to misunderstand what has made Massachusetts' educational system one of the best in the…
Michelle Rheephoto by David Deal, from Atlantic Monthly To my surprise, one of the most-read posts on this mostly-science blog is "Are Teachers Profesionals of Public-Service Workers?", which looked at a NY Times Magazine piece on school reforms by Paul Tough. Tough now has Now there is a piece…
Curse you Gail Collins and your evidence! Last week, Gail Collins wrote a good op-ed about education 'reform.' While it's probably not anything too new for regular readers, it's good to see that the Rockefeller Republican/neo-liberal educational propaganda isn't being swallowed hook, line, and…

This is a well-reasoned analysis of what Obama said. In the long run though tenure for high school teachers is simply going to have to go or be drastically reduced. The essential point of tenure in universities is to allow for intellectual freedom by allowing research in controversial or unpopular subjects without fear of reprisal. The vast majority of high school teachers don't need to worry about this. Tenure serves little purpose other than to keep the bad teachers and the lazy teachers in their jobs.

Obama should take a stronger stand: make it evidence-based performance. The bases (plural) for measuring performance must be proven by evidence that they are sound. Anybody can concoct a rating system, or just pull one out of the air: what is needed is a reliable way of rating performance in terms of how well the teacher does within what circumstances. A Jaime Escalante ought to be amply rewarded.

Joshua Zelinksy is correct: no tenure for k-12. Also, there should be no automatic pay rescaling for teachers taking classes themselves and getting more degrees. If the quality doesn't show in what the teacher does in the classroom, the extra education is clearly worthless.

Anecdotally, there was years ago a famous case of somebody who worked her way up to a doctorate in education but was found out to be dyslexic and functionally illiterate, but well paid.

Whoa... hang on. Do you even know what purpose tenure serves? It keeps teachers like me from getting kicked out for teaching things that the principles and parents in a community disagree with. It makes sure that we can't get fired for, say, teaching evolution in rural Tennessee.

Second, teachers have to go back and get more education - we have to brush up on the latest tools, tricks, and educational standards. The pay-raise is incentive to do so at a much sooner rate, and it's to help compensate - after all, once you've got a masters, wouldn't you want to be paid like you've got a masters degree, and not a bachelor's degree? Finally, once you get your masters degree, it's irrelevant to get anymore degrees, because your pay actually falls off - that's when you move into other fields.

Third, when you start talking about national standards, you have to realize that such a beast doesn't - and cannot - exist in reality. Saying every child will read at third grade level in third grade is a nice pipe dream. It. Is. Impossible. Kids come to school with problems that are beyond our control. They come to school hungry. They come to school abused. They come to school from broken homes. Sometimes, just getting a kid to feel comfortable in a classroom is a huge accomplishment. Getting them to move from hitting someone at the first provocation to simply cussing them out with every word in the English language is an accomplishment. Sometimes, that takes all year long. Is that "accomplishment" by your standards? The child still has difficulty reading and writing. But they're comfortable in the class, so they can start learning - but do I really get a pay increase, or do I get docked because the kid reads at a 3th grade level in 10th grade? Which one of the "geniuses" at the top gets to define progress?

In a perfect world, it'd made sense. But folks, I hate to break it to you: this isn't a perfect world. We need real-world educators calling the shots, not pinheads who think they know what school is like because they graduated. Just because I go to the doctor's office doesn't mean I can tell a doctor how to do their job. Just because you go, or went, to school doesn't mean you can tell teachers how to do their jobs either.


By TheEngima32 (not verified) on 11 Mar 2009 #permalink

I hope to write a longer post on this later, but for now, let me first acknowledge that Enigma raises some legitimate concerns -- and then ask you, Enigma, to consider the possibility that an evidence-based system of performace-based pay based on assessment by a number of metrics (rather than just absolute student scores) used nationwide, but applied to account for local demographic variables, can actually ANSWER many of those concerns.

Such a system could protect you, for example, for being fired for teaching evolution, because if your evaluation showed that your students year after year (or even most years) were learning biology as well or better than most teachers with similar students, you'd keep your job -- or be able to successfully sue the school board. You'd have the data to show you're a good teacher.

Likewise, the sort of evaluations being proposed by Hanuscheck and others would not simply demand that every third-teacher get every kid reading at third-grade level; rather the evaluations would track (among other things) how much that teachers' students tended to advance in a year compared to where they were when they entered your class. Thus if I were teaching third grade and my students, as a whole, advanced as much or more than similar groups of students, I'd be recognized as a good teacher. If they didn't, I'd need to find ways to improve -- and if I couldn't or wouldn't, then I'd need to find another line of work.

It's important to realize the sorts of evaluations we're talking about: This is not simply a demand to raise kids to a certain level regardless of differences in the readiness of the kids who enter the class. The idea is to measure what Hanusheck calls the "value added," which is whether you push forward the students you have -- as compared not to some national aeverage, but to other kids of like demographic and academic background and skill level.

As to the comparison with doctors and whether I'd tell my doctor how to do medicine: This too is actually an argument for the sorts of reform Obama seems to be proposing. I don't tell my doctor how to practice medicine. But what I REALLY NEED -- what I'd kill for and may well die without -- is sophisticated but easily read data that shows me how my doctor's patients fare compared to similar patients treated by other doctors. I don't have that now -- for my doctors or for my children's teachers. We need the data for both.

And note that at present I am quite free to fire my doctor and go to another. A bad teacher? A kid's stuck there for a year, sometimes longer, and the damage -- to the student, to the educational system, to the country's economy and competiveness, and to the credibility of the teaching profession -- is tremendous.

The notion of differentiated pay for teachers needs to be discussed, evaluated and debated nationally and especially locally, at the district level. I wish more teachers and teacher unions would at least entertain the idea of an alternative pay system. I've never been a teacher, but was a school board member for 10 plus years. I tried to open a discussion on alternative pay with our teachers, some of whom were excited about the idea. However, many others had a viseral negative reaction and successfully shut down any further discussion on the topic. That seemed closed minded to me.

The salary schedule, used by virtually by all school districts, bases teacher salary increases on years of service and advanced degrees, or hours toward advanced degrees. The research is pretty clear that these two dimensions do not correlate to effective teaching. So why do we do it? My best answer is a combination of momentum and fairness. Momentum because the salary schedule has been in place for 80 plus years and fairness because teachers they will at the mercy of a principal's subjective evaluation criteria.

Is there a better way? There are some very interesting models for differentiated pay, such as those from the University of Wisconsin, Madison. Although these models have scant research that they are effective, they should at least get discussed fairly, with an open mind. Components of these models include:

Yes...paying teachers a bonus based on test scores, but, doing it as a team bounus (for student academic growth across an entire grade or building) which avoids the concern of pitting one teacher against another and instead actually encourages teacher colaboration. Testing would be based on a growth model vs. a criterion referenced test. This addresses the valid concern of "my class is slower, has special needs kids, etc..

Paying teachers based on demonstrable skills and competencies. Examples of this might include National Board Certification, a second certification (reading specialist, bi-lingual, etc.) a locally developed "Master Teacher" criteria and evaluation system that includes both teachers and administrators as evaluators, or career path options like curriiculum development.

Finally, the design of any new pay system should be based on a set of guiding principles that both teachers and administrators jointly develop and agree to. An over riding principle of "done with teachers and for teachers, not to teachers," is critical to ensure both real and perceived concerns by teachers are addressed and that teachers are empowered as equal participants in the design, implementation and execution of the program.

The current model for paying teachers is inadequate. Differetntiated pay, designed in way that addresses valid teacher concerns has real promise. The discussion needs to start.

By ron carlson (not verified) on 26 Mar 2009 #permalink

The evidence is not "a little thin." The later major study by Vanderbilt University clearer concluded that, "teacher performance pay alone does not raise student test scores." But the pressure is so great that even Vanderbilt felt compelled to include the word "alone." The only factor the study evaluated was whether pay incentives positively affected. The answer was NO. There is nothing in the study to suggest pay incentive with anything else may be effective. Why the word "alone" was inserted suggests that evidence will have little to do with how pay for performance in perceived.

By Gary Walker (not verified) on 22 Mar 2011 #permalink