Ed Note – This week Eric Hanushek and Diane Ravitch are discussing whether removing the lowest-performing teachers is a good idea. Hanushek started the debate on Monday and Ravitch responded yesterday. Here’s Hanushek’s response below. Ravitch will finish the conversation with a post tomorrow.
By Eric Hanushek
Diane comes back to a simple prescription: We should pursue business as usual with a few extensions of current policy. Unfortunately that is not serving us well, because this is exactly what we have done for several decades. We have developed a system that pays little attention to students and their achievement but that supports any adult who has found a job in schools. This policy does not look good by historical evidence on student outcomes. But it is common to defend this basic lack of management by throwing in red herrings whenever any policy change is suggested.
Once more through the evidence:
We have clear and consistent estimates about the variation in teacher effectiveness that exists in schools. The information comes from information on student test scores – something that is directly related to future student earnings and to the aggregate performance of the economy.
Differences in skills on tests of math, science, and reading lead to stunning differences in economic outcomes. Economic outcomes are not everything. But, as we have a continued national debate about both our international competitiveness and the fate of the bottom of the income distribution, we should not ignore economic outcomes.
Undoubtedly other, unmeasured things beyond test scores are also important to students and to society, but there is no reason to believe that being good in these other things is hurt by having greater measured skills. And there is no reason to believe that teachers at the bottom in terms of producing measured skills are anything but the bottom in producing useful unmeasured skills.
It is a red herring to say there might be unmeasured other things that are also important.
Further, there is now consistent evidence that ratings by principals (on metrics other than test scores) are highly correlated with ratings on test scores at the top and bottom of the distributions. While there is confusion in the middle, there is not confusion at the top and bottom.
It is a red herring to say that different evaluation systems produce different results.
If we put together the impact of good teachers on student achievement and the impact of achievement in the lifetime earnings of students, we find that a good teacher (one in the top quarter in terms of effectiveness) each year produces over $350,000 more income for her students compared to an average teacher. But, symmetrically, a teacher in the bottom quarter subtracts $350,000 in income each year of teaching compared to an average teacher.
Ignoring these differences leads to huge inequities and to enormous waste in the potential of our students. I personally think the stakes are large enough that we should consider something other than business as usual.
As Diane points out, I have always advocated other policies that support the improvement of our teaching force: better mentoring, improved professional development, well designed curricula, complementary technologies, and so forth. But, my interpretation of the evidence is simply that none of these policies has been very successful at improving the least effective teachers. Further, our “best” teacher development policies do not go to scale up very well and are not systematically chosen by districts.
It is a red herring to point to other, complementary policies for improving teacher effectiveness without acknowledging the importance of starting with effective teachers.
Finally, it is fine to talk about teacher turnover, but not all teacher turnover is bad. If we move a bad teacher out of a school serving disadvantaged students, it is not a bad thing. The turnover rate in teaching is no different from that in other professions, and the initial turnover in teachers should not be used as a reason for ignoring the effectiveness of teachers. There is an excess supply of potential teachers. The shortage is not teachers per se but effective teachers.
It is a red hearing to say that teacher turnover is high without considering the implications for teacher effectiveness and ultimately for student achievement.
It is unfortunate that the “applause lines” to which Diane gravitates tend to be red herrings, distracting us from the national imperative to improve the effectiveness of our teacher force. I have focused on one specific policy – eliminating the teachers that are harming our children. Diane wants to introduce the idea that, while there are teachers who are harming kids, we should not deal with them because there might be some residual uncertainty about the very last teacher who is in this group.
She repeatedly use the word “firing” to conjure up images of large, unjust, and arbitrary actions. To the contrary, it is simply good management to move a very small number of ineffective people out of the front line of schools. And by so doing, it acts to elevate the status of the vast majority of effective teachers.
Policy by red herring seldom leads to good policy. It certainly does not when considering teacher policy. Diane is very skillful at distracting us from issues of student achievement, but avoiding those issues will be very costly to the nation.
23 Replies to “Ignoring Red Herrings”
Wow! I thought your post Monday might be a step in a constructive discussion. After all yesterday’s NAEP scores offered one more wake-up call for the data-driven crowd. But how do you reply, “Red Herring,” “Red Herring,” “Red Herring,” “Red Herring.”
Name calling is not a policy for improving kids’ futures.
You inadvertantly revealed the original sin of your movement – hubris.
And I am not just talking about your insulting of such a prominent scholar. Diane Ravitch has invested a career in seeking solutions. I doubt you have a fraction of her knowledge of teaching, learning, and schooling, but you foreclose the idea that you can learn how to better use your statistical models by listening to her wisdom.
And the same applies to practioners. Would you absolutely refuse to listen to doctors who disagree with an economist’s interpretation of data from an experiment? Why will you not consider our arguments that your policies recommendations would hurt teacher effectiveness in inner city schools? How about calculating some outcomes for students if your theories backfire?
And if you won’t lower yourself to listening to teachers, why not heed the overwhelming evidence of social scientists and cognitive sciences about the ways that children, adults, and systems learn?
Why not discuss peer review and other methods of efficiently removing ineffective teachers? And also, why not ask why it becomes more difficult to fire bad teachers when you are so cavalier about also firing good teachers caught in your net?
Good response. Look forward to reading tomorrow.
I have a question for Professor Hanushek and other researchers in education:
If the state tests are mostly invalid because of all the test prep, cheating and gaming, aren’t the research studies based on these tests also invalid?
John said, “Name calling is not a policy for improving kids’ futures.”
Calling a red herring a red herring is name calling?
Would you also be offended if one referenced your appeals to authority by name?
I’m not offended that Professor Hanushek says “It is a red herring to say there might be unmeasured other things that are also important.” But I can’t take seriously the idea that anyone who has worked directly with young people and standardized tests believes it.
John and Linda,
I see that you constantly post on Andy’s blog, despite the fact that you obviously disagree with his analysis (it should be said I’m all for debate). And while I won’t get into the weeds on the policy opinion, I do have a simple question:
If what the “reformers” are proposing are wrong, why don’t you open up a charter school and prove it?
Seriously, the time you waste on a blog that you don’t agree with could be much better spent proving your point on the ground.
I am disappointed when thoughtful individuals emphasize tests that examine such a limited domain of knowledge, especially in regard to teacher policy. We may be able to link scores to economic outcomes, but there are major consequences when test scores become the primary measure of what constitutes learning or effective teaching. When I read posts such as the one above, I often feel that economists often don’t understand what is at the heart of other reformers arguments. I don’t want the status quo. The status quo is unacceptable. Outcomes such as educational attainment, achievement gaps…none of it is acceptable, and there needs to be a radical change to the system, but I don’t see tinkering around the edges as a significant solution.
First, even if one accepts VAM as providing useful information about individual teachers, there are so many challenges, such as the difficulty of vertical scaling, that to attach significant policy consequences to it is ethically dubious. That said, my issues with revising teacher policy based on weeding out bad teachers using tests is more about the ways these well-intentioned (one hopes) approaches impact students.
No matter what is said in theory, the tests used in current accountability systems frequently act as the de facto curriculum. They determine the allocation of resources such as time, personnel, and money. Most importantly, classroom instruction time.
While my children’s school has a fairly well-rounded curriculum, they have the “luxury” of including science, social studies, art, music, and even gardening *because* they have high test scores. Those test scores, by the way, can largely be explained by the fact that we live in a university town. At my children’s school, 60% of the students have at least one parent who attended graduate school. Another 29% have their Bachelor’s degree. Step into a school that is not performing as well, and the consequences for underperformance are such that about 4 hours a day are spent on ELA, an hour on math, and the final hour is reserved for any other instruction.
Further, the tests used to measure effectiveness offer little incentive for schools to provide the kind of instruction necessary for students to improve their economic outcomes in a changing world. The “basic skills” assessed on standardized tests are what I — at first — deemed “necessary but not sufficient.” Now, I would argue that many of the elements tested do not even qualify as “necessary.” The skills are often factoids and basic algorithms. Do I want my child to know them? Sure. Do I think that the scores should be used as a way of measuring the quality of their education? Absolutely not!!
One thing other countries understand that puts us at a disadvantage — the multiple choice format relies on such low-level responses, setting the bar too *low* for the skills that our students will need in the future. We are expending so much time and energy on raising scores that we have lost sight of the truly lofty standards. When the highest stakes for our schools are set using such a limited form of measurement, it’s been shown time and again that districts, schools, and teachers will narrow their focus, especially in schools that are at risk of not meeting the necessary level of achievement (e.g., schools that serve large populations of ELLs, ethnic minorities, and the poor). Curriculum narrowing disproportionately impacts the students that we aimed to help by “raising the standards.” Talk about perpetuating — even widening — the achievement gap! I’m terrified about what we will see in the longitudinal data about the outcomes for the students who attend schools that have focused so intensely on test scores. Short term gains at the expense of our ultimate goals?
I have worked in education for 15 years, and I’ve visited over 100 schools in a four-county area. The teachers in my children’s school are great, but some of the best teachers I’ve seen work in far tougher settings and their efforts aren’t always visible in their test scores. Designing teacher policy that even hints at student achievement as a measure of effectiveness in evaluations and publications (i.e., the LA Times) will exacerbate the difficulties recruiting and retaining teachers in urban/low-performing schools. No economist or statistician will be able to convince the classroom teacher that they can “control” for outside-of-school factors in their model.
I am concerned when vocal factions of the reform movement point to other reformers and say, “If you are not with us on raising accountability for performance on these tests, you support the status quo.” It reminds me of when political adversaries rely on binary definitions of what it means to be patriotic or a “true American.” I reject the characterizations offered by Hanushek in this post [and am saddened by the rhetoric], because I don’t think Diane Ravitch and others are advocating “business as usual” with a few minor changes.
Hanushek states that, “Policy by red herring seldom leads to good policy.” I concede to this point. Using the current argument that focusing on a small percentage of teachers will dramatically effect the U.S. education system doesn’t pass the sniff test. In fact, the argument smells much like herring to me.
Mary (above) makes some very good points, including the following: “The teachers in my children’s school are great, but some of the best teachers I’ve seen work in far tougher settings and their efforts aren’t always visible in their test scores. Designing teacher policy that even hints at student achievement as a measure of effectiveness in evaluations and publications (i.e., the LA Times) will exacerbate the difficulties recruiting and retaining teachers in urban/low-performing schools.”
Having worked in several different schools, I’ve seen little difference between teachers in low-income and high-income schools. As Mary noted, some of the best and hardest working teachers are in the low-income schools, where student test scores will be low despite their efforts. The difference between these schools boils down to the students who attend them (and their parents), not their teachers.
Note: By my comment above, I don’t mean to imply that teachers are unimportant (not at all) but simply that in my experience there are relatively small differences between teachers in “failing” schools vs. those is “successful” schools.
Blaming teachers for the huge difference in academic skills that are tied to poverty, English language proficiency, student behavior, parental education levels and MANY other important factors is simply counterproductive. And, as Mary observed above, such policies will likely only serve to exacerbate the problem of attracting good teachers to low-income schools.
@Attorney DC – You may have seen that in your experience, that there is little difference between teachers in low income and teachers in high income schools. Others may beg to differ. Case in point – I’ve had several friends apply to work in wealthier counties in the DMV, with excellent references from the low-income schools we all work in, and a general knowing that they are the best that those schools have got. They were told each and every time that the experience wealthier districts have with lower income schoolteachers is that although management skills are unmatched, oftentimes their content pedagogy was lacking in higher order skill teaching.
Again, just a different experience and I’m not saying yours is invalid, but hopefully it serves its purpose as a foil to any attempts to generalize each of our own limited experiences in school districts.
Hanushek makes a 2-step argument — 1) poorly-performing teachers are a major cause of low student test scores; and 2) low student test scores are a major cause of depressed lifetime earnings for the tested student. He cites no meaningful evidence to support the first step and cites correlation, not causation, evidence to support the second step.
Re the poorly-performing-teachers-cause-low-test-scores: Hanushek asserts: “[T]here is now consistent evidence that ratings by principals (on metrics other than test scores) are highly correlated with ratings on test scores at the top and bottom of the distributions” — that is, that principals somehow know (perhaps based on observations) which teachers are able to raise student test scores and which teachers are not. To support this assertion, he links to a 2006 study in which principals’ ratings regarding teachers’ ability to raise student test scores correlated with the amount of increase/decrease in the students’ test scores. However, it appears that, when the principals in the study rated the teachers regarding the teachers’ ability to raise student test scores, the principals knew to what extent the rated teachers had actually raised (or not raised) student test scores over the prior 7 years. In short, a study that shows, at most, that teachers who raise test scores in YR 1 are somewhat more likely than average to raise test scores in YR 2.
However, assuming arguendo that Teacher X’s students regularly show significant increases in test scores while Teacher Y’s students regularly show minimal increases in test scores, it does not follow that Teacher X is a more effective than Teacher Y. At a minimum, we would have to control for the percentage of difficult students regularly assigned to X and Y, for the class-size/total-student-load/number-of-preparations for X and Y, and for the amount of admin support provided to X and Y. Contrary to Hanushek’s contention, studies of school systems where teacher evaluations were based, in part, on student test scores, show that a teacher’s ratings often vary widely from year to year — that is, a teacher rated A in YR 1 is often rated C in YR 2.
Furthermore, as other commenters have noted, if teacher quality were a major driver for student test scores, one would expect affluent suburban schools to score only slightly higher than inner-city schools on standardized tests. Granted — affluent suburban schools offer better working conditions and therefore attract, on average, more qualified teachers, but there is no reason to believe that suburban teachers are tremendously more qualified than inner-city teachers, particularly given the large numbers of TFA high-achievers who have flooded into the inner-city schools over the past 15 years. The fact that suburban schools score tremendously higher than inner city schools is strong evidence that teacher quality is not the, or even a, major driver of student test scores.
Re the low-student-test-scores-drive-depressed-lifetime earnings step: Hanushek’s evidence shows correlation, not causation. Low student test scores also correlate strongly with low parental SES, student learning disabilities, student emotional disturbance, student non-English-speaking, and the other inner-city demographic characteristics. It is much more likely that these inner-city demographic characteristics drive both the low student test scores and the depressed lifetime earnings.
Bottom line: If we could identify/improve/discharge incompetent and/or lazy teachers, we would have somewhat better schools and somewhat higher test scores. But there is no reason to believe that this would significantly solve the problems plaguing our inner-city schools — i.e., student misbehavior (chronic absenteeism, chronic tardiness, and minor but endemic classroom misconduct) and students reading far below grade level.
Hanushek argues that lower test scores affect future earning potential. And he assigns the blame for those low scores on all those under-performing teachers. Yet, as LaborLawyer (above) points out, test scores are correlated strongly with family income level. So doesn’t the data point to the lack of social mobility in our country rather than poor teaching?
Also, what is with all these charts with dotted lines (which means speculation, no?) Was there a study done to actually prove that an ineffective teacher has that sort of impact? “We have clear and consistent estimates”? That’s not data. Statistics is not my realm of expertise here, but it seems made up.
Lastly, if there’s a red herring here, in my opinion, it is this talk of “bad teachers” which is diverting attention away from real reforms like addressing the mental health needs of our kids growing up in poverty, reducing class size, getting quality Pre-K for all children, resuming talks about equitable funding for high-poverty schools, creating more appropriate alternative schools and classrooms, and improving teacher preperation and collaboration. Why did we stop talking about these thing?
Mary, thank you for your excellent response.
Mary, thank you. Everyone else? Tweet Mary’s post, email it to everyone you know. Print it out and rain it down on the streets.
Oh, and let’s all send a copy to Arne Duncan, shall we?
It concerns me that in one breath it can be stated that “economic outcomes are not everything. But, as we have a continued national debate about both our international competitiveness and the fate of the bottom of the income distribution, we should not ignore economic outcomes.” How does this involve the needs of “we, ““us,” and “our” when the individual student’s needs are not included? As an English teacher in a rural high school that one year made AYP and the next did not, I have to question success as that which defines accountability through the lens of economic impact. Diane stated that “even in a perfect world where all the assumptions of his calculation hold up, we have very little idea of what the distribution of teacher effectiveness looks like when that effectiveness is measured using an evaluation system comprised of multiple measures.” From 2009 to 2011 the measures of effectiveness were indeed multiple as my focus of instruction was driven by what the test gurus, who provided professional development input, considered important.
Mary posted that “we may be able to link scores to economic outcomes, but there are major consequences when test scores become the primary measure of what constitutes learning or effective teaching.” As a teacher I do not want to be reduced to test scores that define my ability. I find this demeaning as a single English class includes 32 diverse students who should not be held accountable for the “income distribution” convoluted by inherited wealth, inherited poverty, immigration regulations, and cultural delineation. Why is teacher/school/district/state/national success ultimately defined by how students meet expectations of standardized test accountability? I have to ask myself, whose economic outcomes are at the root of this accountability based in my ability to be a teacher, and why?
By reading the above posts you can see that John and I are not wasting our time. It appears that we are doing a good job proving our point “on the ground,” where I presume most of these posters live. Remember, the pen is mighty!
Fortunately more and more people are seeing the idiocy of the “reform” movement. For example, Prof. Hanushek probably based all his fancy research on test scores presented to him by school district administrators. Are the “highly effective” teachers the best teachers or are they the ones who drilled their kids on the exact test items? We don’t really know, do we?
Please keep an open mind because one of our greatest institutions, and the bedrock of our democracy, is being attacked.
Looking back at the comments tonight, I’ve enjoyed the conversation in the comments section (once I stopped cringing over the typos that slipped past me in my rush to respond. I appreciate that Eduwonk created an opportunity for Ravitch and Hanushek to interact in this forum. Too often, today’s media is structured so that the writers are preaching to their respective choirs. I have a lot of respect for what Hanushek believes his research can do to inform the debate. I don’t agree with his conclusions about the fundamental assumptions undergirding his work and the way his models can be applied, and I doubt my rushed response will convert him to my perspective. That said, the discussion has pushed my thinking…Thank you!
I’m a bit disappointed with this rebuttal. The strongest point made, in my opinion, was that
“It is a red herring to point to other, complementary policies for improving teacher effectiveness without acknowledging the importance of starting with effective teachers.”
This is where Diane Ravitch loses me almost every time. I often start out agreeing with something she’s saying, but somewhere along the way I get frustrated because she always seems to ignore or deny the need to start with a better teaching force to begin with – which requires, among other things, weeding out the very worst teachers. She wants us to “respect” the teaching profession like other countries do, but she doesn’t want us to deal with the inconvenient truth that teachers in those countries are respected because a lot was expected of them as students and young educators – the complete opposite of the situation in the U.S. They are respected because they were recruited from the top, not the bottom. Making teaching a more respected position requires changing the demographic in that position, and unless we want that to take half a century, we have to recruit and reward high performers AND weed out low performers.
Unfortunately, Rick Hanushek doesn’t seem to have answered the biggest concern educators – rightly – have about what he proposes: Who, exactly, are we talking about firing, and how will we do it? Will we look at every teacher’s test scores and fire the worst 5-10% in every school? Or will it be the worst 5-10% in every district? Every state? What about failing teachers who teach subjects that aren’t tested – how will we identify them so we can fire them? Will test scores be the sole measure? And how will we actually go about this firing process? Will we ask principals to identify their worst 5-10% and fire them, or will we give that authority to people higher up in the chain, like superintendents? Union rules obviously won’t allow any tenured teachers to be fired, so how will we negotiate with them? Should we try to have a law passed requiring that the worst 5-10% be fired? Or perhaps we should propose legislation which would simply make it easier for principals to fire ineffective teachers, removing some of the union protection and bureaucratic hassles that currently prevent schools from getting rid of bad teachers?
Refusing to even briefly address these practical matters gives readers legitimate reason to be skeptical of the whole idea. If we’re just going to use test score data – which gives an incomplete, oversimplified and often distorted picture of students’ learning – to determine who falls in the 5-10%, then I’m not so sure I support the idea. (Hanushek does a decent job of backing this up when it comes to reading, math, and science, but what about all the other subjects?) If we’re going to eliminate the bottom 5-10% of a very small pool – like a single school – then I become skeptical because we’ll end up firing a lot of decent teachers and keeping many worse ones. If we’re going to eliminate the worst 5-10% of a very large pool – like an entire state – without a plan for how to fill all the empty spots that will leave in a lot of very poor schools, then we’re clearly headed for disaster.
With all this in mind, I’m not so sure that firing the worst 5-10% is a realistic goal, though I don’t doubt that it is, in some way, necessary. Perhaps what we should really be striving for is a more general policy that would make it less difficult for principals to fire ineffective teachers and, in cases of budget cuts and similar situations, allow them to decide who to fire based on effectiveness, not just seniority. Over time, this type of policy would likely weed out that pesky 5-10%.
I agree that eliminating very low-performing teachers – coupled with better methods of recruitment, teacher training, compensation, and professional development – is essential to transforming the teaching profession. From a purely rhetorical/ideological standpoint, I don’t understand why Ravitch doesn’t also agree. However, if Hanushek wants to win more than just a rhetorical debate, he’s going to have to address some of the more practical elements of what he is proposing. I’m interested to see if Ravitch will address them in her response.
To pgteach: Sorry for the delay in responding to your comment above (regarding the difference between teachers in low-income and high-income schools).
In response, I’d agree that there are different skills and knowledge that come in handy teaching in different types of environments. I wouldn’t doubt that veterans of inner city schools would have pretty good classroom management techniques while veterans of higher income suburban schools would have more finely honed academic preparations. In addition, I know that higher income schools generally have more experienced teachers than lower income schools (in part because they have a larger selection of applicants for each position and can afford to pick only those with more experience, and in part because teachers like working in those schools so there’s less turnover).
However, you appear to be conflating teaching QUALITY with context-dependent teaching practice, in that inner city teachers HAVE TO concentrate more on classroom management while suburban teachers HAVE TO concentrate more on higher level academics (because their students are taking higher level courses), due to differences in their student populations. This doesn’t necessarily mean that teachers in X school are worse (or better) than teachers in Y school — only that their teaching reflects their students’ needs.
My earlier point was more that, on the ground, walking around in different teachers’ classes, I simply didn’t see much of an overall difference between the teaching in low and high-income schools. Obviously, more class time is spent on classroom management and behavior issues in lower income, minority schools (because it’s necessary) but talking about the actual teaching methods, they seem pretty similar: Teachers use blackboards (or whiteboards) along with textbooks, worksheets, media and other tools to teach material; students take quizzes and tests; teachers assign homework and classwork. If you took the kids out of the picture, and just watched what the teachers did, it is pretty similar. As noted above, to the extent that teachers of low-income kids spend less time on higher-level subjects, that’s because their students aren’t prepared for those subjects, not necessarily because they wouldn’t be able to teach them.
Based on my observations and reasoning, I simply don’t think that TEACHERS are the main cause (or even a significant cause) of the gap between low and high income kids (or between minority and white/Asian kids). In light of the above, I think that reformers should stop their incessant, spotlight focus on TEACHERS as the problem in our schools, because it’s not going to solve the problem.
Trying to draw the line between an “effective” teacher and one less so is probably nebulous at best. Even if you can draw that line, trying to fire a teacher, at least in NY, is like trying shovel dirt with a pitch fork.
All students learn and progress differently and they always will. When I was in HS I loved math and could have done all four years in two. Social Studies was like learning Chinese. But the thing about learning Chinese is that the more you keep at it, the more you do it, the better you get.
Maybe we’re looking in the wrong places for the answer(s). Maybe we should be sending our children to school for 12 months a year instead of 10. Maybe we should look at giving students gym time credit for being in a sport and then using that credit-time to get face time with a teacher in their weakest subject. Maybe we should have more assistants in classrooms. Maybe we should utilize psychologists more in the classroom. I don’t think there’s a magic bullet for this problem. Maybe most of the pieces to this puzzle are broken to some degree and each one has to be fixed in order for the education engine to run better.
“student test scores – something that is directly related to future student earnings and to the aggregate performance of the economy.”
As an economist, I hope you recognize that their are far better predictors of the aggregate performance of the U.S. economy than student test score.
Is this one of those Red Herrings you were talking about?
Q: Was Arne Duncan wrong when he said our evaluation system was broken because 95% of teachers received satisfactory ratings?
A: Not according to Eric Hanushek who says over 90% of our teachers are really quite good.
But lets back up a second. According to Cedar Riener,
“policy efforts urgently focused on improving through selection, rather than supporting all who have chosen a profession, are doomed to fail.”(1)
Not only are there problems of measurement (esp, its accuracy vs. stabiliy),
there are problems with principals who favor the teachers who follow their line and set them up to do better than their peers. There is also the problem of ranking teachers relative to one another, thus creating a
disincentive for cooperation among the staff.
But, leaving those problems aside, what does Eric Hanushek say about the quality of teachers? And where would he suggest his research leads?
On the first issue he says he quality of the average teacher is really quite good. But this seems to get lost in the noise about Finland:
So bringing up Finland, if we could replace the bottom five to eight percent of our teachers in terms of effectiveness with just an average teacher — and an average teacher is quite good in our schools — if we could replace the bottom five to eight percent with an average teacher, our national achievement would rise to the level of Finland. (Eric Hanushek speaking, “Class Size and Student Achievement,” Diane Rehm Show, 8 March 2011;)
Not only is the average teacher ‘quite good, but in defending his position regarding active deselection, Hanushek writes, that this does not apply to “teachers en masse . . . but a small number are dreadful. . . . The typical teacher is both hard-working and effective [but] a small proportion of teachers at the bottom is dragging down our schools. . . . replace the bottom 5%-10% of teachers with an average teacher–not a superstar–we could dramatically improve student achievement.”(2)
If the average teacher is actually ‘quite good,’ does that suggest that the system is failing and needs to be drastically restructured? Or does it suggest that, as regards teacher quality, it is basically sound? Does it seem to justify removing 5 to 8% (10%?) of teachers year after year?
Also, if we need to remove only 5 to 8% (it seem to have gone up to 10% only in the WSJ article) – a big article of faith with the Teacher Quality people – what does that say about the other 90 to 92 to 95%?
So, again,was Arne Duncan wrong when he said the evaluation system was broken because 95% of teachers received satisfactory ratings?
(1) Cedar Riener (“Deselection of the Bottom 8%: Lessons from Eugenics for Modern School Reform,” Scientific American invited Guest Blog, July 19, 2011)
(2) Eric Hanushek, “There Is No ‘War on Teachers’ There is a growing bipartisan agreement on the importance of rewarding good ones,” WSJ Opinion October 19, 2010. As always, he incluided the claim, “The U.S. could move from below average in international comparisons to near the top.” Having to reaccess this, I found it at a website associated with Oklahoma State called Orange Power (http://www.orangepower.com/threads/there-is-no-war-on-teachers-wsj.102284/) and including the following posted comment, “After all, the good employees in any work setting would rarely need a union. Only the bad ones need the shelter of the union.” Nearly all the rest of the comments focused on unions.
By the way, I am currently writing on this subject and would appreciate hearing from those who want to discuss it.