Can you believe how many millions, hundreds of millions, or billions of dollars have been diverted from America’s classrooms in the search for the elusive “bad teacher”? Lest we forget, this was imposed on the nation’s public schools by Race to the Top, and it is a central narrative of the reformster ideology. Find and fire those “bad teachers” and America’s economy will grow by trillions of dollars (so said Hoover Institution economist Eric Hanushek).
Except it turns out that no one has been able to find those hordes of “bad teachers.” They must be hiding. Or they must be good at test prep. In state after state, the hugely expensive teacher evaluation systems–burdened with statistically dubious methods–have been unable to unmask them.
Politico reports that 97% of teachers in New Jersey were found to be either effective or highly effective:
MOST NEW JERSEY TEACHERS RATED EFFECTIVE OR BETTER: Three percent of New Jersey teachers earned a rating of “partially effective” or “ineffective” under the state’s new teacher evaluation system, according to a report [http://bit.ly/1K5q30f ] out Monday. That’s up from the 0.8 percent of teachers rated “not acceptable” under the state’s old acceptable/not acceptable system. The 2,900 teachers rated poorly under the new system taught about 13 percent of the state’s students, or 180,000 kids. “Those educators are now on a path to improvement with individualized support, or will face charges of inefficiency if unable or unwilling to better serve students over time,” the report says. The vast majority of teachers earned high ratings, with nearly three-quarters rated “effective” and nearly a quarter “highly effective.” State officials stressed that teachers are now receiving more detailed and personalized feedback than ever before. “While one year of this new data is insufficient for identifying sustained trends or making sweeping conclusions about the state’s teaching staff, we are proud of this significant improvement and the personalized support all educators are now receiving,” said Peter Shulman, assistant commissioner of education and chief talent officer.
-The New Jersey Education Association said it still has “deep concerns” about the implementation of the evaluation system and the data used in decision-making, but “these results show that teachers are working very hard to meet and exceed expectations.” NJEA is calling for “disaggregated data for teachers with challenging assignments. It is important to know whether the evaluation system is biased against teachers who work in special education, teach English-language learners, or who work in economically challenged communities,” NJEA said. And the union pledged to represent any member who believes his or her evaluation is flawed: http://bit.ly/1AGLG56.
– The results come just days after Gov. Chris Christie denounced [http://politico.pro/1J5ySrL] the Common Core. In remarks [http://politico.pro/1QkQgHX ], Christie also stressed that the state must continue its push on teacher evaluations. “On this we will be unyielding,” he said. “No one should stand for anything less than an excellent teacher in every classroom – not parents, other teachers, administrators or our students. Accountability in every classroom must be one of the pillars of our New Jersey based higher standards.”
It is puzzling to see that 3% of the state’s teachers taught 13% of the state’s students. How is that possible? Maybe the teachers would do better with smaller classes.
The hunt goes on, even though the hunters left empty-handed.

Yes, they should fire 3% of teachers, right after they fire 3% of police, firefighters, and military. Oh no, can’t do that, might be disrespectful to male dominated professions.
LikeLike
This gender issue needs to be made more prominent. What is the other occupation with outrageous demands and stress? Nursing.
LikeLike
in our country, the teacher can rich…much of them teach sincere and live by hard work…for remember “teacher unrespectful in social lives” not like use ago…
LikeLike
Remember what happened in NYS when the teachers received high evaluations??
The evaluation was deemed flawed and had to be reworked to make it more difficult to receive high ratings. Those teachers need to be held accountable and after we make them accountable and they come out on top, make them accountable again.
I’m just waiting for a system of accountability for our elected officials.
Christie has been out of the state (don’t even understand how a sitting elected official gets away with this on our tax dollar)stumping for President and even Bridgegate hasn’t been explained to the public and Cuomo (don’t get me started) is also on varied trips and NEVER has to explain himself or be accountable.
Where is the accountability to the taxpayers. Many more elected officials have no accountability and just railroad the taxpayers.
People stand up, we own them, they don’t own US, the words of Christie, “on this we will be unyielding”
LikeLike
I wouldn’t be surprised if Christie started some “new math” designed to ferret out all the “lazy” teachers of the most poverty stricken students. I am sure he could borrow a “a cup of phony statistics” from his neighbor Andrew Cuomo.
LikeLike
If anybody looks behind the numbers they will find ineffective teachers ranked highly effective and highly effective teachers ranked ineffective. A vast majority of teachers are good at what they do, or they likely would have left the profession within five years.
We personally are dealing with a poor teacher, but since we as parents, and most other parents in our district, are fortunate enough to compensate, the teacher will incorrectly be considered effective. On the flip side, many teachers work with the most challenged students – truancy, health, transience, family breakdown, job loss, poverty, learning challenges, raising children, addiction, crime, mental health, homelessness. All problems I see everyday. Yet the teachers working with these students will be unfairly judged ineffective based on test scores.
The evaluation systems are proving to be arbitrary and capricious. What is irrational is the politicians, billionaires, and Reformers desperately clinging to an idea that simply does not work.
LikeLike
Yes indeed. In Indiana the evaluation process is so flawed that some superb teachers are “JUST” rated effective and poorer ones are OK – according to their evaluation system.
This is DEVASTATING.
Yes, even with the flawed system they cannot find an abundance of bad teachers, at least in some other states.
These people, the politicians are blind, myopic, some other things I cannot put on this blog.
LikeLike
MathVale: as I see it, that’s because behind the massaging and torturing of numbers & stats by the rheephormsters lies their shamelessly open and oft-proclaimed “soft bigotry of low expectations”—
They presume that public school staffs and students and parents are, in their vast majority, by nature and nurture wanting in every respect. So when, by the most sacred measure-and-punish tools of the Rheephorm crowd, you get 97% effective and highly effective teachers, then the fault, Dear Brutus, lies in not effectively employing the “hard bigotry of mandated failure”—
As in ordering up high-stakes standardized tests that are designed, pretested and constructed to produce low scores that show that teachers and students and parents are utter and complete failures.
The rheeality is, the one and only thing that the rheephorm misuses and abuses of math (e.g., VAM) tells us is that the self-styled “education reform” movement will never ever abandon their foundational Marxist beliefs:
“Politics is the art of looking for trouble, finding it everywhere, diagnosing it incorrectly and applying the wrong remedies.”
Groucho. Today. Tomorrow. Forever.
😎
LikeLike
Re: those who did poorly:
If there were 180,000 students and 2900 teachers that would mean each teacher has 62 students.
If these are elementary teachers with 62 students they deserve a bonus. Or is it called ‘hazard pay’?
Maybe the millions spent on teacher evaluations should be spent on SMALL CLASS SIZE!
What a NOVEL idea!
And I did this w/o CC math–
LikeLike
So we have a teacher evaluation system that rates nearly all teachers effective, and we turn around and call it an invalid system. Yet at the same time we mandate that 100% of students should be “proficient”. Should there ever be a test that determines that all or nearly all students are “proficient”, what will be said about that test?
LikeLike
Anyone can make out a test that virtually everyone will pass. Likewise one can make out one where virtually every one will fail.
The makers of these tests are incompetent, ignorant – all of the things that teachers have been accused of.
Would that they had the perspicacity to look at themselves in the mirror.
No, they were voted into office. That makes them instant experts.
LikeLike
Inconvenient Truth for Christie and Anderson. Oh, don’t forget the name of a vicious newspaper editorial group who bashes teachers. These apologists are pretty much screwing themselves for manufacturing ‘failing education BS.’
LikeLike
So far, the use of student growth percentiles (SGP’s) in NJ (based on student standardized test scores) for teacher evaluation is fairly limited.
For 2013-2014 SGP accounted for 30% of the evaluation for only about 15% of all NJ teachers.
The randomness inherent in SGP (having nothing to do with the teacher) has nonetheless undoubtedly already resulted in the mis-classification of some teachers (in both directions) — and it is bound to get worse.
If (when?) SGP gets used for a larger fraction of teachers and/or is made to account for a larger % of the evaluation, we will undoubtedly see a greater % of teachers rated ineffective purely due to random variation associated with student growth scores
See this post by Eugene Stern regarding SGP.
Stern points out that even the people who developed the SGP models have said that “Large-scale assessment results [like SGP] are an important piece of evidence but are not sufficient to make causal claims about school or teacher quality”.In other words, they were not intended for evaluating individual teachers.
Stern also points out that ‘the precise weight [of the total evaluation — eg, 40% or 30%] doesn’t really matter. If the SGP scores vary a lot, and the other components don’t vary very much, SGP scores will drive the evaluation no matter what their weight”
LikeLike
We do know from history when it comes to a witch hunt (think McCarthyism), they will find their witches, accuse them, find them guilty, get rid of them and put them on a black list, but first they have to get rid of due process rights for public employees so innocent until proven guilty doesn’t stand in their way. First the witch hunt and then the gulag—no trial, jury or evidence needed.
LikeLike
Indeed it is very puzzling how 3% of the teachers teach 13% of the students. Could it be that the data is INACCURATE? No matter, it is a fact that the data is misused to evaluate teachers. I have seen teachers who have pass rates on an AP exam 10 points higher than the national pass rate and were rated “effective.”
How are the people who push out these ratings not embarrassed by the obviously inconsistent results? That’s really just a rhetorical question – politicians and sycophant bureaucrats have no shame.
LikeLike
What ever happened to evaluations by experienced supervisors and department heads–colleagues who work with teachers every day and see growth over time? In my experience this is what truly supports professional growth.
LikeLike
Something similar happened here with our teacher evaluation system in Louisiana. Numbers very similar to the “old” evaluation system that was in place before the reform laws. All that money, time, stress, etc. to end up with similar teacher evaluation numbers to those done by principals in the past. Only a small percentage difference at such a great cost. Just think if we had spent those tax dollars on our children, instead of a faux reform. It’s funny that the main stream media here in Louisiana never felt the urge to report those numbers.
LikeLike
They emphasize that under the new system 3 % of teachers were categorized as “ineffective” or “partially effective” while under the old (acceptable/unacceptable) system “less than 0.8%” were deemed “unacceptable”
But if you look at the breakdown of that 3%, only 0.2% were actually found “ineffective”.
And when it comes right down to it, there is subjectivity and uncertainty in any classification system meaning there may be no real difference between even 3% and 0.8%.
It’s very hard to believe that one can distinguish a 2% difference between teacher effectiveness with any evaluation system (and especially not between different systems) . And implying that one can do so without even providing an estimate of the uncertainty is very unscientific.
But worse, as i noted above, the new system is at least partly dependent on student growth percentile scores based on standardized tests.
And though SGP is only being used for a small fraction of teachers (15%) at this point, that could easily be changed in the future (not least of all because it requires less time and effort than observations, which are still being relied upon at this point for the bulk of the evaluation) . When that happens, the randomness in the overall evaluation system will undoubtedly increase and the number of teachers “misclassified” along with it.
LikeLike
“It’s very hard to believe that one can distinguish a 2% difference between teacher effectiveness with any evaluation system…”
I didn’t state that very well.
To clarify: if the uncertainty in the teacher evaluation score is relatively large (noisy scores), it is possible that random “noise” might result in a change of 2% (or even larger) of the total teachers categorized as “partially effective” and/or “ineffective” (eg, from 1% to 3% of the total).
Though this change would obviously be very serious for the teachers affected, it would not mean that more teachers were actually “ineffective”. It would simply be a result of the noise in the system.
LikeLike
i said above that “it is possible that random “noise” might result in a change of 2% (or even larger) of the total teachers categorized as “partially effective” and/or “ineffective” (eg, from 1% to 3% of the total).”
But that’s not right. If it were truly random noise, a rating could go either way, so the percentage of the total would not change.
Never mind. 🙂
LikeLike
If the 3% of the teachers are secondary teachers who teach 5 classes, they would teach 15% of the secondary students (but not all day).
LikeLike
Reblogged this on David R. Taylor-Thoughts on Texas Education.
LikeLike
I hate to see this type of statistic used even to bolster a good argument. It gives credence to using statistics of this sort, which are so highly manipulable as to be either meaningless or truly harmful.
It would be easy for politicians to enact laws that would turn that stat right around. They’re working on it.
LikeLike