Should we pay teachers more?

Libby Nelson reports:

It’s common to hear that teachers should be paid better — more like doctors and lawyers. In 2009, the Equity Project, a charter school in New York decided to try it: they would pay all their teachers $125,000 per year with the possibility of an additional bonus.

The typical teacher in New York with five years’ experience makes between $64,000 and $76,000. The charter school, known as TEP, would pay much more. But in exchange, teachers, who are not unionized, would accept additional responsibilities, and the school would keep a close eye on their work.

Four years later, students at TEP score better on state tests than similar students elsewhere. The differences were particularly pronounced in math, according to a new study from Mathematica Policy Research. (The study was funded by the Gates Foundation.) After four years at the school, students had learned as much math as they would have in 5.6 years elsewhere…

The gains erased 78 percent of the achievement gap between Hispanic students and whites in the eighth grade.

…The $125,000 number was eye-catching, but it was just the start of the school’s approach to teaching. Teachers were also eligible for a bonus of between 7 to 12 percent of their salary. The teachers, who are not unionized, went through a rigorous selection process that included a daylong “audition” based on their teaching skills. The typical teacher already had six years of classroom experience before they were hired.

Teachers at TEP also get more time to collaborate and played a bigger role in school decision-making than teachers in other jobs. Teachers were paired up to observe each others’ lessons and provide feedback, collaboration that experts agree is important but happens too infrequently. During a six-week summer training, teachers also helped set school policy.

There is more hereAddendum: Do read the comments, there are some excellent points in there.


It sounds like an artifact of the screening process. If you really wanted to be sure, you'd do the same type of evaluation as a larger school district and then randomize the teachers selected to teach in the school.

Presumably, the higher pay (and non-unionization) is what allowed the school to do screening. There was a post on here last week about how climate issues might be marketed to the Right []. I would characterize the opening line of this article, "It's common to hear that teachers should be paid better," as an example of how one might market school choice, charter schools, and merit pay to the Left. One certainly wouldn't want to emphasize the income inequality between the TEP teachers and the typical NYC teacher if one wanted to garner the Left's support.

Well it may not need to be marketed to the left at all, but rather to the parents. I think in most cases charter schools, which are growing at a good clip already, are free to pay the teachers as much as they want. However, it usually works the other way.

A relative of mine was a public school teacher in a rough neighborhood in a rough city. Her city "went charter" and in one year all the public schools were turned into charters. The for-profit charter company in-turn rented all the same buildings from the city and the teachers had an opportunity to interview for their old jobs. For those who were able to continue working there, they were paid about 60% of their old salaries with fewer benefits. Clearly something needs to be done to improve education and I really don't know exactly what role charters should play, but if it was as simple as getting out from under the union pay structure and letting the publicly funded private schools pay teachers more you would have seen it by now.

You would have seen it by now? Teacher unions are perhaps the single most intense and concentrated interest group in the country. Simultaneously, labor demand doesn't slope downward at the minimum wage and teachers aren't effected by competition - two 'special areas' where basic logic apparently fails.

Yes. Charter schools compete against each other all the time and have been around for a while. They aren't union shops.

They operate within a public school, unionized teacher marketplace. There are a number of more genuinely private schools and they do demonstrate value through the intense competition for admittance. In my view, if you want to argue that for some reason the supply and demand model doesn't work in education, the onus is on you to explain why.

Thomas, the vast majority of charter school teachers are not unionized. They receive public dollars and can almost always hire whoever they want at whatever salary they want. They only work within a public school marketplace from the perspective of the students. However, students do not always have a truly public alternative, as some districts have gone 100% charter. So, charters have competed for students and none of them have adopted a model of paying teachers a lot more than alternatives, but they do typically pay a lot less.

Jan, as Steve Sailor mentioned above, and as I assume you already knew, some private schools do pay a premium for teachers. This happens at the preschool, elementary, middle-school, high-school, and college levels. These schools typically charge high tuition and have highly competitive admission. They are enormously insulated from the public school/ union teacher environment by virtue of their excellence and exclusivity.

Charter schools and voucher systems are much more effected by the education environment. Simply enough, public education is analogous to price controls, and there is a vast, sparsely populated space between the heavily-subsidized public system and the elite system.

In a free market we find a spectrum of goods and prices, yet largely in education there is a bifurcated system of public and elite, without much in between.

Where was this?

Highland Park. A school district surrounded on all sides by Detroit.

Surely the point is that you pay teachers more so that you attract better people to be teachers? Is anyone really claiming that you can make teachers into better teachers just by giving them more money?

There's a difference between "should we hire a new set of teachers offering a competitive salary and hire them based on merit and expected performance" and "should we start massively increasing the salaries of current public school teachers" but you can imagine how this will be interpreted.

@Jeff--very true. But also the study shows, again, 'people respond to incentives'. My pet peeve is lack of innovation due to the low rewards for inventing, by and large. If you pay money to people to invent, they will invent. If not, not. People respond to incentives, it's pretty basic.

Don't underestimate the effect of a strong screening process and a culture of exalting the craft of teaching. Yes, yes, money helps. But I would hazard a guess you could achieve two-thirds of these results just with screening and school professional culture. The vast majority of public schoolteachers aren't very smart, and aren't very well-educated. Just weeding out the imbeciles and the time-card punchers produces a vastly better pedagogy. Of course, getting them weeded is greatly assisted when you can offer that large a salary increase.

I'm for tripling public schoolteacher pay, right after we make teaching the hardest profession in the country to get into. Our teachers need to come from the tops of their classes, rigorously screened and murderously winnowed down to the elite. This is the most notable difference between our culture and those with really excellent schools. They draw their teachers from the best, and they have a corresponding high cultural standing. Status is more powerful than cash. Doctors and lawyers have status not only because they are relatively well paid, but because it is very hard to become one. A schoolteacher? 80% come from the bottom quarter of their college class. And everyone knows the education majors, they sit with the communications majors and the the humanities. It's what you go into if you're a nice girl, and can't hack nursing school. Their status is correspondingly low.

Top of the class in what? The day to day running of a classroom is not an academic skill, but a craft. Techniques that can be learned and practiced to handle the various situations that arise.

I love how this article and discussion abstract away the reality. Decimal points in the tests score improvements. We wouldn't want to get our hands dirty with the real details involved in helping a child to learn to read for enjoyment or to be conversant in math, or to have a working knowledge of where the things he sees everyday come from.

You'll notice I mention the power of the craft of teaching early on. But it is not enough to teach this craft (although we'd be better off if we did). If we want better teachers, we need to be able to offer more capable people an incentive to be one. This requires both money and prestige. The more of one you can offer, the less of the other you need. Plumbers can make an obscene amount of money, but they have to, because they get no status. College professors can be paid next to nothing, and the status will make up for it. Teachers get a moderate amount of money, but they are slowly seeing their status eroded. This is because the entry bar is so low and the average quality of teacher is terrible (plus some really poor political choices by teacher's unions). The lower their status drops, the fewer really capable people will try to become one, and the ranks will be filled by the sort of people for whom steady government work is the best they can do in life. It's a death spiral, and unless it is reversed, we will see the complete dissolution of public education at some point.

You want reality? This year, 70% of high school graduates will go to college. This means that roughly a third of college students are below average intelligence. And the vast majority of our public schoolteachers in ten years will be from this group. IQ is solidly correlated with educational attainment, and while this isn't exactly the same as the ability to then teach the subject, it is a prerequisite. Necessary but not sufficient. So yes, we do need drastic changes to pedagogy, we need a better way to teach the craft of teaching. But we also need better raw material, because making the dullards of one generation the teachers of the next only produces more dullards.

You bring up a good point, but I think the current role of the teacher involves multiple roles that should be divided up. Throw out the entire concept of 1-teacher-1-classroom. This is beyond archaic.

Doctors and lawyers are supported by armies of nurses and paralegals. The brightest of our teachers who actually study how their students learn and design curriculum should make up the doctoral level teachers, while a more abundant class of (lower paid, BA) teacher aids provIde the man power of supervising students, administering exams, enforcing discipline and helping identify students having trouble with material so they can receive additional attention. The doctoral teachers are going to be fewer in number and operating more at the level if an entire age group, while classrooms no longer need serve as the basic unit of organization and can be more fluidly managed by the teaching aid staff. Add to this picture a statf of specialists in psychology, learning methods, and information technology, and you soon have something more resembling a hospital or law firm.

This hasn't penetrated the public consciousness, but in the US its hard to become a teacher now. I have had several friends try to break in. And the main reason actually is the unions. Turnover is so low that there are not many openings.

It may be hard in places to get hired as a teacher, but in terms of qualifications (degrees, testing etc.), it's far, far easier than any other profession they seem to want to be compensated similarly with. I agree that unions make it impossible to fire bad teachers, and the generous pensions and nonexistent physical requirements make for long careers.

While in the school in this story, there was a turnover of 50% from year one to year two for the teachers. Including some teachers being asked to not return.

"The vast majority of public schoolteachers aren’t very smart, and aren’t very well-educated."

This is untrue. The average high school teacher SAT score is 580 in their content field (math for math & science, English for history & English). Both are considerably above the average for college graduates. Elementary teachers are at 500 V, 480 M--slightly above and slightly below the average for college graduates.

Anyone who thinks teachers are stupid is operating from a major knowledge deficit and isn't qualified to opine on education policy. Not that this ever stopped anyone before.

By the way, there's no research showing smarter teachers leads to better academic outcomes.

By the way, there’s no research showing smarter teachers leads to better academic outcomes.

Good. Let's hire food-service workers for teaching jobs, dirt cheap.

I don't find 580 to be very good, especially in one's content field. So while I'm glad you have figures on this question, the figures aren't very reassuring.

And if that's the "average," what is the standard deviation?

500 V and 480 M are both below the average for a college graduate according to your own linked figures.

I'm curious how this is reconciled with the GRE results for Education degree students, which are at the lowest of the low end.

Many of the crappiest education majors drop out, raising the average of the remainder.

You can see how Ed majors stack up against the other majors. Hint: when looking for "education," start from the bottom of each chart and read up.

Some of the worst students are education majors. In Illinois many education grads need multiple tries to pass basic language and math tests to become teachers. Just paying them all more is a stupid idea.


"I don’t find 580 to be very good, especially in one’s content field. "

Well, then, you're pretty damn silly. 580 is in the top 30-40% of all SAT takers, and in the top 50% of all college graduates, give or take. There are 1.7 million high school teachers and these days middle school teachers have to pass the same test as high school teachers, so that's an even larger number of people who are getting that average. You can't populate a teaching profession with teachers getting 650 or higher on every section without pulling smart people from other professions. And given the weak to non-existent relationship between teacher academic ability (once past the basement, see more on that below), it's absurd to start cannibalizing from other professions just because you're a snob about SAT scores.

Mind you that's the average of 11 states on the Praxis. It doesn't include California, which notoriously has difficult licensure tests in every profession, including teaching. It does include the states with the lowest cut scores on the Praxis. So in all likelihood, the average score of all high school teachers is higher than that.

No report on the standard deviation that I can find. However, if the standard deviation is huge, it would go in both ways, suggesting that there are lots of teachers with scores higher than 580. More likely, though, the deviation is fairly small, because the African American and Hispanic failure rates on the tests are a pure bloodbath. In math and science, something like 60% of blacks fail the tests in most states I've looked at--and that's the population that has already gotten through the easier tests (CBEST in California, Praxis I in others). The cut scores are kept a standard deviation below the average in most states (as is true for bar exams and most licensure tests). My guess is that the SAT score needed to pass the tests in math is about 500.

"Good. Let’s hire food-service workers for teaching jobs, dirt cheap. "

That assumes there's no basement, which is a monumentally stupid assumption. The logical inference is that a basement exists, and we haven't found it yet--but that we're above it. If we weren't, the research would have found more of an impact.

Selecting teachers by their test scores is like selecting coaches by how good a player they were.

How good can a coach be if he has never played the sport? What if a teacher likes the idea of teaching (education major) but doesn't know any subject well?

Evidently pretty good:

That assumes there’s no basement, which is a monumentally stupid assumption.

No, chum. It assumes I took you at your word.

I said "teachers", which means the research involves the intelligence of people who have been accredited to teach.

"What if a teacher likes the idea of teaching (education major) but doesn’t know any subject well? "

These days, a teacher can't major in education without passing the credential test at some point prior to graduation. Most ed schools have chosen to stop committing affirmative action:

And a lot of high school teachers didn't major in education. Those who did usually majored in something else as well.

I am pretty sure the central argument a lot of people are making is that we should pull smart people from other fields into teaching. That would be the reason to pay more, for instance.

I'm sure that is it. But it's a silly central argument. Anyone who thinks further sculpting the teacher pool from the top 50-75% to the top 75-90% will change outcomes hasn't ever taught advanced math or literature to the average, much less the below average, high school student.

Look at the chart in the linked article. The number's are fishy. The much touted math impact remains at 0.3-0.4 for the first three years, then in the last year the gains quadruple discontinuously. Similarly the english results have no statistical effect in the first two years, then shoot up by year three out of the blue.

It's hard to imagine a legitimate mechanism causing this effect. If kids moved from crappy teachers to really good ones, you'd expect quite the opposite. They would quickly pull ahead initially, then the gains would level off as the superstar teachers pushed them closer to their upper-bound of potential. It'd be much more likely that these type of results are driven by selection bias, statistical manipulation or even cheating.

@Doug--maybe there's a three year learning curve, literally?

good point !

And, after looking up the wiki, I might add, that this is

1. a result of just 7 turbo teachers, of which "half" are told to have quit after the first year alone,
2. you can be sure, that the 120 kids were somewhat (self) selected
3. the effect seems to be more like 0.5 years, or 12 % in total extrapolating the first 3 math years to 05,
and 0.4 and 0.6 years for english and science

Coming from a country with working free public schools and universities (we also ended the experiments with tuition last year), Germany

I see nothing to get even a little bit excited about

Bringing up Germany doesn't help your case at all. As the number of German B.A. titels inflate, their market values go down. Enrollments in private schools are sky-rocketing.

Studies show student achievement drops any time that student switches schools. That should at least partially explain the pattern you describe.

As expected from vox☭, the original study isn't linked on vox, and when I go to the mathmatica study, I find that the original study pdf has been 404ed away, however, I found it on

That graph on vox shows the gains in "years of learning" a classic lib ploy to make gains look bigger than they really are. The original study reveals how they were calculated:

This figure converts the effect sizes presented in Tables IV.1 and IV.2 to years of learning using the
following benchmarks derived in Bloom et al. (2008) for average annual student achievement gains: from
the end of grade 4 to the end of grade 8 (for four-year impacts)—0.40 standard deviations (SDs) in math,
0.30 SDs in ELA, 0.30 SDs in science; from the end of grade 4 to the end of grade 7 (for three-year
impacts)—0.42 SDs in math and 0.32 SDs in ELA, from the end of grade 4 to the end of grade 6 (for two year impacts)—0.49 SDs in math and 0.36 SDs in ELA; and from the end of grade 4 to the end of grade 5(for one-year impacts)—0.56 SDs in math and 0.40 SDs in ELA. Effects are normalized such that theaverage annual achievement gains made by comparison students during the period of treatment equal zero.

We can calculate how many SDs these gains are that way. Their proudest achievement, improving scores 1.6 years over what they would otherwise be is 1.6 years*(.40SD/years) = .64 standard deviations. But that's the outlier. Look at the three year improvements in math 0.4 years*(.42SD/years) = .168 SD or the 4 year improvements in ELA .4 years(.3SD/years) = .12 SD There was no improvement for the one year cohort in ELA, I guess my headline will be "No improvement in fifth grade ELA scores from supposed miracle school."

The .64 SD gain would be impressive, if it were repeatable, which I see no reason to assume that it is. It's an outlier, the typical results are around the .1-.2 SD range, nothing that can't entirely be applied to sorting. We all know that in any of these lotteries it's the smart(er) parents who enter their smart(er) children into these lotteries. And the study does not control for the dumber kids leaving, only claiming there is no evidence that the school expelled problem students. I'm sure if the teaching was as rigorous as proponents claim some of the kids, not wanting to fail out would have transferred. But suppose these gains were real, what's to say it's because the teachers were paid a huge amount of money? Maybe smart(er) children benefit from being in a classroom with smart(er) children away from the dumb(er) children?

We will see study after study about miracle school after miracle school, but I guarantee that in 30 years you'll still be told that the Solution to the Problem is right around the corner, if only you give our school a little more money, if only you apologize a few more times from the crime of being White.

*the typical results are around the .1-.2 SD range, nothing that can’t entirely be explained by sorting. -

"After four years at the school, students had learned as much math as they would have in 5.6 years elsewhere…" - the point being there was statistical improvement, but it was not radical. It was not 10x but incremental. Your points may be valid, but it's also valid that people respond to incentives, in this case teachers to teach better. Even better would have been to give students some money for learning more.

Those kinds of SD gains are common among these kinds of "miracle" charter schools. What's to say that it's the teachers being payed better? Sure, incentives matter, but ultimately once you get to a certain level of competence how much can the teacher do?

Think about the effects of paying teachers like lawyers or doctors. In addition to the obvious financial problem of paying for it, you'd have people who would otherwise be doctors or engineers teaching pre algebra in middle schools, nothing better for our 21st century economy than that, right?

The statistical test to be made is on the outcome for students under such system, Not on the effect of pay in the system.
The exchangeable part is pupils results.

As for the economy, it is not obvious that there marginal impact of a doctor is above one of a school teacher. It all depends on what you do with both, and the school teacher has the advantage of being able to trigger many new doctors vocation over the years.

That said I completely agree with the culture of excuse and all the political motivation. Beside proponents of such culture are generally the one who sell out the common good, respect for science and work first.

Isn't measuring it in standard deviations misleading? A sigma of 0.1-0.2 sounds trivial; but "SD" you mean standard deviation of the test scores among individual kids. Since the sample as 120 kids, the signal to noise ratio better than 0.1 * sqrt(119) = 1.1 sigma [2]

Of course the effect is still small. On average, the better teachers shift the median kid into what would have been something like the 54th percentile. The shocking thing is that if the 0.1 SD/per year calibration is true[1], then roughly 85% of kids are already TWO YEARS ahead or behind the median.

List of dubious assumptions above:
[1] Pretending everything is Gaussian.
[2] Pretending kid-to-kid variations are independent.

Clover, thanks for finding the study. I think the charter school runs on the same budget as other New York schools. Also, it seems the students are not selected because they are likely to be successful but the opposite.

It is just chance, small schools will show a big result in one year and be unable to duplicate the result. Educational results are almost all controlled by the student. Applying for a charter school requires an unusually high level of parental involvement.

thanks for the link. I agree that this is probably a fluke like all of the other small studies.

What are teachers at elite private schools in Manhattan paid these days?

My vague recollection is that teachers at, say, Marlborough in Los Angeles averaged around $100,000 a half dozen years ago.

But since I wrote this, I've heard that the average at a top SoCal private school is a lot lower than I thought. So don't trust me!

But I did hear this evening third hand that the principal of the private school pays himself a half million. And is worth it because he routinely lands seven figure donations.

CEO markets in unexpected industries.

My wife, after retiring from a 35-year career in banking, consulting and executive-level management took a job teaching chinese at a very good, but not elite, private high school in the northeast. She was paid $48,000 per year. At the elite school down the road, the salaries were even lower.

It's much more pleasant to teach in a private school classroom, thus cash wages are lower.

But what is the value of this private school benefit... It could well make up for more than the difference. Thus implying private schools pay more even when payroll is less.

There are schools where teachers accept the lower pay because it makes it easier to get their own kids in (tuition-free, too, which is probably an IRS issue).

Anecdotal, but teh schools might simply be hiring more people rather than paying more to fewer. I know a "teacher" who is a nutritionist specifically for the sixth grade in a Manhattan private school (at $65k/yr). If they're down to that level of granularity with staff, I doubt they can afford to pay them all a ton.

Yeah, private school teachers get paid less with less job security, I'm pretty sure all research shows.

I live in an area that has tons of highly regarded private schools, and for years I tutored kids that went to these schools. The teachers aren't terrible, but they aren't extraordinary. And every so often, they are just as awful as the stories you hear about public school teachers.

I'm happy to have Bill and Melinda Gates pay teachers as much as they wish. As long as it's Bill and Melinda paying them.

Why not have a sample that prioritizes merit and removes unions to see the result. Then compare with a similar group that has higher pay to isolate the effect of pay.

Generally, I'm not sure the teacher is as important as people think, when compared with the influence of engaged parents and a curriculum that is rigorous and demanding.

Asians for example aren't doing better because they keep lucking out with great teachers. My kids have attended schools in the west and in Asia. They consistently do well in Asia because they're expected to - by us and the standards set by society.

I just love how Asians are proving wrong 30 years of political propaganda.
It is way worse in Europe which has much more "equality redistribution mechanism" and yet way, way, more excuse culture.

"Nearly half of first-year teachers didn't return for their second year, either because they resigned or because they were not rehired"

I think it's important to know what proportion of this is teachers resigning.

Alternative title- "Average is Over for teachers"

Think the range of the question is too limited, too myopic. Would argue that monetary code itself is an insufficient cultural coding mechanism for calibrating complex network relationship value information.
Would add that our cultural genome, that is, the set of coding structures that the cultural network uses for its interface with reality, is complexity inadequate.

“The most fundamental phenomenon of the universe is relationship.” Jonas Salk
“There were 5 exabytes of information created by the entire world between the dawn of civilization and 2003; now that same amount is created every two days.” Eric Schmidt, former Google CEO

Code is fundamental infrastructure for complex relationships.

Our moral, religious, legal, etiquette, and monetary codes (cultural genome), can't process the onslaught of complex network relationship information w/ sufficient speed, accuracy, and power. This contributes significantly to externalities, whether obesity, income disparity, extinction, climate change, infrastructure decay, etc.

Imagine a biological genome that ignored myriad relationships in its environment. That would be a genetic set unlikely to pass the natural selection test. Think that's what we have monetary code, and our cultural genome in general.

In the transition from simple hunter-gatherer social structures to the exponentially more complex information architecture of city-states, we added writing, legal, etiquette, and monetary coding structures to our cultural genome. Arguing that we need to do this again, get fundamental about complexity at the cultural level.

For more, please see:

I truly cannot determine whether the above is automatically generated or merely incoherent. Therefore I cannot click on the link.

Ah, forum comments and spam. It's like Blade Runner, but in a text only universe. I need Voight-Kampff.


Even if this were entirely what the blurb makes it out to be (I suspect there are a lot of factors and biases in play) is there any evidence that this would work for all teachers, or that there are anywhere near enough super high end teachers to do this kind of thing in all schools?

From all the teachers I know, their incomes are pretty good, usually better than mine. I never thought "I don't want to be a teacher because pay is too low" but always thought the burocracy and legal atmosphere coupled with parent issue and lack of support and follow-through from school boards and management would make it a completely terrible experience.

If they're looking at teacher pay they're looking at the wrong thing, IMO.

I haven't seen anyone point out that, even if the conclusions are valid, this is not broadly repeatable because:

"The typical teacher already had six years of classroom experience before they were hired."

It also states in the study that only 2 of 42 teachers hired had fewer than 3 years teaching experience.

In other words, the salaries weren't drawing in highly skilled teachers from other professions. It was just cherry-picking among the best ones who were already teachers. If you merely raised all teachers salaries tomorrow, you'd be left with roughly the same set of teachers as before, including the ineffective ones.

I was about to post the same thing but you said it first and best.

I was going to make the same comment. Nuts, you beat me to it.

I was even going to illustrate it with a point on why this is not an economics website based on the comments by giving the following example"

Assume a bell shaped distribution of teacher aptitude. Each teacher in the distribution earns $64k. Now, offer teachers who go through an interview process where they can discuss their past success in teaching (probably also have a list of references and awards), and also demonstrate it as well.

1. Who will be attracted.

2. What part of the distribution will apply and be selected.

Tail selection problem.

"The typical teacher in New York with five years' experience..." has tenure if they are in the public system, which vests in NYC after a 3 year probationary period. NYC public system teachers with tenure are well nigh un-fireable. Any public school teacher who moves to a charter school is giving up tenure, with very little chance to return to the public sector and get it back.

Many teachers are risk averse, so any public school teachers in this study would have a strong appetite for economic risk or be from such wealthy backgrounds that economic risk is not an issue.

Or, they are confident of their abilities, or that the prospect of $125k beats out tenure at $64k. But, you have a good point about the cost, er, value of tenure.

Of course, some could be teachers who were going to quit anyway, for which a charter represent a less-risky bet.

Hmmh ... high quality public school teachers frustrated with the education status-quo sounds a LOT like the Teach For America crowd. I wonder how many teachers in the sample started off as TFA?

If the study school is just skimming the cream off the high-human capital endowment, high family wealth TFA's the results will not be reproducible in a wider context. Just like almost every other edu-awesome achievement study.

Or, how first year TFA's would be accepted into the TEP program. Probably none. So, what you have, as you pointed out, and I agree with, is a program whereby the public system trains or selects the first round, and the TEP system gets its pick of the second round.

Agreed, except for the part about having the "same set of teachers as before." Teacher salaries tend to be higher than those in the community that support them, and increasing their salaries (and related) pension benefits by 50% or more will almost certainly result in far fewer teachers than before.

Reply fail, was responding to Dave.

"The teachers, who are not unionized, went through a rigorous selection process that included a daylong “audition” based on their teaching skills."

From a pool of teachers, the most skilled are selected and then paid more and the results are better. From a pool of unpaid junior hockey players, pick the best ones and then see if they beat other teams of randomly selected unpaid juniors. What will the results be?

How did they select students and did that population change any?

We should definitely pay teacher better, whether that means more on average, I don't know. I think that teacher like many public sector workers are paid too much of their compensation as deferred compensation which probably increases the premium for seniority. Also apparently it is difficult to pay teachers more for scarce skills or teaching in less desirable schools and situations. Also, I doubt the utility for paying more for additional training per se. If more training results in better performance, that should be recognized. Finally, I doubt the utility of "education" as a major component of teacher training as most learning of how to teach happens on the job. Certification should occur after a few years of classroom experience, not as a requisite for teaching.

All public sector employees should annually bid for their jobs, with the low bidder getting the position. That's how it works with everything else, paper clips, computers, bus maintenance, etc. Giving people more money doesn't make them better at what they're being paid for, giving Josh Hamilton $25 million a year only got the Angels 10 home runs in 2014.

Let's start with execs in the Fortune 500 and hold them to a vote by shareholders.

The Board of Directors is elected by shareholders. Do you have a point, or is this just inchoate flailing?

I have represented at least 5 Fortune 500 companies, know their execs and boards, and can tell you that you are living in a fantasy world if you believe board members are "elected" by the shareholders. The "election" process begins with the selection process. Whoever "selects" elects, and it aint shareholders. Its management and a few other board members.

Sometime do some research and find out how many times a person nominated by the board was rejected by the shareholders.

How come new companies haven't gotten to some better corporate management methods, like proportional voting by the shareholders to elect a board?

I largely agree that in the long term boards go rogue from shareholder interests, but why don't the shareholders do something about it while they are still in charge?

Most ridiculous comment on here so far.

Tell me how many private firms hire staff with the low bidder process?

I bet it is a very, very small minority. Obviously the Angels do not.

Private firms aren't required to hire the low bidder for anything. They are, after all, private, and they can do as they wish with their money. The public sector is very much different. Ordinarily, when a government agency needs a new building, motor vehicles, furniture, etc. they are required to accept the specified product from the firm that offers it at the lowest price. Why should it be any different for individual employees? If an applicant for a job offers to do it with no fringe benefits, for instance no health coverage, or for a lower wage than the incumbent, why shouldn't they be hired?

Chuck, comparing skills honed over time by employees to purchasing inanimate objects like paper clips is illogical.

The extra money paid to the teachers would get better results if the money had been paid directly to the students who got the highest grades.

It is academic performance by the students themselves that is most likely to respond to financial incentives. It works in every other field of human activity.

This was the argument made by Roland Fryer, Harvard economics professor and chief equality officer at the New York City Department of Education from 2007 to 2008. He undertook several randomized controlled trials where students were paid for performance. Financial incentives for students did not improve test scores, but they did improve behavior.

Did they examine how it affected academic achievement beyond test scores?

Like what?

Arnold Kling would ask whether this study rejects the null hypothesis. That is, does it tells us anything about how to improve life outcomes for students.
I'm not sure there's been enough time elapsed to measure that, but I don't see definitive proof to reject the null hypothesis.

"It’s common to hear that teachers should be paid better — more like doctors and lawyers."

The highly paid lawyers work in the private sector, those that are employed as prosecutors and public defenders or in the attorney general's office are recipients of a form of attorney welfare. There's probably a kind of doctor in the same situation, at the VA, Public Health Service, CDC, Bureau of Indian Affairs, etc.

There are certain lawyer positions that are low pay compared to private practice but very high prestige, like USAA.

We need more immigrants to go into teaching if anything teachers should be paid way less. Plus teachers are basically morons they should be paid in scraps.

I skimmed the comments so may have missed if someone else also makes this point. Most of the reaction here is about incentives (pay) for the teachers but I didn't see anyone commenting on some of the organizational changes -- the broader role of teachers (and I assume the elimination of administrative staff -- meddling middle management?) and whether that may not have had as much impact of the outcome as the pay did.

Repeating what one commenter said in a slightly different way, if the production process is poor -- high error rates in the units produced then selecting better line workers and paying them more may well result in a higher quality output form that production line but the trade-off may well be in fewer units actually getting to the end of the production line. The improved QA activity by the better and more incented worker shifts the fail rate in the line from say 5% to say 25%. Is that approach really an organizational structure that best meets market needs? This would be the nice try but no real improvement here for society in regards to education.

If, on the other hand, the process is improved and the quality of the units produced increased from 25% error rates to 5% error rates because the production line is a better technology the pay incentive may have been marginal in producing the results -- though one might still argue that the productivity originates with what the teachers are coming up with so the pay deserved.


Before we alter the administered price for teaching services to sluice mo' money to yet another Democratic Party client, can we please alter the institutional architecture of primary and secondary schooling, recruitment methods, and financing?

All of these have been pointed out, but it's worth putting them all in one place: the "improvement" was absurdly small, incremental improvements. To declare improvement at all is to use statistics in a way that most people would find....unconvincing.

"The gains erased 78 percent of the achievement gap between Hispanic students and whites in the eighth grade"

When you look at it closely, this will turn out not to be true. Either the achievement gap didn't exist between this particular group of students (low achieving whites, high achieving Hispanics), or they are talking about the percentage that met a particular baseline, which was unusually low in eighth grade.

Teacher pay doesn't make a difference. The auditions aren't designed to select excellent teachers, but rather teachers who adhere to the teaching style desired by TEP (presumably either no excuses or progressive, I'm not sure which). The teachers worked less than the average teacher--they had two prep periods a day. They got bonuses, so presumably they didn't get bonuses for the two years their scores were lower than average.

Almost certainly, the huge improvement in the third year is due to something simple: they convinced the kids to try on tests, maybe. Or they cheated.

"“The gains erased 78 percent of the achievement gap between Hispanic students and whites in the eighth grade”
When you look at it closely, this will turn out not to be true. Either the achievement gap didn’t exist between this particular group of students (low achieving whites, high achieving Hispanics), or they are talking about the percentage that met a particular baseline, which was unusually low in eighth grade."

The population was 89% Hispanic and 9% African American. So they didn't measure any Hispanic-white achievement gaps at all, and don't have any reason to think that white and Asian students wouldn't have also shown similar gains. They just mean that in one year there was an outlier gain in one subject, math, driven primarily by the classes of a single math teacher of 0.64 standard deviations, and the general Hispanic-white gap is 0.8 standard deviations. So the theory is that if Hispanic students all got outlier math teachers, and white students got average ones, then the gap would be reduced by that amount.

It's possible the super math teacher cheated, taught more to the test, or motivated students more on a low stakes (for them) test, but maybe he or she was actually a superstar. Intensive tutoring often increases math scores substantially, even a large fraction of a standard deviation. Why couldn't a top teacher have similar effects to personal tutoring?

"Intensive tutoring often increases math scores substantially, even a large fraction of a standard deviation. "

Not anything that closes the achievement gap, or anything close to it. Kids who tolerate intensive tutoring probably want to learn, which is also true of kids who self-select for a school like TEP, or any charter.

If the teachers were exceptional and the root cause of the improvement, why were the kids worse than average the first two years? It's not that suddenly they got better. Curriculum doesn't cause that. Something else changed.

As I've said before schools are a horrible way to convey information to young people. Successful education needs to be student-driven with rewards (earlier graduation and entry into the workforce) clearly conveyed to students. To the extent teachers are to differentiate themselves, they need effective control over the curriculum and environment rather than national-level common core micromanagement. Teachers need to be selectable by students on the individual level.

Education in the US is already too expensive, but much of the cost is for administration. No personnel above the school level are useful. There may be a lot of waste in construction and procurement as well. If we can get administrative costs down, I'd be OK with teachers being paid more as long as it isn't publically subsidized. The student teacher ratio might need to be increased, either by increasing class size or having classes meet with their teacher less often. Ideally, students would just read textbooks and access equipment like Bunsen burners on an as-needed basis.

Lawyer pay has decreased for many lawyers, and lawyers are increasingly seen as unnecessary middlemen. We also need to cut medical pay or increase the number of tasks that are done by other than doctors. I have more respect for plumbers that make an honest living than doctors hiding behind medical licensing or teachers that coerce their students into attendance.

Doesn't Eric Hanushek's work show that you can get close to those results just by firing the worst teachers?

I wonder if people on the left would generally accept the compromise of doubling teacher pay scales while eliminating unionization and tenure altogether. Rather, do they think that is a net improvement?

Probably not, it would pull a lot of smarter, more qualified White people into the profession, who would push out the NAMs.

No they would not. The point of unionization is political power not improved education. And they would no doubt be highly skeptical of these claims anyway, as am I.

Actually, there are very few black or Hispanic teachers in high school academic subjects, precisely because the credential tests keep most of them out. While there are more elementary school black and Hispanic teachers, it's still a relatively small number. Blacks are about 10% of elementary/middle school teachers, and about 6% of high school teachers (which includes special ed and pe, relatively low-scoring majors)

"People on the left" are mostly anti-teaching these days. But doubling teacher pay scales is a moronic idea. It wouldn't improve outcomes, would still be hard to fire teachers, and unions don't really do much except keep pensions high. Easier to fix that with a law that covers all government employees than double teaching salaries. As I observed above, the teaching pool is as intelligent as it needs to be already--probably more than needed in elementary school.

We always seem to tap-dance around the real problems. Of course there is lots of pressure to raise teacher's salaries, and a lot of claims that raising salaries will automatically lead to better education outcomes - the teacher's unions see to it that the debate is framed this way.

Having just put a kid through 12 years of public school in Canada (where the schools are supposed to be good), the problems were obvious - and shocking:

- Bad teachers have a disproportionate impact. My kid was a straight-A math student until running into a lazy, vicious teacher who did little but berate the kids and tell them they should pick non-science careers because they were all too stupid to handle math. Many of the kids chose to take math by correspondence rather than be exposed to that poisonous teacher. Our complaints to the administration were met with a knowing nod, and "Yes, yes, we've heard about her many times. Unfortunately, there's nothing we can do." Other bad teachers were simply lazy, or incompetent. There are a lot of good teachers, and a few great ones, but the few bad ones can be devastating if your kid's path through school intersects with too many of them.

- Institutional laziness. Kids here are no longer allowed to see their unit exams - only their score. Homework is not individually graded by most teachers. Parents aren't allowed to ask about their children specifically at parent-teacher interviews, and are also not allowed to see the completed exams. I'm guessing this policy is the result of the internet - schools and teachers like to save money and effort by re-using exams, but if the kids can take them home they can post them on the internet and make the exam useless for the next semester. Rather than deal with this in a better way, they just keep the exams a secret, meaning the kids and parents can't know where the gaps in knowledge are. Engaged parents are critical to education, but it's hard to be engaged when the school refuses to give you information, when your kid doesn't even know what he got wrong on an exam, homework isn't graded, and where attempts to discuss your child with a teacher are met with indifference and/or stonewalling.

Here in Alberta, teachers make well above average salaries - a teacher with 10 years' experience makes $84,000, plus generous retirement benefits. This while working about 500 fewer hours per year than the average salaried employee. And yet, I've found that if anything the general quality of teachers has declined over the past couple of decades - probably because of union rules that prevent new entrants into the field while guaranteeing lifetime employment to anyone who makes it to permanent teaching status. Union systems like this have no place in white-collar work. Not only do they prevent the kind of filtering and churn required to build and keep a high-quality workforce, they actually act as disincentive for the best teachers and an incentive for the worst. Every chair warmer in a union job puts additional burdens on the people actually trying to achieve something.

Another problem is the closed-shop nature of teaching, which prevents older people with experience from entering the field. Instead, the ranks of teachers are largely made up of people who have done nothing but take a 4-year education degree and then 'teach'. Many of them are shockingly ignorant about the subjects they are teaching. And if my college is any guide, the education faculty is where people land if they can't hack it in the other faculties. So we start with something less than the best and brightest, then put them in a job where there is no reward for excellence, where it's almost impossible to be fired for laziness or being a bad teacher. We should not be surprised when this results in poor outcomes.

If you want to really see the effects of higher pay, instead of giving them all $132,000, announce that next year the bottom 10% of teachers will be fired, and their salaries will be split among the top 20%, with everyone else remaining at the same level. Then see how teaching behavior changes.

Nonsense. The unions are not the main problem here. The de-professionalization of the teaching profession is. Why do people want their children to become lawyers or doctors or businesspeople? Mainly because they will make a good salary. It's not that being a business person is in some way better than being a teacher -- except for the paycheck. If you want to attract great people to become teachers, you need to motivate them to put up with the bureaucratic pressures, long hours, and condescending attitude of the community. One way is by compensating them financially. Pay people better and create a competitive market of much better professional teachers.

The teacher governance and collaboration model are what appeal to me. I would work at such a school for less than 125k.

Comments for this post are closed