Improving Schools

by Harry on May 27, 2004

Excellent post from Laura about improving schools. She makes several school-improvement suggestions, in response to an article in the NYTimes arguing that all you need for good schools is good teachers and small classes. As Laura points out, the research on class size is completely inconclusive. I’d add two points. The first is that even if class size matters we have no reason to believe that there are no threshold effects; it may be pretty much as easy to teach 30 as 25, and much easier to teach 22, for all we know. Incremental across-the-board reductions in size are expensive, and may have miniscule benefits. Second, I have a feeling (based only on anecdotal evidence) that small classes, in making it more feasible for teachers to individualize instruction, may encourage them to engage in trendy, experimental, but ultimately less effective teaching methods.

All Laura’s comments are worth thinking about, and this one hits the nail on the head:

bq. I would also work on job programs for poor areas, on parenting classes, on adult education, and complete overhaul of some neighborhoods.

The idea that you can provide an excellent education for children raised in broken communities, with high levels of unemployment and family separation is, well, silly. Its no accident that the consistent high performers in PISA tests are countries with low levels of inequality and very low levels of child poverty.

The one thing Laura doesn’t point out (so I’ll add) is that it is incredibly naive to think that all you need is good teachers. How good a teacher is depends on 3 variables: i) that person’s qualities; ii) the fit between that person’s qualities and the needs of his/her students and iii) how well the institution as a whole is managed. Incompetent and/or ill-willed adminstrators are entirely capable of preventing much learning from going on. Think of the way that American public school lessons are routinely interrupted, without warning, by administrators who think they have an ‘important’ announcement to make over the PA system. Important announcements about the football results, or the availability of the school T-shirt or…well, you get the idea.



Robert Lyman 05.27.04 at 6:57 pm


Now if only the left-wing education establishment would adopt the views of the left-wing bloggers of CT (which happens to coincide with much of the right-wing critique of the state of education), then we could actually start educating children.

(Extra point: we should also lay to rest the myth that additional funding is all that matters–many schools have arguable too much money, and waste it so extravagently that they don’t have enough to educate children).


jdw 05.27.04 at 7:07 pm

I think the biggest challenge in reforming public education is that the system is so thoroughly busted that it’s difficult to come up with a real solution short of firing all the teachers and admins, burning down the education colleges, and starting from scratch, which means nobody’s getting any school at all for 5 years or so. The teachers of tomorrow are in education classes today, learning about the pedagogy of self-esteem education through togetherness training, and they’re going to teach the teachers of the day after tomorrow, who are then going to go to an even more dumbed-down education college. And the teachers have a union, which makes things even more difficult!

I think that this is one of those problems — along with the military-industrial-complex, hedonistic morality, and the ubiquity of lawyers — that we’ll only really be able to solve _after_ the Revolution.


Laura 05.27.04 at 7:30 pm

so glad that you mentioned the leadership factor, harry. I had in my post originally, but it got lost in some hasty editing. Research has shown that good principals and administrators can have a huge impact even in bad neighborhoods. Sadly, leadership programs across the country suffer from the same problems as the education schools. They suck, too.


freethinker 05.27.04 at 7:35 pm

The first poster is wrong about most shools have too much funding. Cause if they had some much funding then they would hire more teachers etc…

What is wrong is way schools are funded. Using property taxes to pay for funding causes poor areas to have poor school. We need to create a even play field for everyone.


Steven 05.27.04 at 7:56 pm

“… even if class size matters we have no reason to believe that there are no threshold effects; it may be pretty much as easy to teach 30 as 25, and much easier to teach 22, for all we know.”

Studies that purport to show class size doesn’t matter must be the work of statisticians moonlighting from the Bush administration.

Picture a class of 35 kids in a special needs, working class district. Both parents work, so the TV is a big part of their lives. This means less ability to sit and listen. Diet is poor — sometimes they come in without getting a proper breakfast. Yes, it happens. Some of the kids act up, mouth off, cause trouble. It doesn’t take more than two or three to consume big amounts of the teacher’s attention. There are a few other kids who are this close to understanding something, but they need a little extra time. There are the average kids, just getting by, who won’t get that extra bit of push because the teacher is burning off energy and time dealing with the laggards and the troublemakers. And there are the really smart kids who want the extra stimulation and could help lift everything to a higher level through their participation. Instead, they’re learning to coast and becoming alienated from the very idea of education.

Small class sizes are vital. Teachers know this. The trouble is, they get drowned out by backroom theoreticians with notions that giving them a manageable number of students
“may encourage them to engage in trendy, experimental, but ultimately less effective teaching methods.”

Are you kidding me?


teep 05.27.04 at 8:16 pm

My mother is a teacher. She teaches in Baltimore. You can go here and see some statistics for her school.

Mom was awarded a thousand dollars this past fall for being a superior teacher, first recipient of the award in the city of Baltimore. She gets observed by teachers from other schools because she is so good at what she does.

She is retiring in the fall because she has had enough of the “No child left behind” bunkum. In order to “prove” that every effort has been made to teach every single child, she is supposed to, starting next year, track every single discrete skill on every single child with a pre-test, measured practice, and a post-test. This would have to be be documented and laid open for inspection at any time. So, for example, if you were teaching the notion of a metaphor, you would have to pre-test to prove that the children don’t know what a metaphor is. Then you must practice (and DOCUMENT same) metaphors. Then you must post-test metaphors to prove that they have, in fact, been learned.

While this level of granularity does mean that one’s “teaching skills” are very easy to bean-count, it adds a level of drudgery and dumbing-down to one’s job. The amount of documentation that one must slog through these days is already insane. As an example of currently-in-practice documentation, my mother is not ALLOWED to fail a student until she has met face to face with the parent or guardian of the student. It doesn’t matter that little Johnny hasn’t been to school but 15 days in the quarter (there are 45 school days in a quarter, so he’s missed two-thirds of the available instructional time). It doesn’t matter if you’ve called down the truancy folks on his ass. You must meet, face to face, with little Johnny’s mother before you may assign him an F. So… you call his mother. Repeatedly. You set up times to meet, on evenings after school, on weekend mornings, on mornings before school. Johnny’s mother, who is less-than-coherent, never makes it to a meeting. Never. You’ve scheduled six. She’s made it to none of them because she’s a crack whore or something similarly responsible. You are still NOT ALLOWED TO FAIL JOHNNY. You must give him a D unless you can meet face-to-face with his mother and explain to her that her son is going to fail. You cannot do this over a cell phone. The RULES are that you have to meet face-to-face. I suppose that if he gets at least a D, he’ll not be left behind…

Here is part of an email my mother wrote to me last year:

This past Friday I received a copy of a memo from Dr. Frank Whorley. It informed me that I must recalculate my students’ grades. No longer is a 50 the lowest grade to be given a middle school student. In fact, oh wonderful day, the main computer had already changed, CHANGED, all the 50’s and 55’s we had given earlier in the year to 60’s. God bless them, every one.

Moreover, we were to now recalculate all previous grades with no test score being averaged as less than a 60. Get the drift here. Joe Student is a total non-worker and scores a 12% on a city written and asssigned Curriculum Based Assessment. We are to average in a 60% , as if the 12% never existed. These assessments, we were told earlier in the year, were to be 30% weight of a quarter’s grade. Now, Joe Student has a 60% on his big city test, thanks to downtown. The 12% never existed because we can’t enter it in our grade books and average it as less than a 60%.


That incident hit the news. The Baltimore City School District allowed that “mistakes were made” and that an “overzealous” person had been chastized. Every single Baltimore City teacher got that memo. Every single one. I do not know what the message to children is, when memos like this get sent out, but the message to Baltimore City teachers was pretty damn clear.


harry 05.27.04 at 8:24 pm


no-one argues that huge classes (in excess of 100, say) are as easy to teach as small ones (15) just as no-one says that classes of, say under 5 would be better than classes of 10. The issue is whether within the range of, say 15-40, there are obvious differences. As I say, careful social scientific research is utterly inconclusive on this — the range that most studies look at is about 15-35. We have experimental evidence that supports the claim that classes of 15 are more effective (measured in terms of test scores) in K-1, but the improvements yielded disappear in later grades (so the kids who got small classes early on do no better later on in test scores).

‘Teachers know’ cuts no ice with me, I have to say. Teachers like small classes, and so do parents and administrators. I’m interested in how best to get at, and deploy, extra resources, for the benefit of the least advantaged kids. So, for example, the 1997 pledge of the UK Labour Party to reduce all elementary class sizes to 30 was deeply wrong. It ignored the possibility of threshold effects and the lack of evidence in support of smaller class sizes, and, because it was across the board, did nothing to benefit the least advantaged relative to the more advantaged.

The scenario you decribe is all too real, and I wouldn’t argue for classes of 35. But, my anecdotal evidence comes from watching teachers with small classes indulge, rather than treating with appropriate severity, bad behaviour, and relinquishing ccontrol of the classroom for the sake of working with the kids in small groups. As with any strategy, this benefits some kids (the self-starters) and disadvantages others (the easily distracted) relative to a more teacher-centered classroom.

Before reading through the secondary studies on class size, btw, I shared roughly your reaction to them. But it is very compelling. The secondary studies can only say, of course, ‘not proven’ either way. But that matters. for any signifcant amount of money, if we were choosing between putting it into smaller classes in schools or into eliminating child poverty (eg by introducing a family allowance of some sort) I’d go for the latter every time. And if it had to go into smaller class sizes, I would advocate *much smaller* class sizes in k-1 in schools with low-income kids over incremental across-the-baord reductions.


Jason 05.27.04 at 8:44 pm

Look at universities as a role (OK, the student enthusiasm levels are different), a better teacher/better organised class are MUCH more important than size.

Of course, size is so much easier to measure.


Robert Lyman 05.27.04 at 8:46 pm

The first poster is wrong about most shools have too much funding. Cause if they had some much funding then they would hire more teachers etc…

First off, I didn’t say “most” schools have too much funding, I said “many schools have arguably too much funding.” And my point–perhaps unclear–is that excessive funding doesn’t necessarily result in hiring more teachers, or any educational improvment at all, but may get squandered in corruption or trendy nonsense.

An example of the difference management can make: the “best” public high school in Seattle (Ballard) spent $3,800/student a few years ago. Parents fought to put their kids there, for the top-notch facilites, good teachers, positive, safe environment, etc. The “worst” school (Rainier Beach) spent over $5,000/child, and at one point the students held a walkout/protest march because…there weren’t enough textbooks to go around. Both of these schools are far, far better than many Washinton, D.C. schools, which spend in excess of $10,000/child, but neither could match the local Catholic schools, spending $1,500-$2,000/child.

Note also that the Seattle public schools are funded from the same property-tax pot, so disparities in home values don’t enter the comparison, and the school board can fund based on relative need. Note further that Seattle has a complicated “parental choice” system so that students don’t necessarily attend school near their homes, which means that kids from poor areas might be brought to “richer” schools and vice versa.

My broad point is that funding isn’t the problem, even in poorer areas, except for a small number of extreme cases. Management of the funding is (or, at least, its one of the problems).


Xavier 05.27.04 at 9:13 pm

The focus in education reform should be based on marginal benefit/marginal cost analysis, not the pursuit of equality. If our top-tier students can more effectively use additional resources than bottom-tier students, then resources should be shifted toward the top-tier students. Equality should not be a significant goal of education policy.

I fully agree with Laura that due to social influences outside the public school system many students are incapable of being educated. But I think the solution is to simply accept that those students cannot be educated and stop wasting time and money trying.


h. e. baber 05.27.04 at 9:25 pm

I wonder why with all this dough shools don’t do dumb little things like having study halls where kids are made to do their homework. You could easily lock them into the school cafeteria for a couple of hours, with minimum wage workers on patrol. I’ve investigated and there’s no way to buy study hall for kids short of sending them to boarding school.


Robert Lyman 05.27.04 at 9:25 pm


Education, at least publicly-funded education, is about promoting social goods. It’s not clear to me what social goods will flow from abandoning those at the bottom. In other words, I think you may be focused in the wrong place on your cost-benefit analysis: while individual top-tier students may well derive more personal benefit from marginal increases in spending, society as a whole may get less, as the “abandoned” at the bottom turn to crime and welfare dependancy.

Your idea would make more sense if we abandoned the most “uneducable”–students who insist on, for instance, committing violent crimes in the schoolrooms–in favor of the next-from-the-bottom students, who have many disadvantages, and may not be strongly focused on their studies, but who are not actively hurting others or undermining the school’s mission. It seems to me there’s a lot that can be done with those students, and some huge marginal gains for society might flow from it.

Boy, does it feel weird for me to be arguing from the left on this one.


jdsm 05.27.04 at 10:02 pm

To follow up one of Harry’s points on class size, I think he’s right and I think from experience I know why. The teacher has three potential roles in the classroom, one as an presenter to say what everyone should be doing and explain stuff, one as a disciplinarian and one as a tutor, to give personal attention to the kids who need it. If you have ten kids you can do all of these pretty well. If you have fifteen it’s much harder to discipline them and almost impossible to give any real tutoring. The same is true of 40 kids. Your role is split between presenting and disciplining with no time for the tutoring which often really makes the difference. You may as well have a video and a bouncer.


DJW 05.27.04 at 10:04 pm

Robert Lyman,

Do you speak from actual knowledge about Seattle schools? If so, would you be interested commenting on why it is you think Ballard does better on less $$? It can’t all be about competent management, given that the principal there up until a year or two ago was widely known amongst the staff to be profoundly incompetent?

Just curious as to your thoughts–I’ve got friends who teach at Ballard and Rainier Beach (small world, this one), I’ll be seeking their perspective as well.

Harry–one thought your overlooking on class size. Even if in-class teaching dynamic is not improved by going from 35 to, say, 20, the quality of the out of class preparation, grading, etc. might be improved given the reduced workload for the teachers (and possible morale boost from that as well). I’m a lazy college teacher, the idea of 35 students per class, 5-6 sections, five days a week makes my head explode.


djw 05.27.04 at 10:05 pm

“you’re” not “your.” My editor has been fired.


Xavier 05.27.04 at 10:29 pm

If the benefit in putting underperforming students in school is that it helps keep them away from crime and welfare, then that should be the goal of education policy. We should try to improve discipline and vocational skills rather than standardized test scores. My point is that traditional academic achievement is not for everyone, and a government policy that fails to recognize that is not going to work. Our economy will always have some opportunities for unskilled labor. I don’t see why it’s so important that every child receives an education that would allow them to do more.

This isn’t the first time I’ve forced you to argue from the left, Rob. I often take a pretty extreme anti-egalitarian position. Someone has to take that position. Almost all of the debate surrounding No Child Left Behind is about the details of funding, standardized testing, and school choice. No one questions the underlying assumption that no child should be left behind.


Robert Lyman 05.27.04 at 10:43 pm


The numbers for per-student spending are from a Seattle Times (or maybe the P-I) article about the student walkout I mentioned. I don’t have time to find it right now but it should be online.

I myself am not a product of the Seattle system (having grown up in the suburbs) but I knew and know people who did attend them, and the consensus was that there was a good education there for the taking at all of them, and if you missed out, that was your own fault.

As for “why,” I wouldn’t want to hazard an all-inclusive guess. But one factor might well be family involvement and student effort: if one school is populated by a large number of students whose families fought to put them there, and another by students whose families were uninterested, the outcome would be pretty unsurprising. (And Ballard is newer, so it has a better physical plant, for what that’s worth).

Also, I think that Rainier Beach has more “at risk” kids, which would explain at least partially where their extra funding is going, rather than textbooks.

Please post what your friends have to say; I used to follow this pretty closely but I’ve fallen off since I went to law school in VA.


Michael G 05.27.04 at 11:15 pm

fascinating stuff…as a longtime teacher and current administrator at a public school outside of detroit, i only feel compelled to add one thing to the debate: teachers do not receive three months of summer vacation. check out your local school’s calendar and you’ll likely find that “summer” lasts about 8-9 weeks. sorry about being snarky…


Jonathan Goldberg 05.28.04 at 1:27 am

I’d like to make explicit a point make indirectly above: per-student expenditures can be drastically driven by problem children. The things we buy them these days are expensive, capital E.
This accounts for some (I don’t know how much) of the differences noted above; schools in good neighborhoods have fewer problem (“special needs”) children, and parochial schools often have none, since they have the option of refusing them entry. Public schools don’t have that option.

As the spouse of a teacher in one of the richest districts in the country, I’ve heard about what a lot of money can buy. If I were a parent in that district, I’d certainly rather have those things than not. Much of this is a large number of highly qualified special service people. It also pays highly for cream-of-the-crop teachers. And yes, the money helps get them.


Sam 05.28.04 at 2:14 am

In one of my other lives, I served five years on my local school committee. I negotiated a contract with teachers. It was excurtiating and almost came to a strike. Although I pushed then for the smallest pay increase in the county, I can say, with some certainty: teachers are not overpaid. The corollary to this is that, pace Laura, money does matter, and mightly so (just notice that four of the five of her highlighted points, after the first that asserts its not just about money, are, in fact, about money!). Just to get teachers salaries up to something close to what they deserve (it is hard work, harder than college teahing, my day job), requires big chunks of money. And to give them health insurance, has become extraordinarily expensive. For all of you who say money is not the problem, just look at colleges and universities: they are “controlling costs” by moving more and more to underpaid, unbenefitted adjuncts – and we all know what a great idea that is. Should that be the future of public education, too?

One more thing (I could go on at much greater length): class size does matter. It is the reason why I took my daughter out of public school, where she was not getting the attention and repititions of basic skills (yes, I am a basic skills guy who likes memorization – that is how I learned Chinese!) she needed in a first grade of 21 children. She was able to deflect the teacher’s attention away from herself by being a “good girl” but her reading and math were stagnating. When she was placed in a class of 10 (it costs a lot!) at a private school, she had no way of avoiding the teacher, and all of her basic skills improved markedly. I don’t know how you quantify such a dynamic for proper social science testing, but it illustrates a way in which small class size matters.


Laura 05.28.04 at 2:42 am

Sam, I never said that we shouldn’t pay teachers better, I just think that money alone won’t bring the best people into the profession.

After I got my Masters from the Univ. of Chicago, I took a couple of years off from grad school and tried desperately to get a job teaching. The obstacles were enormous — paperwork, union regulations, redundant education course requirements, the incompetent Board of Education. I eventually got a job, through a connection, teaching special education in the South Bronx. Did it for two years until ambition pushed me back to graduate school.

Sure, teachers should be paid better, but other measures (and perhaps costly measures) have to put in place, as well, to really attract the best.


Sam 05.28.04 at 3:11 am

My chief worry is that when we accept the “money is not the problem argument” we will find ourselves supporting, wittingly or not, those Washington politicians who refuse to fully fund IDEA or NCLB. We don’t know what adequate federal funding of public education looks like because we have never had it! They have never fully funded IDEA. In my small town, we just cut $119,000 from the elementary school budget and will have to cut about $700,000 from the high school budget because local taxpayers refused to increase property taxes to make up for the 20% cut in state funding that happened last year. Massachusetts faces a funding crisis in public education. And New York has a multi-billion dollar problem as well. It is no time to be saying money is not the problem. To my mind, it seems to be the the major problem, even recognizing the bureaucratic issues mentioned above.


kc 05.28.04 at 3:58 am

I’ve never responded to a blog before but this seems like an issue where at least I feel like I won’t be wasting people’s time. I’m a high school teacher, have been for 25 years. I’ve had small classes, 10 – 15 kids, and large ones pushing almost 40. Honestly, there hasn’t been a reliable correlation between class size and what I’ve been able to judge concerning the quality of the classroom experience for all involved, myself included. I’ve had small classes that were painful because they lacked a critical mass of inquiring personalities, and I’ve had large classes that hummed with energy and hard work. But there is one number that has consistently affected my ability and my desire to deliver the kind of instruction that every kid deserves. That number is the cummulative number of students a teacher is responsible for during any given grading period. Anyway you slice me, there isn’t enough to go around for more than about one hundred kids at a time. Beyond that I start losing track of names and people, and I start pretending to know everybody and pretending to care. It is dispiriting, to say the least, to work in an environment where survival dictates pretending on so many levels. So to sum up, it’s not how many kids are in a single classroom; it’s how many kids are in one teacher’s care.


John Quiggin 05.28.04 at 5:20 am

The weakness of the evidence on class sizes reflects the difficulty of measuring the benefits of schooling. Meta-analysis, experimental analysis, and comparisons of the earnings of workers who went to school in poorly-funded and well-funded states all support the view that, on average, more resources produce better outcomes.

And it’s striking to observe that private and public schools spend money in very similar ways. Well-off private schools go for small classes just as do well-off public schools. Given that there are presumably some competitive forces at work, this is hard to explain if small classes have no educational benefits. (I know revealed preference arguments aren’t much in favour here at CT, but this seems like a pretty good one to me).


Davis X. Machina 05.28.04 at 5:29 am

Anyway you slice me, there isn’t enough to go around for more than about one hundred kids at a time.

Interesting to note that Ted Sizer’s reform group, the Coalition of Essential Schools, stresses total teacher load — not to exceed 80 — and not class size.


Liz 05.28.04 at 7:20 am

I have two comments to make at this time.

Number one: Why is it that we talk about “schools” and “schooling” as if teaching environment for the k-3 age range is identical to the 9-12 (grade, not years) age range? It makes for really sloppy thinking and argumentation.

Early education needs lots more hands-on instruction. I could go on, but here is just one aspect: I’d venture to claim that there is more variation in maturity in the first-grade classroom than in the 5th grade classroom.

Number Two: Jonathan Goldberg talked about “problem children”. I don’t know what he meant in there–especially the “good neighborhoods having fewer special needs”–but I am assuming he means children with physical, emotional, intellectual or mental disabilities, all of whom are guaranteed:

Special needs students are provided rights to a free appropriate public education (FAPE) by three federal laws. These are the Individuals with Disabilities Education Act (IDEA) 1997, the Americans with Disabilities Act (ADA) 1990, and Section 504 of the Vocational Rehabilitation Act. These rights are to insure that all school children in the United States are provided with the educational opportunities to allow them to be independent productive citizens.

Well, yes, SpEd is expensive….but if isn’t that part of the social contract we have here in the United States about education? Mr. Goldberg, would you walk into a room with parents who pay property tax AND have a kid with severe disabilities–Down’s syndrome, say–and be able to look those parents in the eyes and say, “Sorry, your kid costs too much. We aren’t going to pay for the education that will help your child live an independent life.”?

I don’t have an answer to the SpEd dilemma, I just know what the law mandates.


DJW 05.28.04 at 7:46 am

KC, that’s exactly what I suspected, and fits with what I’ve heard from teachers. Studies about what goes on in the classroom miss an important aspect of class size–the cumulative, outside work effect (it’s hardly an issue that us college teachers are blind to when we talk about our own workloads, we really should pay attention to it in this case as well).


From Rainier Beach, I get the following response:

“My job is 90% disciplinarian, I just don’t have time to do much teaching.”

From Ballard:

“I never really thought my students were all that great, but this is the only place I’ve taught. They do, as a group, seem very eager to please, though–themselves, me, parents, imagined college admissions folk.”

That seems potentially consistent with part of your thesis.


Monica 05.28.04 at 2:27 pm

I think KC has hit the nail on the head (thanks for writing). Does someone familiar with the scholarship know of studies that examine total students per teacher, rather than size of individual classes?



David Salmanson 05.28.04 at 2:32 pm

Another comment on class size. In languages – which I don’t teach – class size is incredibly important. The teacher has to be able to hear each kid pronounce with some regularity. In history and English, it is less class size than total load. If I have two classes of twenty or two of fifteen and one of ten, I am still assigning the same number of essays but if I have a total load of eighty, I am not going to assign as much writing. (It takes me about an hour to grade an essay effectively: that is provide feedback on strucutre, argument and evidence, style and grammar).

Second, on “teacher-centered” pedagogy. Having been a pretty good college teacher, I am trying to unlearn this and fast. The brilliant teachers at the upper school I work at have found ways to let kids try things out in small groups or as individuals and still hold them accountable. This isn’t “fashion.” Face it, the only way to learn what history is – an act of interpretation based on available evidence – is to actually do some history and kids can’t do that if the teacher is doing it for them. Same goes for literary criticism. Model, teach technique, than get out of the way. I don’t do this enough in my classrooms and I see that my students are not as successful as those of my department chair. Of course, she has been doing it for thirty years and has a lot more practice.


Robert Lyman 05.28.04 at 2:39 pm

DJW–very interesting. Perhaps the people who study “school performance” should be looking more closely at the students than the teachers.

David–small groups and discussion works well for self-starting students who want to learn, but what about kids who are less motivated? Isn’t more rigid structure appropriate for them?


Tom Morris 05.28.04 at 2:42 pm

Schools? Pah! Give me a library any day…


harry 05.28.04 at 3:33 pm

I don’t know the literature on teacher-load, but there must be some. Intuitively it must matter a lot — and I’ve no idea whether class-size literature (including the STAR experiment — I’ll have to go back and look) controls for overall teacher load. Intuitively it makes a lot of sense that teacher-load would matter more than class size.

John, parents might demand small classes because they falsely believe them to be educationally beneficial, perhaps because they believe the widespread mis-characterizations of the results of the STAR experiment. Anecdotally I can tel you that I talk to a lot of college professors about quality-of-school issues and find them almost uniformly ignorant of the problems in determining the quality of a school. They frequently believe, for example, that a school with good ACT and SAT scores is thereby a good school. Which is nonsense. They else, almost uniformly believe that small classes are good, and when I ask why it turns out that they believe it because they have read some misleading representation of the results of STAR. They are much better attuned to the particular needs of their own kid within the particular situation she finds herself in.
Just to say, too, that the class-size studies I say are inconclusive are all framed in terms of effects on test scores. Maybe small classes are nicer to be in, which would be a reason for parents to want them, or maybe they make it easier for the parent to lobby for his own child (less competition for the teacher’s attention) which might be a good or a bad thing depending on the kid and how well the parent is clued in. I suspect that middle class parents like this aspect of small classes.

Thanks for arguing from the left, Robert. I should clarify that other CTers are not to be held responsible for my eccentricities. But also comment that in the UK my views about education are considered to be, fairly uncomplicatedly, on the left, whereas in the US, as you observe, they don’t fit well with the (much less left) left. Odd that.

I agree with Sam and Laura that money matters, and with you and Laura that management matters. Having disavowed locutions like ‘teachers know…’ I would add that teachers know that management matters — *teep’s* mother’s story is horrible, but not at all atypical. I agree that on the whole we should raise teachers’ salaries, but also that low salaries are not what account for the high attrition rates among teachers (especially good teachers) — in my experience many enter the profession with a sense of idealism (and often because they have a higher earning spouse whose income effectively subsidizes the schools), but leave because they feel undermined and undervalued by administrators. Not just at the school level but at the district level. In nearly ten years of (public school) teaching my wife has yet to be observed by someone who has, or even claims, any expertise on teaching; she has every in-service day wasted by incompetent and ignorant district or school administrators; her classes are shortened without notice, and interrupted frequently but irregularly (so she can’t plan for it) by administrators making frivolous PA announcements or walking into her classroom without knocking to ask an unnecessary question. She teaches in a high-spending ‘good’ district — a lot of that spending is simply wasted by mismanagement; that is incredibly demoralizing and accounts for a good deal of early attrition.

Noting all this about management I should say that I tend to be conservative about funding in my local (high-spending) district. I tend to oppose real budget cuts because I think the management is so bad that cuts would be made in all the wrong places. I tend to be sceptical of increased spending that doesn’t go to teacher’s salaries because I think management is so bad that the extra money would be wasted. I’m ok about teacher’s salaries going up not because it benefits me (though it does!) or because I think it benefits education much, but because I think that money can at least be some compensation for the disrespect and ill-treatment they receive from their managers and too many of the parents and kids.


riume 05.28.04 at 5:17 pm

So many points so little time…
On class size reduction. There is very little to no evidence that it matters in high schools at all. Common sense dictates that there is a level at which it does matter though. There is also one interesting study of middle schools by Harold Wenglinsky at ETS that uses a Lisrel model to get at the relationship between class size and a better discipline environmnet and the relationship between that environment and better performance. But its one study. Tennessee STAR, the only randomized experiment on a large scale is quite compelling. Particularly compelling is the “Lasting Benefits Study” add ons that show kids in small classes did better all the way through high school. That is a concentrated intervention, k-3. That research also shows that lower income students benefitted particularly from the intervention. All the meta analysis research that shows class size doesn’t matter, particularly the stuff by Hanushek, is subject to alot of criticism. His findings depend on what the meaning of a “study” is. Alan Krueger at Princeton is merely the latest to have done a good job of sorting through it. The point about faddism in teaching is well taken though. The otehr point is that scaling up a class size reduction is difficult because you need to have a pool of available qualified teachers to step in. California’s efforts in this regard stumbled becasue class size reduction implemented state wide created opening in the burbs that qualified inner city teachers took, meaning that the teacher quality problems in places like LA were exacerbated. Implementation is hard. Generating the political will to only do this in low performing or poor districts is also hard.

No one has mentioned prekindergarten or expanding kindergarten to a full day. After reading the reading research and the research on preliteracy and poverty – in particular the early childhood longitudinal study that shows how the achievement gap exists before school stops- I’m one of those hoping early childhood ed will pay off in a big way. And I’m not talking about “childcare” or daycare as it exists now. I’m talking about a prekindergarten curriculum with an emphasis on phonemic awareness, numeracy and vocabulary development.

As for tenure, in just about every state teachers are at will employees for their first two or three years. In hard to staff schools -like the one I taught at – you don’t see that option getting used, even though there is no impediment beyond managerial will. In part its because there isn’t a supply of fantastic young teachers willing to jump in. There are teachers in the system where I taught who failed the entry test multiple times and yet were hired and retained becasue there was no one else who wanted to be. (Harry’s point about administrative incompetence also comes to mind but I think that the first incompetence is in recruiting or in allowing defacto segregation by attendance zone). And there is no systematic effort to help these struggling teachers improve. Sink or swim. The real problem is that the most qualified teachers simply leave the poorest schools. In urban school districts at the least, the teacher supply problem makes any problem with the tenure system moot or at worst marginal.

In some places (not many, but some) the unions are actually taking this issue into their own hands, negotiating peer assistance and review programs. Toledo Ohio is a leader in this. The New York City union -after working with a pilot program on this for some time – put it on the table in negotiations this year. Its the law in California. I wish there were more enthusiasm for this by some unions, but its a positive reform.

Laura’s points about recruitment are generally spot on though. And reforming teacher education -either by working harder to line ed schools up with school districts they feed too (which is something you are seeing in some places) and/or by raising standards both of programs and entrants – is an important step. And I would say that rather than “making it easy” for second career teachers to get in, we need to find ways to “make it easy” for them to succeed. A lot of support before and during that first year is necessary.

The part of Laura’s post that calls for teachers to work a professional day as opposed to receiving overtime doesn’t gibe with the facts about teacher workweeks. Teachers are professionals and do not get overtime for working more than the contractual day. (Thinking back on my time in the classroom, that would have been sweet!). Teachers may in some instances get extra pay for extra work like managing clubs or coaching or somesuch, but that’s different. Moreover, teachers do work lots of overtime. According to the US Department of Ed’s Schools and Staffing Survey, in 2000 public school teachers spend an average of 48.6 hours a week working (in school and out). Secondary school teachers spent 51.7. Research that has had teachers keep work diaries to track out of school work confirms this finding. So yes, the required school work day is less than 40 hours (37.8 hours in 2000 again according to SASS)- but the real workweek is longer.

But there is something to the “professionalism” argument. Teachers should have offices of their own or shared, with computers and good internet connections in the schools. That is something I’ve seldom if ever seen outside of “teacher centers” in some NYC schools which are corun by the union. It would increase the professionalism quotient alot.

On poverty, healthcare. schooling and the achievement gap – check out Richard Rothstein’s new book at the Economic Policy Institute web site (

As for social promotion, my first disciplinary problem as a teacher was when I took over a class a bit into the year. I was the third teacher in a bit more than a month and that was because the kids had a lot of behavior and discipline issues. I had two weeks of hell to endure before the first marking period. It was ugly. Kids getting up walking around, coming in late, throwing things, talking, sleeping. I based that grade on the performance in those two weeks and failed three quarters of the kids. In part I did it to let them know that in fact I was in charge and also to let them know there was some accountability for their actions. And its not like the grade was part of the final record or any such thing. It was a statement that something was wrong and if they as individuals wanted to get something out of the semester they needed to readjust. I got no backup whatsoever from the administration that threw me in there. That was more years ago than I’d like to admit. Its amazing how little has changed on that front….


krebscycle 05.28.04 at 8:15 pm

Xavier and Jonathan Goldberg,

I hope you both have special needs children of your own some day.


John Quiggin 05.28.04 at 11:45 pm

Harry, I doubt that many people have based their beliefs on the STAR study, with the possible exception of a few who had shifted to an agnostic position as a result of the null results coming out of the “production function” literature and poor aggregate performance in the 1970s and 1980s. Such people may have taken the STAR study as evidence that their common-sense beliefs had been right all along.

I think the pattern of wealthy private schools offering small classes is standard across many countries and time periods, but I was referring specifically to Australian data from the 1970s and 1980s.

I’m prepared to accept that parents paying for small classes may be doing so for a variety of reasons such as that small classes are nicer, but isn’t that equally a legitimate concern for public education as well?

In the Australian context, a big expansion in education spending in the 1970s was followed by a rapid increase in school completion rates until the early 1990s, when spending was cut and completion rates fell.

There are all sorts of confounding evidence in time series, I know. OTOH, much of the force behind US doubts about class sizes and so on comes from the fact that test scores stagnated in the 1970s even as spending rose.

As implied above, I think that for a wide range of practical purposes completion rates are more significant than test scores.


Mike Huben 05.29.04 at 12:39 pm

I’m in a mid-career transition to teaching in public schools. (And I have two kids in public schools.) So I’m working on an MEdu at UMass Boston, which is fairly well respected as an education school.

That said, compared to my undergraduate courses at Cornell, the education courses are not very difficult or rigorous: the standards are lower. But there are several valid reasons for this.

First, passing rigorous courses has little to do with how good a teacher you’ll be.

Second, you won’t want to discourage good teachers with whatever portion of the requirements doesn’t apply to them personally.

Third, much of the content is a very soft social science.

Fourth, MA is in transition into requiring all teachers to be certified with MEdu’s, which means you don’t want to fail the many teachers that now suddenly have this new hurdle in their career.

The courses are mostly designed to change the ideas of the teachers from the viewpoints they developed as students. The most interesting point that I’ve taken away recently is that because students can turn off to anything they want to, based on whim, bad experience, distractions, etc., teachers must be continually recruiting interest in their subject AND NOT OFFENDING OR REJECTING students. We all can remember teachers that offended us or that we disliked and how we tried to ignore their classes and subjects. That’s why cultural relativism is so heavily stressed at UMass Boston, because cultural clashes are a major bone of contention between students and teachers.

There are a bunch of other things I’ve read here (and at associated blogs) that sound misleading to me.

People who complain about tenure and propose merit pay and firing of non performers as alternatives miss the basic point that those alternatives are worse: they lead to systematic political abuse of teachers and (by setting teachers against each other) disrupt what little bargaining power teachers have to raise their miniscule wages. One of the major attractions of teaching is the departure from corporation culture into a different kind of professionalization.


Elizabeth Dwight 05.29.04 at 1:34 pm

Where do I start with you folks? Perhaps I’m jaded at the end of a school year teaching at an urban public school, but my reaction to your class size debate is to point out the NCLU factor, aggravated by the MMF quotient. If I had more than 28 upper grade school or middle school students in my room at a time, what we in the business affectionately call “Middle School Stench,” would escalate so dramatically that we would be forced to lobby the (surely unresponsive) feds to increase our funding for the “Leave No Child Undeodorized” program. On a serious note, I can say that when I’ve been most effective as a classroom teacher, I worked within an efficient grade-level teaching team. By working and planning together, we can minimize the clerical stuff to teaching (one person is writing and making 50 copies of a parent letter, while the other packs up the science kit), we can flexibly group the kids into ability groupings for part of the day (or the more pc term “readiness groupings”) to better meet individual needs, we can confer on hard-to-reach children or the best approach to take with HMPs (high maintainance parents have taught me all I need to know about relentlessly advocating for my own children in a kind caring supportive fashion- don’t forget the chocolate on conference night), and together we can don our shades and a big floppy hat to hide out in a coffee shop consoling each other when the test scores come out and, big surprise! Not every poverty-level or middle class child we established a healthy rapport with, browbeat into reading every night, and mentored through the mysteries of the benefits for social adjustment of wearing deodorant showed growth on a random selection of short reading passages which seem to have been chosen for their startling irrelevancy to the limited background knowledge of our students. So,you all have perhaps more time (and interest) than I do in perusing THE STUDIES-
anything out there on the benefits of putting more than one educator in charge of a double or triple load of students? Can we let go of our quaint American pioneer notion that one teacher should teach one room of children behind one closed door?


Elizabeth Dwight 05.29.04 at 2:03 pm

Whoops! Okay, so that acronym is MSS not MMF. I guess I was thinking of its sister acronym “Middle School Stench” which still doesn’t explain the extra M. What do you expect from a psuedo-professional, soon to be “vacationing” (on 10 months of pay spread out over 12 months for seasonal work), ineffective, ambition deficient, “dumb” public school teacher?


harry 05.29.04 at 4:14 pm

bq. I’m prepared to accept that parents paying for small classes may be doing so for a variety of reasons such as that small classes are nicer, but isn’t that equally a legitimate concern for public education as well?

John — yes, I agree with this, and haven’t said otherwise (but can see how my remarks could be taken otherwise). In fact the subjective experience of school, given that it is something we force vulnerable people into having, matters a lot more than most of us think mst of the time. The issue is about opportunity costs — getting small classes across the board costs A LOT — and if you are an egalitarian like me you want to find ways of diveting resources to the least advantaged, and getting those resources used properly. (I promise a post in the next few weeks about why we should make schools nicer places to be even if that doesn;t raise scores and achievement at all).

I’m skeptical about completion rates counting for much in the US, where the school itself is the only detreminant of whether you complete (ie: it, and it alone, decides whether you graduate).

elizabeth — lots of things have come up here that I doubt have been studied and should be. Your issues are all about bad management (you function well when you are in a well-managed school, which it sounds like you have been sometimes, lucky thing). I’m struck by your last comment:

bq. Can we let go of our quaint American pioneer notion that one teacher should teach one room of children behind one closed door?

to which my response is YES. But also: this is just one way in which school managements are *SO* unimaginative. All of my (US) high school teacher acquaintances (though none of my UK high school teacher acquaintances) work in environments where no-one collaborates, even though many would like to. Crap management.

Again, I wouldn’t propose, at all, that we raise class sizes above 28, or above 25, or whatever, but just affirmed Laura’s impication that the evidence is inconclusive, and that there is a lot more wrong, and to do, than the NY Times piece claimed. Good teachers and small classes can be undermined by lousy managers, and they very often are.


Elizabeth Dwight 05.29.04 at 10:12 pm

Harry- Actually I wouldn’t give a whole lot of credit to management for any of my teaching success. On my own initiative, I can analyze the needs of my students, team with like-minded colleagues and together offer a comprehensive program to meet a diversity of student needs despite a lack of effective managerial direction.


John Quiggin 05.30.04 at 4:21 am

Harry, a discussion on making school a more pleasant/valuable experience would be well worth having. A lot of discussion of these issues treats school as if it is solely a preparation for adulthood rather than a major life experience in itself – rather as some religious discussion treats life as preparation for Heaven.


DJW 05.30.04 at 6:10 pm

John, that’s an excellent point for several reasons. A tremendous proportion of middle and high school students are, frankly, miserable, and that hardly seems like something we ought to ignore or treat as inevitable. My happiness and overall satisfaction with life took a huge leap forward when I switched from High School to College, which was also the same time I started taking school, studying and learning seriously. I imagine there are any number of reasons for this, many of which I don’t have the position to assess myself, but one of them certainly must have been not being depressed is one part of that story.

Harry, I see via the NYT announcement today you’ve got a Carnegie Corporation of New York fellowship for a project on justice and educational reform. Congrats! I’ve learned a great deal from this conversation and I hope you’ll consider sharing some of your thoughts as your research progresses in this forum.


harry 05.31.04 at 1:04 am

Thanks djw. I don’t get the Times — where was the announcement? I’m delighted obviously, and I, too, have learned loads from this discussion. Yes, I’ll try to float ideas and provide findings as I go along. This is the sort of feedback you couldn’t get by paying for it!


DJW 05.31.04 at 10:13 pm

It was an add, about 1/8 of a page, in the Week in Review section I think, listing and congratulating the 15 or so winners. Name, affiliation, and title of project were listed.

I’d tell you more but I think it got recycled.


djw 05.31.04 at 10:14 pm

argh, ‘ad’ not ‘add’. I always do that.


Wendy 06.01.04 at 1:53 am

Ont thing to keep in mind is that the public school’s expenses *include* several expenses of private schools. Public school districts must take on the costs for bussing and textbooks for any student with a disability, even if he or she attends private school. In my dad’s former district (he is a retired high school teacher), the public school district *paid* for a foreign languages teacher to teach at the local parochial school. So those costs were transferred to the public school district.


Catalina 06.01.04 at 4:12 am

This is an important debate but we would all do well to acknowledge the complexity of what it means to become educated as opposed to becoming “schooled” in a democracy. First, few of the posters have had direct experience in the classroom and so unwittingly fuel the debate with opinions that simplify rather than illuminate the complexities of public education. Second, it is hypocritical to hold teachers to some of the highest standards of any profession and pay them slave wages, underfund the schools they teach in and the school systems that ought to support their endevour, and apply the lowest acdemic cachet (in terms of recognition, and yes, funding) to schools of education where teachers are “schooled” rather than educated. But, rather than wring our hands and sling reproacheful and clever retorts at each other and the system, which some of you claim is busted, you ought take your outrage and organize your local communities, educate yourselves and others so that we, as citizens of a democracy, collectively put our money where our mouths are and work for a fully sustainable public education system that purports to educate all children not just our own or those of the people that look like us. Education policy, left to the politicians, creates inequitable outcomes. Although it sounds good to say that “we” ought to move aggressively into poor areas with massive social and other initiatives to “fix” those communities, “we” ought to ask the folsk who know their communities best what would work best for them first. The poor, the English learners and students of color and their families, upon whom the largest burdens are heaped with each newly legislated dictum, are the biggest losers in this so called public education system and debate. They, too, deserve a place at this and every other table.


Elizabeth Dwight 06.01.04 at 4:17 am

The effects of rampant depression is an important issue which relates to class size. This year I’ve seen more obviously depressed kids than ever. The skin cutting, the suicidal poetry they write, the mood swings… In my daughter’s school a 5th grader commited suicide this year! With a class approaching 30 students, it is difficult to make a connection with all of the kids. Granted we are there to educate, not counsel, but we also care about them and recognize that they aren’t getting the counseling they need, and can’t participate in academics without feeling safer and connected. The depression factor seems to correlate with underachieving giftedness as well. This year, my colleagues and I made tremendous headway with a classically underachieving gifted student with clear emotional problems (and a family history that explains much of it). He became able to participate, improved his relationships with peers and adults, wrote amazing poetry and historical fiction, and became visibly more together and even cheerful at times. However, when he took his standardized test, he did not show a full year’s growth in reading despite remaining in the 90th percentile. Should I be fired? Should the school be burned down? Should he be sent home to spend more hours in a tense, difficult situation? Or…should we perhaps find some additional measures besides a standardized test score to help us ensure accountability in education without throwing out the baby with the bathwater? And…to stray completely from the topic of class size, why aren’t teachers mandatory reporters about emotional problems like we are about suspected physical and sexual abuse. Every time there is a school shooting, I am certain there was a teacher who suspected emotional imbalance, but had no real recourse. In my district, we are told we can’t recommend counseling to families because then the district would have to pay for it. What is the social cost for that policy?


Brian 06.01.04 at 5:33 pm

I am currently in a graduate level teaching program at Columbia University…it would be very helpful to me if all the people who stated that research says this or that would please post their references. It would be very helpful to me.




Another Damned Medievalist 06.02.04 at 10:01 pm

Late entry here, but I’ve been away from the box. Where to start?
Parent accountability?
Teacher accountability?
Bloody government accountability?

Like Laura, I looked into teaching in secondary school. Not possible in Washington State — a phD in History and several years of teaching college freshmen and some experience teaching kids ESL count for nothing. I still have to have the MEd. And then, because of WEA rules, I have to teach whatever Social Science (because I know so much about Econ and Psych) the low man on the totem pole teaches, because someone with two classes in History with seniority wants to teach her favorite classes. That’s so good for the students — not!

I do think that overall, it’s student load, not class size, in 9-12. That said, I do a lot of small group work and Socratic lecturing in my classes, and that’s generally more effective with groups of 25 than 40. I think, from a student’s perspective, smaller classes are both better and harder — but the better is because students feel more that the instructor cares about them as people. This is important in college, but I think much more so in High School, and even more so in Middle School. Kids that age are really testing and trying to spread those proverbial wings, but they want a safety net. Again, I think that slightly smaller classes make it easier to assign interesting assignments and shift approaches depending on the class dynamic — this is much harder in a large class.

And teachers aren’t paid enough — although where I live they all make more than full-time adjuncts and the salary scales top out at higher levels than the top end for Community College tenured faculty, although Tenure-track positions generally start at a slightly higher wage.

As for money — when I was a student in the glorious pre-Prop 13 days in CA, one of my teachers told me that the property tax money went into a pool, and then was parceled out according to attendance, with some extra for inner-city schools. I believe the number was $144 ADA, but on the off-chance it was only $44 ADA, I checked an inflation calculator. At $44 ADA x 180 days, we’re talking $7920 per student in 1978 dollars. In 2002 dollars, that’s $22,607.

Oh and as for the person above who argued that not everyone was academically inclined? I don’t think we’re talking college, here – it’s basic skills that are killing us. Plain old reading and writing and maths, plus a background in history, civics, art, music (I know — what are they?), a second language and the sciences. If we made sure every young person (as much as physically possible) got a decent basic education that included instruction in how to live in society (cooking, cleaning, writing a check, changing a tire …), I think students would be more confident in their abilities and would not look at college as the only way to go.

Comments on this entry are closed.