Cheating Fate

Pretty much all the evidence shows that demographic and socioeconomic factors have huge implications for student achievement. This fact, in turn, sometimes leads people to take a very strong "demographics is destiny" view of kids' outcomes. That, however, is far too simple. At my reunion over the weekend, I met up with a good friend who now works at the Excel Academy in East Boston and they get very impressive results:

Excel Academy students — 64 percent of whom are Latino and 73 percent of whom qualify for the Federal free and reduced lunch program — outperformed their local and state peers as well as their state-wide Caucasian peers on the 2007 math and English MCAS exams at every single grade level, thereby reversing the achievement gap. Furthermore, Excel Academy’s eighth graders were ranked third out of 280 Massachusetts school districts and fourth out of 461 Massachusetts public middle schools, placing them in the top one percent statewide.

Adequate resources, deployed correctly, can achieve great results even with disadvantaged kids. The trouble, however, is that when you look at the schools that are having success it's not as if they're slight variants on the education we're giving to most poor kids -- it costs more money, it involves more hours, it's more staff-intensive, etc., etc. Basically, to make up for the disadvantages that come from being disadvantaged, you need to give poor kids more and better instruction than middle-class kids need. In this country, however, our general practice is to give them less and worse.

Comments

I admit I don't know anything about Excel academy. So I have this question - is it cherry picking the bright disadvantaged kids or does it have blind enrollment? Taking bright disadvantaged kids gives them a chance they wouldn't have a definitely deserve, but disqualifies the school from being a model of education for all low income underserved communities. Blind enrollment, however, would definitely prove the theory that more money, more time and better teachers work.

Posted by Evan | June 10, 2008 8:47 AM

Based on their web-site, they appear to be cherry-picking - they have an applications process and they advise that "space is limited." I would expect that they don't consider kids with a history of behavioral problems. They may also consider past academic success.

And note that Excel requires every family to sign a "Family Accountability Contract" that imposes a high level of parent or gardian involvement. There's an implied threat of expulsion for failure - the family and student must "accept the consequences" for failure to abide by the "Contract."

The biggest problems for public schools are (1) kids with families that can't or don't provide support, and (2) disruptive kids who eat up time, resources, and energy. Excel seems to leave both classes of kid in the ordinary public schools.

I know your just a blogger, Matt, but this is your livelihood, after all. Make an effort here. Ask your friend about Excel's admissions and retention policies, and get back to us.

Posted by Bloix | June 10, 2008 9:03 AM

is it cherry picking the bright disadvantaged kids or does it have blind enrollment

Presumably the students at Excel (or their parents) have to have sufficient ambition and initiative to enroll there. Whether this kind of self-selection qualifies as "cherry-picking" is, I suppose, up for discussion.

I don't know about the name, though. "Excel Academy" sounds like a weekend workshop in the use of spreadsheet software.

Posted by James Gary | June 10, 2008 9:08 AM

I'd be curious to see the results from a school where the disruptive students were sent home and/or failed. Would the non-learners drag the average down for everyone?

Posted by freddiemac | June 10, 2008 9:14 AM

Even cherry-picking's better than nothing: I'm sure every inner city has a good number of bright kids who would otherwise never get much in the way of resources directed towards them.

The question I always ask is, is this widely replicable? That is, can we taxpayers make this happen elsewhere by opening our wallets to roughly the same degree elsewhere - is the pay good enough to get the quality, motivated staff to make it work, or not? And if not, why not - could more money do the trick, or are there just not enough such teachers to make it work if you took this nationwide?

If it isn't replicable, then it's useful for that particular set of kids, but it doesn't advance the discussion about how to help bright inner-city kids generally.

Posted by low-tech cyclist | June 10, 2008 9:15 AM

Cherry picking is very, very, bad if the point is to evaluate what works for the majority of disadvantaged students.

We need random selection from disadvantaged populations, varying levels of resources applied to selected students, and a statistically valid analysis of the results.

This is how we do science, and really, anything else is just storytelling.

Posted by jvoe | June 10, 2008 9:21 AM

disqualifies the school from being a model of education for all low income underserved communities

Low income communities have many bright children who are disrupted by underachievers. So this type of school can actually be a model for all low income communities, but they might not work for every student. Some kids will need more discipline (just like mid and high income communities).

Posted by Owen | June 10, 2008 9:24 AM

I don't know about the name, though. "Excel Academy" sounds like a weekend workshop in the use of spreadsheet software.

Hey, that's going to really important for some kids.

Posted by Captain Noble | June 10, 2008 9:25 AM

This Excel Academy seems fine, for what it does, but it's hardly unique. There are plenty of schools which do a good job of educating poor and/or minority children who are academically inclined and whose families are committed to education. Indeed I suspect our system adequately serves the vast majority of students in that category. Our problem, nationally, is not those students, but the students who are not academically inclined and whose families are so dysfunctional or culturally alienated that they don't support their children's educational careers.

Posted by y81 | June 10, 2008 9:34 AM

Well, I hope that what y81 is saying is true about the ubiquity of inner city schools being able to well-serve the poor but academically inclined minorities.

The school system is optimized for students who come to school well fed, have few mental/behavioral problems, and have a home where they have enough personal space and peace and quiet to do their homework. Ensure all students have those things, and the school system will produce better results.

Meanwhile, though, cities that want to keep families from moving elsewhere will serve their needs by providing attention and resources for the academically inclined and motivated.

Posted by Tyro | June 10, 2008 9:43 AM

One teacher in a room with 30 or more children is so 18th century...its time for a change.

Posted by jerri | June 10, 2008 9:43 AM

What y81 @ 9:34 AM said.

NPR had a piece Monday morning on a charter school in Colorado Springs that emphasizes character and discipline and reports very high test scores. The problem is its student body is mostly white and affluent--so we really learn nothing about whether what they're doing transfers to more challenged communities and populations. The result was a very frustrating story.

I think it's fine to have public systems that allow parents to make some choices and magnet schools and even non-profit charters. (DairyStateKid#2 will go to such a charter for 6th grade in the fall.) But any system that takes care of the motivated families also has to have intensive resources for that families y81 identifies above.

Posted by DairyStateDad | June 10, 2008 9:49 AM

I echo the comments about cherry-picking made above. I'd also point out that, according to the website, Excel has a 9-hour school day, compared to a 6 1/2-hour day in my kids' suburban middle school. The last 2 hours & 20 mins. seem devoted to academic support and homework club so it may be optional but, even without these minutes, the schedule still has 20% more instructional time than a typical school. That's huge, both educationally and money-wise. Lengthening the teaching day by 20%, even without considering the academic support obligations, would be enormously expensive, underscoring Matt's point that it's about the resources, stupid.

Posted by mert7878 | June 10, 2008 10:06 AM

Reading is fundamental.

Parents, read to your kids.

There is no reason why any kid can't start reading before the go to Kindergarten. Remember, it's just a matter of learning what sound each letter makes.

Posted by Swan | June 10, 2008 10:16 AM

Basically, to make up for the disadvantages that come from being disadvantaged, you need to give poor kids more and better instruction than middle-class kids need. In this country, however, our general practice is to give them less and worse.

It's politically impossible to give disadvantaged kids more and better instruction than middle-class kids. Middle-class (white) parents simply won't stand for ghetto kids getting more resources than their kids.

Posted by David | June 10, 2008 10:35 AM

If you don't bother to check the incoming characteristics of the students, you're talking through your hat. I've heard the same anecdotal crap for decades and it keeps going nowhere.

Posted by gcochran | June 10, 2008 10:36 AM

This area is where I have become less of a liberal than I used to be. There is no end to the 'give disadvantaged more and better education', as there is no downside to 'giving the underclass or "disadvantage" less and worse.' We need to define a strong, well-supported middle ground in public education, supplement it with some extras for those who need it and then let the educational chips fall where they will. We need good schools -- but for many, probably most of these kids, the difference will not be in the eduction per se but in the milieu: mainly, do their parents or parent have a good job, health care, decent place to live in a safe neighborhood. Attempting to launch small numbers of disadvantaged youth with super and expensive educational opportunities is the tail trying to wag the dog.

Posted by Chris | June 10, 2008 10:37 AM

This area is where I have become less of a liberal than I used to be. There is no end to the 'give disadvantaged more and better education', as there is no downside to 'giving the underclass or "disadvantage" less and worse.' We need to define a strong, well-supported middle ground in public education, supplement it with some extras for those who need it and then let the educational chips fall where they will. We need good schools -- but for many, probably most of these kids, the difference will not be in the eduction per se but in the milieu: mainly, do their parents or parent have a good job, health care, decent place to live in a safe neighborhood. Attempting to launch small numbers of disadvantaged youth with super and expensive educational opportunities is the tail trying to wag the dog.

Posted by Chris | June 10, 2008 10:39 AM

The biggest problems for public schools are (1) kids with families that can't or don't provide support, and (2) disruptive kids who eat up time, resources, and energy. Excel seems to leave both classes of kid in the ordinary public schools.

This is a crucial point. Catholic schools in NYC have great results with disadvantaged kids via an extensive scholarship program, and they spend less per pupil than NYC public schools. It's not about cherry picking "bright" students or any special Catholic approach to education -- it's about self-selection of kids and parents who are committed to education. So the problem with public education in poor communities is not just lack of resources. I don't know what the solution is, but ignoring the deeper social and cultural problems that make public education in poor communities extremely difficult will not help kids in those communities.

Posted by Led | June 10, 2008 10:54 AM

Genes have considerably more impact than demographics (?) or socioeconomic factors. Plenty of adoptive studies have demonstrated that both IQ and income are largely heritable - JR Harris, Holt/Sacerdote, etc.

Posted by Chris | June 10, 2008 11:03 AM

"do their parents or parent have a good job, health care, decent place to live in a safe neighborhood."

None of these factors matter much. If good jobs mattered, you would have seen radically depressed academic performance in American kids living through the Depression. Japanese kids who lived through fire-raids that burned down their cities would all have become dum-dums: it didn't happen.

The worst medical care in the US today is a hell of a lot better than the best available in 1940: yet if you look at test scores in Iowa, which go back that far, there has been very little change. Which simultaneously invalidates all the talk about how education has gone to the dogs.

Almost every commonly stated notion about education on either the left or the right is easily shown to be false.

Posted by gcochran | June 10, 2008 11:09 AM

"Adequate resources, deployed correctly, can achieve great results even with disadvantaged kids."

Sigh, no amount of resources will be enough if the kid's parents, or in most case, parent or grandparent does not care about the kid's education.

Why do liberals refuse to acknowledge this basic truth?

You could send every "disadvantaged" child to MIT with around the clock tutoring thrown in and still it wouldn't matter if the child's parent doesn't care and becomes an active partner in the educational process.

You can damn well bet that the parents of the children at this Excel Academy are active participants in their child's education.

If my tax money is to be used in the providing education to other peoples' kids, I would rather the money be spent on parents who will make sure that their child is educated, rather than speading it on parents who won't.

Posted by Chicounsel | June 10, 2008 11:14 AM

IOW, Chicounsel is down with punishing kids for picking the wrong parents.

Posted by Davis X. Machina | June 10, 2008 11:23 AM

I live in Boston (area), and I believe Excel cherry picks to the extent that you have to want to get in, there's the commitment expected from parents and students, and they aren't afraid to lose students who won't abide by the rules. They are not picking by test scores, but they are picking by commitment to education.

While not a perfect solution, frankly I have no problem seeing that a school that selects for "wants to be there" would do better than one in which classes are regularly dragged back to the level of the most disruptive student. The issue then becomes: What do we do with students who don't want to be there, whose parents say "hey, I have no control here, the schools are supposed to fix them"? I don't have a good solution; I suspect the answer might be to focus resources downwards (Head Start, extra help in Elementary school) and accept some losses as you go upward.

I did a few presentations for my kids' classes this year (1st and 6th), and one thing that struck me was how engaged they were--they listened, didn't giggle and talk, asked intelligent questions at the end that showed they'd listened to what I said. I live in a fairly affluent suburb--parents move here for the schools--and every back to school night there is at least one parent, usually two, for each tiny chair. I'm sure there are disruptive students and disengaged parents, but they're in the minority. So if Excel basically selects to make itself more like the best suburban schools--parents strongly emphasize education, are involved to help with homework etc--I can see where that would help. But not that it's a bad thing.

Going off on a tangent--a friend's son (in another state) has Asperger's, so he needs special services both for being gifted and for being handicapped. She is a lawyer and knows EXACTLY what the state is required to provide, and is willing to go toe to toe with the teacher, principle, and on up to ensure he gets it. It struck her that if you were an immigrant parent with limited English this would be lost to you, and I think that's true in a lot of poorer urban settings--I've heard it from the NYC schools--that just to master the system and make it work for your child, it helps to be fluent in English, highly educated, etc. That's where a lot of the above goes off rails--not in dealing with the problem kids, but that for a normal kid getting into a good school isn't just a matter of showing up, but of your parent understanding whatever school choice is in place and how to work the system.

Posted by Deborah | June 10, 2008 11:54 AM

Led @ 10:54 AM:

So "self-selection of kids and parents who are committed to education" isn't cherry-picking? That seems like a distinction without a difference.

Chicounsel @ 11:14 AM:

I'm not sure why the problem is with "liberals". The mandates in NCLB that require all schools to advance achievement for all students are going to most punish the schools with the greatest proportion of students whose parents or grandparents do not care about their education (among other things).

And what Davis X. Machina @ 11:23 AM said.

Posted by DairyStateDad | June 10, 2008 12:00 PM

Calling Steve Sailer! calling Steve Sailer! Discussion of poor black and hispanic kids and school reported on M. Yglesias site. Respond immediately!

Posted by cw | June 10, 2008 12:00 PM

I stated how a simple scientific study could establish how resource application affects student performance. Matt's nice story really tell us nothing about whether resources affect performance--and that is the truth.

In this debate the left or right will ever advocate real rigorous science. Both sides are afraid the findings might upset their deeply held beliefs.

So students who could get help don't, and those that don't use the help they get, get more. This debate in our country has become a debacle.

Posted by jvoe | June 10, 2008 12:10 PM

Excel is a charter school; how does it spend more than other schools? How much more do the teachers there make than at regular public schools?

Tell me, which party works to limit the number of charter schools in Massachusetts? Is it the party that Matt cheerleads for?

Posted by Thomas | June 10, 2008 12:11 PM

I stated how a simple scientific study could establish how resource application affects student performance. Matt's nice story really tell us nothing about whether resources affect performance--and that is the truth.

In the education debate, the left and right both fear real rigorous science. Both sides are afraid the findings might upset their deeply held beliefs.

So students who could get help don't, and those that don't use the help they get, get more. The education debate in our country has become a debacle, and both sides are guilty of making it so.

Posted by jvoe | June 10, 2008 12:12 PM

Tell me, which party works to limit the number of charter schools in Massachusetts? Is it the party that Matt cheerleads for?

So? Is it now against the rules to try to pressure your party to change its positions?

Posted by Adam Villani | June 10, 2008 1:35 PM

So "self-selection of kids and parents who are committed to education" isn't cherry-picking? That seems like a distinction without a difference.

I said it wasn't cherry picking "bright" students. It may be cherry picking motivated students, although the phrase is a little inapt because it's mostly the parents that are doing the picking.

Posted by Led | June 10, 2008 1:35 PM

In my opinion, Bloix and Led are basically right on the money about this. I am a public school teacher and both my father and mother were public school teachers, as well as several other members of my family. Having a well-funded school is important but nothing is as important as having a student body with stable well-adjusted homes. Trying to fix that deficiency is far easier typed than done. Obama gingerly touches upon this issue, saying, “Parents need to turn off the TV and video games and make sure that your child does their homework”. This needs to be said a lot more.

To that end, let me step out of my professional capacity for a moment. It seems like the left can’t seem to link the school system’s struggle to the breakdown in the American family. The right can but doesn’t propose anything to solve the problem, it just rants about gays, abortion and Hollywood. So, the problem persists.

Posted by Raindog | June 10, 2008 1:58 PM

To get away from the pejorative connotation of "cherry-picking" and back to the point - Matt's post is a comparison of school performance statistics which he uses to draw a conclusion: that "adequate resources, deployed correctly, can achieve great results even with disadvantaged kids."

This may or may not be true. The problem is that we can't tell if it's true because we do not know how Excel selects its students. And even if Excel took all comers, we would face the problem of self-selection: it's likely that the children who apply come from more motivated families. Who is responsible for their better performance? The school or the families?

What we need to do is to determine the relevant characteristics of a random sample of Excel students upon entry (family life, income and educational attainments of parents, prior academic performance, behavioral issues, health status, etc.) and compare the performance of those students over a period of years to the performance of a sample of public school students with identical characteristics.

The result would tell us whether Excel does better than public schools in educating students with those characteristics. Maybe yes, maybe no.

Unforunately, it would tell us nothing about how best to educate students with other characteristics. E.g.., if Excel screens for behavioral problems, then finding out that Excel does a better job in educating kids without behavorioral problems tells us nothing about how kids with behavorial problems would do in an Excel-type school. Still, a good study would let everyone know whether the extra costs of the charter school were actually worth it.

A charter school that really wanted to know how it was doing would set up a study from day one. But the school districts don't require it, and no charter school administrator wants to incur the costs and the risks.

Social science research is slow, incremental, and expensive. As Matt shows, pundits and policymakers generally don't care for it and prefer to generalize based on their preconceived ideas buttressed by a few bits of out-of-context data.

Posted by Bloix | June 10, 2008 2:09 PM

If stable, well-adjusted homes made a difference, test scores would have tanked when the divorce rate soared in the 1960s.

That did not happen.

Posted by gcochran | June 10, 2008 2:21 PM

I think we all can agree that stable homes, motivated parents, access to good health care, nutrition and a stimulating pre-school environment will all contribute mightily to greater educational achievement. But, these are difficult, multi-source social problems that are hard to understand, let alone try to fix given the political and economic realities. By contrast, schools are publicly run, limited-mission institutions that everybody understands, or thinks they do [we all went to 2nd grade, right?]. Superficially, at least, they should be subject to control. That's why we focus on schools; they are the obvious target. I think Matt's point is that, assuming we are not going to level the socio-economic or cultural playing field any time soon, the only real chance of addressing the achievement gap is through more resources.

Posted by mert7878 | June 10, 2008 2:50 PM

Like others, I lament the lack of a scientifically controlled study of what works, though that's pretty hard to do when kids are the guinea pigs. But, if you look at any successful charter school or other educational initiative that has resulted in real, sustained gains in test scores, all of them -- even with their cherry-picked students or limited enrollment -- have dedicated more resources. Think KIPPS. Even anecdotally, I don't know of any charter school that has achieved similar results without more resources.

And, also, as far as replication goes, let's not forget that charters attract the most motivated teachers too. I'm not sure there are hundreds of thousands of similarly motivated would-be teachers hanging around with nothing to do just waiting for their chance to close the achievement gap.

Posted by mert7878 | June 10, 2008 3:01 PM

"Basically, to make up for the disadvantages that come from being disadvantaged, you need to give poor kids more and better instruction than middle-class kids need."

The funny thing is how lots of people think that way about K-12 but nobody in America thinks that way about college.

Matt's parents paid a fortune to send him to the most famous college in the world with the smartest, highest paid professors. And all their friends complimented them on how proud the much be for Matt to have been so smart and motivated that he was one of the select few to get into Harvard. (Of course, in reality, Harvard professors aren't chosen to be the best teachers, they are chosen because they are, or were in the past, the best researchers.)

The logic of this so obvious that few bother to spell it out: The smartest teachers should teach the smartest students (i.e., with the highest capacity to learn) because more total learning is imparted that way.

When thinking about colleges, everybody can see that Matt's doctrine for K-12 lowers the total amount of knowledge imparted -- putting the smartest college professors in with the dumbest college students and the dumbest college students in with the smartest students is a recipe for destroying America's higher education system.

Posted by Steve Sailer | June 10, 2008 3:50 PM

SAT scores predict later GRE scores, but college attended has no effect.

So, as far as we can tell, putting the smartest college professors with the smartest college students has no effect. So you're wrong, Steve.

Posted by gcochran | June 10, 2008 4:06 PM

So, America's colleges, with their steeply elitist pyramid of resources and prestige (much steeper than at the K-12 level in America), are the envy of the world. Compare America's higher education system to Germany, which used to have the world's leading colleges a century ago, but has destroyed them through egalitarian policies.

Matt argues that for K-12, however, America should have an inverted pyramid where the greatest investment should go toward students with the least capacity to learn, those who will give us the least return on investment.

Posted by Steve Sailer | June 10, 2008 4:12 PM

Stephen Jay Gould, America's foremost liberal critic of cognitive testing, spent his entire career teaching at Cal State Dominguez Hills, where he proved that he could turn the children of Compton into high-achieving scientists.

Oh, wait, sorry, my mistake. I was assuming that Gould's private career choices were in anyway related to his public opinions. It turns out that Gould was a Harvard professor! Well, what to you know ...

Posted by Steve Sailer | June 10, 2008 4:49 PM

Ah Steve, Steve, Steve. I knew you couldn't resist.

Posted by cw | June 10, 2008 7:01 PM

The Matt Yglesias Plan for Public Education is to lower the return on investment by investing more in those who are, on average, least capable of learning.

Sounds like a great plan! I'm surprised, though, that the Chinese haven't beaten us to it. Matt should go over to Singapore and explain his plan to Lee Kwan Yew. He'll probably hire Matt right on the spot to revolutionize the education system in Singapore.

Posted by Steve Sailer | June 10, 2008 9:16 PM

The Matt Yglesias Plan for Public Education is to lower the return on investment by investing more in those who are, on average, least capable of learning.

Actually, MattY is saying we should invest lots of resources in areas where those most capable of learning would otherwise not have the opportunity to take advantage of their abilities because of other obstacles in their way.

Posted by Tyro | June 10, 2008 9:32 PM

I teach at another charter school in Boston and have the following comments:

1) Charters are not allowed to "cherry-pick" the "brightest" students. All charter schools in MA accept students from an open lottery that can only be restricted by geography (for Excel that means preference goes to kids in Boston and Chelsea). To "apply" parents need to fill out an application that includes name, age, sex, address, etc. The only schools that cherry-pick in Boston are the exam schools (BLS, BLA, O'Bryant). We lose a large chunk of our 6th grade class to those schools every year, despite the fact that our seventh and eighth graders tend to score higher on many tests than kids at BLA and O'Bryant.

2) Parents of charter school students may be more motivated than the average parent, but what does that mean in practice? My sense is that most charter school parents choose charters because they are safer, have a uniform, have more discipline, have a longer school day and seem less chaotic that public schools. I think going the charter route is often a vote of no-confidence in the district re: safety. I think hard-core academics are less of a concern, though many parents are attracted by the idea, too. Most of my students' (7th graders) parents claim they can't help their students with the work we assign - it's not like most of them were super successful as students themselves. Also, many work multiple jobs and aren't around for their kids out of necessity.

3) A key thing to notice about many of the successful charters in Boston (no, not all of them work) like Excel, Roxbury Prep, Boston Collegiate, Boston Prep, Edward Brooke, and Pacific Rim, is that their scores tend to start low (proving that they are not getting some super advanced set of kids in their lotteries) and rise as students are in the school longer. Some people argue that the cause is the fact that some students don't make it into the higher grades because of low skills or behavior problems, are retained, and then leave the school. Though I don't have numbers to support this, at the two charters I have worked at significantly more students leave for exam schools in 6th grade every year than ever get retained.

Posted by mathteacher | June 11, 2008 7:54 AM