My not-very-informed impression is that, in terms of measurable school performance, the benefits of selecting your students and your parents swamp almost anything else. If I can select a subset of students from intact families with an average IQ around 110 and few or no special needs, my school will shine by comparison with any school that doesn't do get to do this.
Now, you can object to this because maybe it supports charters or private schools and you don't like that for political or social reasons, but I think there's a more fundamental problem with it: this effect swamps the effect of any actual improvements in education. Maybe some of these charter schools are doing stuff that really does help their kids learn better. Maybe some are doing a worse job teaching their kids than most schools do. But the student selection effects swamp both of those effects. If the school gets to select only kids with years of good grades and high test scores, you can give them an education that's 10% worse than what everyone else gets, and that school will still show up as a massive success story.
I have kids who've gone through selective schools, with interesting and innovative programs. Are those programs better than the default public school offerings? Are they better, conditioned on getting smart and motivated students? I suspect they probably are, but how would we even really know?
One argument I've seen for school choice is that it allows schools to do things that are sensible and correct, but politically untenable in public schools, like track kids by ability and by behavior. If school A can't expel kids even when they're constantly beating up other students and disrupting their classes, but school B can, then school B is going to perform a lot better, just because they'll have fewer kids worried about getting beaten up and fewer classes disrupted.
As a parent, I want my kids in the school where beating people up or disrupting class isn't something you can get away with, and where you don't have some poor teacher trying to teach algebra to a group that includes kids who've already learned it off Khan academy and kids who don't remember how to do long division, multiply big numbers together, or multiply fractions. For a variety of reasons, it seems difficult to get the public schools in many places to provide that, which is a good motivation for me to put my kids elsewhere. But there's not actually any reason public schools couldn't do that stuff, except the politics.
Yes, and... The students entering Harvard might do fine wherever they went, but that's not so true for those entering Kindergarten who have the potential to go to Harvard. For some of those, their 13 years of schooling will take Harvard off the table. For others, it will boost their chances. The Commissioner of Education of Tennessee once said, as she was trying to dismantle Tennessee's best-in-the-world Value-Added accountability system, "Kids that are not going to Yale need just as much as those who are." The problem is, who is and is not going to Harvard or Yale or name your school isn't nearly as clear at Kindergarten as it is by the time they are seniors in high school.
Even allowing for selection bias, there are real differences in schools. Some schools and school systems with highly capable incoming students coast on those innate abilities and do little to develop them. Others give up on students with lesser innate capacity. Thirteen years of great schooling can make a huge difference. Even two or three bad teachers in a row can drastically reduce life options for many students. All of this has been clear for almost three decades now based on the work of Bill Sanders and his Value-Added Analysis system.
Then, of course, there are the monumentally bad ideas that gain traction and influence leadership in school systems (and there are some new whoppers floating out there today). When I was on the school board in Nashville, I had a principal of a middle school in my zone tell me he couldn't support my effort to make algebra available to all qualified 8th graders (instead of just those in the magnet schools and the zoned middle school in the rich part of town) because "they serve as role models for the others." Really? In the first place, who signed them up for that role at the expense of their own education. And, second, really? So an 8th grade student seeing his or her peer coast through class without studying, acing the tests and answering every question, is going to say, "Wow! I really need to work harder!" Or, is that student going to say, "I don't have what it takes," and, in the rational response, quit trying? If you want growth mindset responses, the scenario I've just outlined hurts both students. The whiz is learning that it ought to be easy and is set up for a hard fail when things get harder. The less-math-oriented student is giving up too soon and taking away future options and abilities.
My point is this: schooling matters. Selection bias is often used as an excuse for poor performance by teachers, administrators, and systems. In recognizing the one, we need to be careful not to excuse the other.
My not-very-informed impression is that, in terms of measurable school performance, the benefits of selecting your students and your parents swamp almost anything else. If I can select a subset of students from intact families with an average IQ around 110 and few or no special needs, my school will shine by comparison with any school that doesn't do get to do this.
Now, you can object to this because maybe it supports charters or private schools and you don't like that for political or social reasons, but I think there's a more fundamental problem with it: this effect swamps the effect of any actual improvements in education. Maybe some of these charter schools are doing stuff that really does help their kids learn better. Maybe some are doing a worse job teaching their kids than most schools do. But the student selection effects swamp both of those effects. If the school gets to select only kids with years of good grades and high test scores, you can give them an education that's 10% worse than what everyone else gets, and that school will still show up as a massive success story.
I have kids who've gone through selective schools, with interesting and innovative programs. Are those programs better than the default public school offerings? Are they better, conditioned on getting smart and motivated students? I suspect they probably are, but how would we even really know?
One argument I've seen for school choice is that it allows schools to do things that are sensible and correct, but politically untenable in public schools, like track kids by ability and by behavior. If school A can't expel kids even when they're constantly beating up other students and disrupting their classes, but school B can, then school B is going to perform a lot better, just because they'll have fewer kids worried about getting beaten up and fewer classes disrupted.
As a parent, I want my kids in the school where beating people up or disrupting class isn't something you can get away with, and where you don't have some poor teacher trying to teach algebra to a group that includes kids who've already learned it off Khan academy and kids who don't remember how to do long division, multiply big numbers together, or multiply fractions. For a variety of reasons, it seems difficult to get the public schools in many places to provide that, which is a good motivation for me to put my kids elsewhere. But there's not actually any reason public schools couldn't do that stuff, except the politics.
Yes, and... The students entering Harvard might do fine wherever they went, but that's not so true for those entering Kindergarten who have the potential to go to Harvard. For some of those, their 13 years of schooling will take Harvard off the table. For others, it will boost their chances. The Commissioner of Education of Tennessee once said, as she was trying to dismantle Tennessee's best-in-the-world Value-Added accountability system, "Kids that are not going to Yale need just as much as those who are." The problem is, who is and is not going to Harvard or Yale or name your school isn't nearly as clear at Kindergarten as it is by the time they are seniors in high school.
Even allowing for selection bias, there are real differences in schools. Some schools and school systems with highly capable incoming students coast on those innate abilities and do little to develop them. Others give up on students with lesser innate capacity. Thirteen years of great schooling can make a huge difference. Even two or three bad teachers in a row can drastically reduce life options for many students. All of this has been clear for almost three decades now based on the work of Bill Sanders and his Value-Added Analysis system.
Then, of course, there are the monumentally bad ideas that gain traction and influence leadership in school systems (and there are some new whoppers floating out there today). When I was on the school board in Nashville, I had a principal of a middle school in my zone tell me he couldn't support my effort to make algebra available to all qualified 8th graders (instead of just those in the magnet schools and the zoned middle school in the rich part of town) because "they serve as role models for the others." Really? In the first place, who signed them up for that role at the expense of their own education. And, second, really? So an 8th grade student seeing his or her peer coast through class without studying, acing the tests and answering every question, is going to say, "Wow! I really need to work harder!" Or, is that student going to say, "I don't have what it takes," and, in the rational response, quit trying? If you want growth mindset responses, the scenario I've just outlined hurts both students. The whiz is learning that it ought to be easy and is set up for a hard fail when things get harder. The less-math-oriented student is giving up too soon and taking away future options and abilities.
My point is this: schooling matters. Selection bias is often used as an excuse for poor performance by teachers, administrators, and systems. In recognizing the one, we need to be careful not to excuse the other.