« What do on-line daters want? | Main | Why Can't We Have a Better Press Corps? »
Making school choice work
Here is a new policy paper by Caroline Hoxby, written for a New Zealand context. It is an admirable brief summary of the U.S. evidence on school choice, almost like a response to Ezra Klein. Here is a summary. Her conclusions are that successful school choice requires:
- Supply flexibility, which means that schools should have the ability to open where there is demand for them, expand with increased demand and contract with reduced demand
- Money should follow students, which means that funding policies must be designed so that schools that are in demand have the funds to expand and those that are not in demand lose funds and must contract; and
- Independent management of schools, which means that schools must be free to innovate in a range of areas, including pedagogy, teacher pay, budget allocation, and the way the school is organised.
Hoxby stresses that these conditions are rarely found together. By the way, the paper does not cover either Chilean or Colombian experience with vouchers.
Posted by Tyler Cowen on September 22, 2006 at 07:48 AM in Education | Permalink
Comments
From many years watching over my children in DC and Fairfax county public schools as well as private schools, I have formed a rule of thumb on schools. The more the bureaucracy (supervisors, school boards, city councils, state legislators, congress, and presidents) intrudes into the classroom, the worse the teaching. Charter schools would minimize bureaucracy to save money, and they would put pressure on the public schools to let teachers serve the need of the students instead of the bureaucracy. The downside to school competition is it produces the sorting of students by parents and students motivation, and, probably the students ability. In school systems where parents are knowledgeable and self confidant enough to pressure the schools I doubt that competition would have as much of a positive effect, and the sorting effect might produce an over all negative for the school system as a whole. In the Colombian case, where the program was not restriced to the poor, there was also sorting by income it is not surprising that it did not result in overall improvement.
Posted by: joan at Sep 22, 2006 10:21:27 AM
Here is an important paper showing how hard it is to start a private school in California:
http://www.reason.org/ps329.pdf#search=%22seireg%20%22addition%20and%20subtraction%22%22
Posted by: Daniel Klein at Sep 22, 2006 10:48:23 AM
Hoxby has done outstanding empirical research on school systems
and choice for years and I wonder why it doesn't get more attention from
the press or even academia.
Posted by: Dan L at Sep 22, 2006 11:52:22 AM
From Ezra:
>I'm always impressed by the faith libertarians, and some contrarian liberals, put into education markets. They speak of them in the rapturous tones of Bill Kriston contemplating slaughter, or me talking universal health care. But none of the evidence I've seen on charter school outcomes has been very convincing. My understanding is that while they've not cream-skimmed, taking only the rich and white as some Democrats feared, they've failed to improve outcomes among their students. Nor have they been found to improve the performance of surrounding public schools -- a RAND study (pdf) said there was "no measurable impact" and "no evidence that charter schools create a competitive environment." But surely the libertarians have seen these studies too, and I've spent very little time studying education policy. So please, someone, show me what I'm missing, or have to read, or where the compelling evidence lies.
How is that new report in any way a response to the charge that "while they've not cream-skimmed, taking only the rich and white as some Democrats feared, they've failed to improve outcomes among their students"?
Posted by: joeo at Sep 22, 2006 12:03:17 PM
From Ezra:
>I'm always impressed by the faith libertarians, and some contrarian liberals, put into education markets. They speak of them in the rapturous tones of Bill Kriston contemplating slaughter, or me talking universal health care. But none of the evidence I've seen on charter school outcomes has been very convincing. My understanding is that while they've not cream-skimmed, taking only the rich and white as some Democrats feared, they've failed to improve outcomes among their students. Nor have they been found to improve the performance of surrounding public schools -- a RAND study (pdf) said there was "no measurable impact" and "no evidence that charter schools create a competitive environment." But surely the libertarians have seen these studies too, and I've spent very little time studying education policy. So please, someone, show me what I'm missing, or have to read, or where the compelling evidence lies.
How is that new report in any way a response to the charge that "while they've not cream-skimmed, taking only the rich and white as some Democrats feared, they've failed to improve outcomes among their students"?
Posted by: joeo at Sep 22, 2006 12:03:35 PM
It is a response to the charge in that it asks "what does it take to make a voucher program effective?" If some voicher programs are ineffective, but also don't do what it suggests, then this paper claims that it is not that voichers are ineffective, but rather that bad implementations of vouchers are ineffective. That, coupled with a listing of how to differentiate a good from a bad voucher program is quite the response.
Posted by: Kyle at Sep 22, 2006 12:37:49 PM
One issue I've not seen raised by opponents is risk. It seems to me as a non-economist that one part of school choice is the acceptance of more risk. Just by the odds, some (most?)new charter schools will do better than existing schools; some will do worse. A parent making a choice has to weigh the possible gains (child spends a year learning more than she would have in her public school) against the possible losses (child spends an unproductive year in a very bad school).
Posted by: Bill Harshaw at Sep 22, 2006 2:24:10 PM
Milton Friedman has a website about vouchers. To say that the site has a lot of links and articles is an understatement.
This is a summary of John Stossel's Stupid In America piece he did a few months ago on 20/20. I don't know if you can get it on the internet or somewhere, or if they will show yet another repeat of it, but the show itself is more damning that the article in my link, and that is pretty dmaning by itself. One quote: Belgian school principal Kaat Vandensavel told us she works hard to impress parents.
She told us, "If we don't offer them what they want for their child, they won't come to our school." She constantly improves the teaching, saying, "You can't afford 10 teachers out of 160 that don't do their work, because the clients will know, and won't come to you again."
"That's normal in Western Europe," Harvard economist Caroline Hoxby told me. "If schools don't perform well, a parent would never be trapped in that school in the same way you could be trapped in the U.S."
Posted by: happyjuggler0 at Sep 22, 2006 2:27:57 PM
The problem with school choice is a lack of solid consumer information for making informed choices.
The most important mechanism for making school choice work would be mandatory value-added testing: test all students each school year to see how much smarter and better-informed they've become, relative to themselves.
Posted by: Steve Sailer at Sep 22, 2006 4:33:04 PM
Steve, although I see your point, I doubt this: "The most important mechanism for making school choice work would be mandatory value-added testing: test all students each school year to see how much smarter and better-informed they've become, relative to themselves."
I teach in greater Boston, which has an extraordinary amount of school choice for folks with money in the form of a vast number of private schools. Private schools are conspicuous in part for having much less testing than public schools, and less availability of the testing data they do have, and yet there is a flourishing market.
One reason is that parents have alternative methods for finding out about school quality; word-of-mouth is huge. I'm not sure how scalable this is; maybe it works because the subset of the local population which is interested in private schools is relatively small and highly interlinked, so word gets around. But even without the personal connections parents have many other sources of information about schools, such as web sites and tours.
Another reason is that the testing isn't going to capture what parents necessarily find important about schools. It won't capture everything they care about academically because it's not going to reflect everything students do academically; it's difficult to make standardized tests which reflect higher-order thinking, particularly if you want their results to be comprehensible to the general population, and there are plenty of subjects (such as mine, Latin) which are not generally tested. But more than that: academics isn't all that parents care about. Parents remark, when they tour our school, about how warm and caring it seems. Parents remark on how their children find niches, acceptance and ways to develop their passions. You walk through the school and you see lots of things you will never see on tests, like adults having a thirty-second conversation with a kid which shows they really know each other, or this fantastic art project the fifth graders do, or enthusiasm. And these things really do matter in kids' educations and in parents' evaluations of school quality. You can have a half-dozen testing results in front of you and that won't tell you what school is a good fit for your child.
Evaluating the quality of a school is a difficult problem I spend a lot of time mulling over. I have a very good handle on the quality of my school, because I work there. I am very unclear on how to evaluate the quality of a school I have not spent substantial time in. So many of the things that matter are difficult to capture, and certainly impossible to capture on paper. Even though I have strong opinions about the features schools should have, and a much better-than-average ability to evaluate the quality of middle and high school curricula (being a middle school teacher), I am not at all clear how I will approach the school search process in a few years when it becomes relevant for me.
Posted by: Andromeda at Sep 22, 2006 8:34:50 PM
I agree with Andromeda. I like the idea of choice because I don't think that the one size fits all produces the best educational outcomes. When I looked for the "best schools" for my children I did not always use the same schools for them because they had different interest and needs. Test scores as long as they were acceptable were not the deciding factor. The one thing I always looked for was flexiblity which is where public schools are most deficient.
Posted by: joan at Sep 22, 2006 9:07:17 PM
One aspect of Chile's experience with vouchers that has gone unmentioned is that the school system is presently racked by massive protests against the government's education policy. Earlier this year, practically every school in the country was shut down by sit-ins for fully two weeks; the spectre of more strikes is ever-present, of course.
Change seems inevitable, so soon we may have some contemporaneous data about the effects of de-privatizing education...
Posted by: neil at Sep 22, 2006 10:11:20 PM
I am against vouchers. It is the new busing. If any of this stuff passes (and when people find out is the new busing, it won't), watch for a lot more kids to be yanked out of the public schools and home-schooled, and for more private schools which won't participate in the plan.
Posted by: Arnold at Sep 22, 2006 10:49:53 PM
Steve, I'm surprised you didn't point out that part of the reason for bad schools are bad students whose parents don't particularly care how well they do, and that letting the Nelson Muntz's into the classes the Martin Prince's go to escape them will not be good for either.
If you ask me, I say just get rid of public education.
Posted by: TGGP at Sep 23, 2006 11:48:02 AM
Wonderful example of what is wrong with our public schools.
LAKEWOOD, Colo. -- A Jefferson County geography teacher was placed on paid administrative on the second day of school for hanging several flags from other countries in his classroom.......
for insubordination, citing a Colorado law that makes it illegal to display foreign flags permanently in schools.
http://www.thedenverchannel.com/news/9726287/detail.html?subid=22100484
Posted by: joan at Sep 23, 2006 12:43:36 PM
Having spent vast amounts of money over the years on schooling for my kids, what I've found is that you are buying a pig in a poke. And educators, including private school educators like Andromeda above, want to keep it that way.
Posted by: Steve Sailer at Sep 24, 2006 12:21:57 AM
Neil, it was not Chile's school system that was racked by massive protests, it was Chile's public (non-voucher) school system that was racked by massive protests.
Besides the fact that the protest movement was to a large degree manipulated for political motives (it's only been a few months and most of the student leaders from the movement are now making conspicous appearances as the latest figures for the major political parties), the three main objectives of the movement were to get a "free rider card" for the public transportation system (it's currently subsidized for students), free entrance exam into the university system (it's a standard test, so you have to take to get into an university), and an overall improvement in the quality of education received in the public (non-voucher) system. Improving quality was not even the main objective.
In regards to the "voucher" system, it's not really a voucher system, it's a susbsidized private schooling system. Each private school can choose wether to receive the subsidy (it's per-student and it amounts to something like $500 per year if I recall correctly) for all it's students or not at all. That means you don't get to choose any school you want, though most private schools do accept the subsidy.
If you want to know what customers (parents) think of this system, well, when it started in 1982 subsidized schools were attended by less than 20% of all students. Now they are attended by more than 40% of all students (public schooling takes a 50% and private-non-subsidized schools around 10% of all students).
Does that seem like a failure to you?
Posted by: xavier at Sep 24, 2006 12:10:55 PM
I agree with Steve. An all private system would take the problems of applying to college and move the problems to the kindergarten level. At the top there will be the established private schools that will be able to choose a few students out of a huge number of applicants. At the bottom will be the storefront schools for parents who cannot get their children admitted anywhere else.
I like to say, school choice does not let parents choose their children attend, it allows parents to choose the school where they apply.
I also totally disagree with Andromeda in the benefits of picking schools based upon reputation (word of mouth). Reputation is the one thing that a newly opened school does not have and the one thing that is slowest to change about a school that is on the downward slope of performance. Parents pick schools on reputation while the schools hide their problems from the parents.
Posted by: superdestroyer at Sep 24, 2006 1:24:45 PM
xavier:
What you are describing is the sorting effect. Initially 80% are satisfied with their school. When the 20% of the most motivated students leave, some of those that were previously satisfied become disatisfied and they leave etc. This is what the opponents of choice worry about.
Posted by: joan at Sep 24, 2006 1:50:38 PM
Sweden has had school vouchers since the early 1990s, and a recent paper reports positive effects:
"Since the introduction of school vouchers in 1992, independent and public schools in Sweden operate on equal terms. We analyze the effects of competition on the public schools using data on the results of 28,000 ninth graders. Because the decision on which school to attend is a choice variable, sample selection models are used. To account for the potential endogeneity of the share of students attending independent schools, we use instrumental variable estimation. We also estimate panel data models on 288 Swedish municipalities. Our findings support the hypothesis that school results in public schools improve due to competition."
The reference:
F. Mikael Sandstrom and Fredrik Bergstrom (2005). "School Vouchers in Practice: Competition Will Not Hurt You." Journal of Public Economics, 89(2-3): 351-380.
URL to the journal: http://www.elsevier.com/wps/find/journaldescription.cws_home/505578/description#description
URL to the working paper version: http://ideas.repec.org/p/hhs/iuiwop/0578.html
Posted by: Niclas Berggren at Sep 26, 2006 5:45:30 AM
Speaking from that "New Zealand context": she misses a prerequisite.
You also need efficient transport between possible school sites. "School choice" requires it: else you will have too few providers for a given consumer to choose from and "the market" is inefficient.
That, in turn, means that you require population density. Else you lack school choice.
May I remind you (again) that she's writing for a New Zealand audience.
Posted by: meno at Sep 30, 2006 12:30:03 AM
