« Assorted links | Main | Book Forum: Harford and Kevin Grier on Cities »
The Power of Vouchers
Many studies of education vouchers have looked at the achievement of children who are given vouchers and who transfer to private schools. Generally these studies have found small but meaningful improvements (e.g. here and here). A voucher program, however, is about much more than transferring students from lousy public schools to better private schools it's about creating incentives to improve the public schools.
Florida's Opportunity Scholarship Program rated schools. Students at schools that received an F in multiple years became eligible for a voucher that allowed them to attend a private or higher-rated public school. In Feeling the Florida Heat? (ungated version) a paper sponsored by the Urban Institute Rouse et al. look at what happened at failing schools.
...we find that schools that received a grade of “F” in summer 2002 immediately improved the test scores of the next cohort of students, and that these test score improvements were not transitory, but rather remained in the longer term. We also find that “F”-graded schools engaged in systematically different changes in instructional policies and practices as a consequence of school accountability pressure, and that these policy changes may explain a significant share of the test score improvements (in some subject areas) associated with “F”-grade receipt.
Thus, this paper shows two things. First, that the test scores of the students in the public schools improved when vouchers gave the schools better incentives to perform. Second, at least some of the improvement comes from changes in how students are taught. The author's note, for example:
...we find that schools receiving an “F” grade are more likely to focus on low-performing students, lengthen the amount of time devoted to instruction, adopt different ways to organize the day and learning environment of the students and teachers, increase resources available to teachers...
It is not true that "nothing can be done to improve the schools." Incentives matter.
Notice that Florida's program worked even though the program was very weak. It offered vouchers only to students in the worst schools and only after those schools received F grades in multiple years. The vouchers were relatively small and could not be topped up. In addition, the program lasted only a few years before it was declared unconstitutional by Florida's supreme court.
A true voucher program would be national, would not discriminate among students, would offer funding equal to that spent on students in public schools and would be permanent. Competition in such a system would be more intense and even more productive than in Florida's program.
Posted by Alex Tabarrok on February 18, 2008 at 07:46 AM in Economics, Education | Permalink
Comments
Of course scores improved in schools that received an F...It is easy to pick low hanging fruit. Tim Harford had a chapter in his book, the Logic of Life, in which he shows that moving children doesn't increase test scores, but simply improves their utility. In my opinion, the only pedagogy that will increase a student's attendance, achievement, and standardized test scores, is a caring teacher no matter where that teach is employed.
Posted by: Mike Fladlien at Feb 18, 2008 8:22:28 AM
If slavery still existed and someone proposed a voucher system that allowed a slave to move from one plantation to another, would you favor it? If houses were "owned" by the State, would you favor a voucher system that let people move from one house to another? If not because maybe you wanted to abolish slavery and state ownership of housing, why not propose abolishing public schools?
Vouchers are a political football, and can be controlled and restricted by educrats and politicians. They don't address the real problem and don't offer a market-based solution, which is private schools.
It's high time to stop being a utilitarian already. If slaves and home occupants should have their freedom, so too should school children.
Posted by: Bill Stepp at Feb 18, 2008 8:28:27 AM
> A true voucher program would be national, would not discriminate among students, would offer funding equal to that spent on students in public schools and would be permanent.
That's a touching statement of faith. But you should know about the perils of generalizing from highly selective examples to the population as a whole.
Posted by: Stephen Downes at Feb 18, 2008 8:32:45 AM
Would you know some similar studies for vouchers IN THE ARTS? Thanks in advance.
Posted by: Moggio at Feb 18, 2008 8:33:27 AM
The basis of this study is flawed because of the focus on test scores. Standardized testing is not an effective measurement of intelligence (see Flynn, "What is Intelligence?: Beyond the Flynn Effect"). The two key flaws with this reasoning are that 1) tests are not consistently baselined every year, but instead every 3-5 years or more, resulting in a false increase in intelligence scoring; and 2) schools that perform poorly on tests will focus more of their time on teaching to the test rather than on educating students.
The ultimate fallacy in education reform today is that improved test scores equates to improved education. Nothing could be further from the truth. As a result of this focus, students are not learning how to think creatively and extensibly. The learning they're being exposed to is rote memorization and not education, which in the long run does not serve society well, assuming your goal is to continue the shift toward a service-oriented job market where many jobs require problem-solving skills. It's a slippery slope that programs like vouchers and NCLB have put us on; one that leads to a far weakened position for the US.
Posted by: Ben at Feb 18, 2008 8:39:30 AM
If F-rated schools improve, this might just be regression to the mean. To study whether this is the case, F-rated schools that are given vouchers schould be compared to F-rated schools that aren't.
Posted by: LemmusLemmus at Feb 18, 2008 8:46:33 AM
@Ben
Please explain how "vouchers" are the cause of standardized testing? or vice versa? They are not related. The emphasis on testing comes from the desire to be able to "measure" a school;s and individual's performance. It doesn't work.
My time in high school was spent in independent study and mentorship programs which I believe are in jeopardy because of the emphasis on testing. Essentially I was off the grid so to speak. Choice in how a child is educated is important as one method of instruction does not work for all children. This is why vouchers and school choice are more important than standardized testing.
Posted by: tim at Feb 18, 2008 8:53:27 AM
Vouchers have great promise but they may also have a hidden cost.
A benefit of public schools is the (imperfect) mixing of different groups. With vouchers, we may exacerbate tribalism with schools catering to different religious sects, ethnic groups, races, languages, cultures, political ideologies, income levels, athletic interests, fashion levels, etc. Many, perhaps most, parents will not choose schools based on their academic quality (as most academics might think or wish), so much competition will be on different margins.
Tribalism is one of the most destructive forces on the planet. Because a proclivity for tribalism appears to be written into our genetic code, active, ongoing measures are needed to tamp it down.
Posted by: A student of economics at Feb 18, 2008 9:07:21 AM
I tend to be pretty free-market in general, but I've always been skeptical of school vouchers. This type of research goes a long way towards convincing me that they would be effective. I'd be interested to know if other states experimenting with vouchers have seen similar effects.
Comment on a comment - excellent analogy between slavery and education! Let's privatize the State Department and armed forces as well. Markets in everything!
Posted by: Greg at Feb 18, 2008 9:08:23 AM
I would rather not have my tax dollars subsidizing the religious tripe that is taught at the vast majority of private schools in this country. Vouchers are really about subsidizing religious education. Increased tribalism is in fact the most likely outcome, as "A student of economics" has pointed out.
Posted by: Diogenes at Feb 18, 2008 10:03:52 AM
Stephen Downs wrote:
"That's a touching statement of faith. But you should know about the perils of generalizing from highly selective examples to the population as a whole."
No, what is not touching are the smug attitudes of the pro government side in this debate. There is no greater example of mass government failure in the United States than secondary public education and we get the same deny, deny, deny, deny response from its advocates. The way the pro government side bashes vouchers or any other attempt to reform the public schools is ideology at its WORST.
Posted by: john pertz at Feb 18, 2008 10:31:40 AM
Student,
Have you been to a public secondary school in the US? The conditions many of the students endure would almost certainly not be permitted in the workplace, and might in some cases be considered torture. These schools tend to have "tribes" of greater intensity than the outside world. I'm not going to claim definite causation here, but I think there is a very strong chance of it. If you deprive people of freedom of association (whether in jail or school), what else would we expect to happen? Do we think all those students who would otherwise never associate with each other will just play nice and say "thank you Mr. School District creator-guy, you know how to choose my friends better than I do"?
I do not think parents should choose schools based entirely, or maybe even mostly, on educational quality. There are many more important things in life than attending a school which supposedly provides a good education.
I'm always surprised at how quickly economists jump to defend the status quo, when that state was reached decades ago by irrational processes.
Posted by: Grant at Feb 18, 2008 10:34:18 AM
Vouchers sound a lot like Medicare to me.
Under Medicare a patient is free to select their doctor with the government paying for it. Moreover, individuals are free to top-off Medicare coverage by buying supplemental private insurance.
Now doesn't that sound exactly like the voucher system being advocated by libertarians? Each student gets a voucher that the student is free to select their school with the government paying for it. moreover, student would be free to top-off voucher coverage by buying supplemental private instructions.
Medicare has gotten much more complex because the political system does not just pass out money without having to impose extra regulations, etc., to try and prevent waste and crime. Wouldn't exactly the same thing happen to a voucher system the first time some entrepreneur found a way to exploit the system without actually providing a good education?
Posted by: spencer at Feb 18, 2008 11:09:36 AM
First, let's not forget that when you are "failing", almost anything is an "improvement."
Second, almost every voucher system I have seen proposed would not actually cover the true costs of a quality private school, even before figuring expenses like uniforms and transportation. Thus, it would be a cruel joke on the lower and middle class families who are supposed to benefit. Most "average" figures for private school costs include very small schools, religious schools, unacredited schools, and homeschool support organizations.
Third, even if vouchers did work, and the students of failing schools fled to private schools, where exactly are private schools going to get the capacity for all these students? Where will they come up with seats and books and instructors? Anybody? Bueller? Gee, I here that Failing Grade School has laid off 20 licensed teachers!
Posted by: ShortWoman at Feb 18, 2008 11:36:05 AM
Re: schools rated F do better the following year. At least in part, this is the result of regression toward the mean. Is there any attempt to address this fairly obvious and very important point?
Posted by: steve at Feb 18, 2008 11:39:06 AM
No matter how problematic vouchers may be, I think there is enough evidence to argue that a nation wide voucher system would be superior to the current situation. I feel relatively certain that the voucher skeptics lack enough evidence to argue against their implementation. Many of the arguments against are on the shakiest of grounds at best and involve way too many if, buts, maybes, subjective opinions, and just plain old ideology. I like government therefore I like public schools is not a persuasive argument in favor. The pro government side looks like the Iraqi information ministry prior to the U.S invasion of Iraq. "Everything is fine, leave us alone" is just plain crazy talk at this point.
Posted by: john pertz at Feb 18, 2008 11:46:41 AM
Steve - you didn't RTFS, did you? Section 5.A.3 -> "Mean Reversion and Selection" describes how the researchers deal with this precise issue.
Posted by: Colin at Feb 18, 2008 11:54:18 AM
ShortWoman,
Those small, inexpensive private schools which pull down the mean cost of private schooling are also included in student performance studies, which often rate private schools as being both better performing and cheaper. In addition, vouchers would create a market for cheaper schooling that was previously taken up by compulsory schools. As for how the existing private schools would handle the capacity, I'd imagine they'd do it the same way Paris manages to feed itself despite lacking the paternalistic wisdom of those posting on this thread.
Seriously, problems in education are simple compared to problems faced by other industries on a daily basis. In my opinion, the difference in efficacy is due to the difference in decision making processes. Thats not to say I know how to fix education; I don't know any more about doing that than how to engineer injection molds. I'm equally confident that, a) people out there do know how to fix education, b) I am not qualified to vote on the issue at all, c) most people here probably aren't either.
Posted by: Grant at Feb 18, 2008 12:10:42 PM
So would not the act of getting feedback itself cause an incentive to improve. Getting an F might have led to the gain alone.
We might also consider that the actual testing is the incentive to improve, as it creates a core curriculum that can be taught to. The so-called "teaching to the test" nightmare of educators who don't care about standards whatsoever.
Posted by: akatsuki at Feb 18, 2008 12:15:24 PM
Reading some of the comments regarding the success (or lack thereof) of vouchers has only reminded me why I am skeptical of supporting voucher programs. I am someone who considers himself more liberal and yet I have found many of the arguements regarding a voucher program to be enticing; improved test scores and the improvement of underperforming schools to name two cited in this article. I will even give more evidence supporting the implementation of voucher programs; the success of school choice in places like Denmark, the possibility of breaking up class (and race) based segregation in schools and the possibility of encouraging teacher performance through higher pay.
Saying all this, I can't in good faith support vouchers due to my suspicion of motivations animating the pro-voucher movement. Previous posts here and other publications make me question whehter voucher supporters actually believe in universal education. There is a social darwinist aspect of the pro-voucher movement that at times I find somewhat repugnant. While I understand many supporters of vouchers are genuinely interested in improving education for the general populace, I become unerved when I detect a more ruthless sense of "weeding out the weak" that permeates much of the writing on the subject.
Beyond this more emotional reaction against vouchers, the pro-voucher movement often overlooks other reforms that may need to be addressed. A recent Atlantic Monthly article illustrated one of the biggest problems with education; local financing. Our own sense of local educational control is in fact a hinderance to improvement of education for the masses. A previous poster mentioned the danger of tribalism if school choice was implemented. However, a class-based tribalism exists due to the way schools are locally financed. A HUGE factor in driving where people live is to place their children in top public schools. We already have a version of school choice already that divides the have's and the have nots. With this in mind, I wonder how long it would be before people start to complain their children have to share the schools with the "wrong" sorts of people (I witnessed a version of this phenomenon when a group of mostly non-white children started attending my school from other towns and even states). I have more to write on the subject, but I have written long enough.
Posted by: Colin at Feb 18, 2008 12:34:50 PM
Section 5 A 3 seems to me to be quite mixed up with the concepts of "mean reversion" and "regression to the mean".
Posted by: dsquared at Feb 18, 2008 1:01:25 PM
I would just put the burden on the government people. Its their mess. If vouchers are not the solution then what is is? Vouchers, private schools, and public schools are just attempts to try and arrive at the same thing which is a quality secondary education. I feel like people are too bogged down on the means and not focusing on the fact that for the vast majority of Americans, the public schools in this country are failing them.
Posted by: john pertz at Feb 18, 2008 1:54:32 PM
Strange that a group of economists seem pretty hell-bent on forcing consumers to nationalized education in the name of diversity. Especially the comments that further move in the direction of forcing students to attend schools outside their neighborhood to avoid any kind of class separatism.
I grew up on the edge of a bad neighborhood with bad public schools. Because I was white and didn't qualify for the busing programs, my parents decided to take extra jobs to send me to private school to keep me away from the gang and drug elements of the public. I'd be curious to know about the criminal element in some inner city schools (which maybe I'm wrong but could be correlated with lower test scores) and the impact school vouchers would have on juvenille crime (if any).
As for the tribalism, any former public school student knows quite well that it is alive in well in every school, and not necessarily on the basis of class or race. It seems to be impossible to have a large heterogeneous group without self-grouping. And I'm not sure what the benefit would be in forcing students to abstain from pursuing those relationships (unless of course it is crime-related).
Posted by: Jarick at Feb 18, 2008 2:04:00 PM
It's amazing how human animals will think up any excuse possible to resist change. The nay sayers are basically afraid of change. Anything that disturbs their own idea of how the world should work. The religion of public schools won't couch having any competition. The publicans want to keep their monopoly and thereby prevent those "religious fanatics" from getting state funding. This is the mind set of the opposition. Preaching the secular religion is fine with them. Diversity, multiculturalism, they don't see anything wrong with that. How did the human race ever survive without public schools?
It is apparent that unions and the public monopoly mind set won't allow vouchers on the State level. The best way to introduce it is Federally. Abolish all Federal aid to elementary and secondary education (except aid to the handicapped) and replace it with a voucher system of say $5,000 per student. Let the parents cash in the vouchers at the school of their choice. Of course, the public schools won't accept them. It is against their religion. Let the private schools take them and show the publicans what is wrong with their schools. In no time, you will see the public schools change their tune. Now that's revolution...!!
Posted by: jorod at Feb 18, 2008 4:57:48 PM
"A benefit of public schools is the (imperfect) mixing of different groups."
Yeah, as in the benefit of the smart kids like me getting beat on by the older, dumber, meaner kids. That kind of mixing?
"Tribalism is one of the most destructive forces on the planet. Because a proclivity for tribalism appears to be written into our genetic code, active, ongoing measures are needed to tamp it down."
Give me a break.
Posted by: Andrew at Feb 18, 2008 5:14:24 PM
