Arnold Kling  

Adverse Results for School Vouchers

PRINT
Doubts about Planning... Kotlikoff on Social Security...

Chang-Tai Hsieh and Miguel Urguiola find evidence that in Chile school vouchers caused schools to compete for the best students rather than compete to deliver better education.


Although statistically insignificant, the point estimates suggest that, if anything, test scores experienced a relative decline in communities where the private sector made greater inroads.

The results of this study would appear to me to suggest that vouchers need to be "progressive," in that higher vouchers should be given to needier students. Progressive vouchers would change the incentives so that schools would compete for needier students, which in turn might create more incentive to compete on quality.

It could be that the majority of early experiments with vouchers are likely to prove disappointing. However, if we learn from successful and unsuccessful experiments and adopt best practices, my guess is that ultimately education systems that adopt vouchers will be superior.

UPDATE: More from Tyler Cowen and from Brad DeLong.

For Discussion. In the Chilean system, poorly-performing public schools were not shut down. How could this explain the authors' results that areas where private schools made larger inroads were not areas in which overall education outcomes improved?


Comments and Sharing





COMMENTS (24 to date)
Eric Krieg writes:

>>The results of this study would appear to me to suggest that vouchers need to be "progressive," in that higher vouchers should be given to needier students.

Agreed. I've been thinking about this lately, with the discussion about catastrophic health insurance, and the middle class entitlement that HMOs really are.

If the fact that HMOs are paid for with pre-tax income is a middle class entitlement, so is public education. What makes education so special that the government should provide it for free to people who can clearly pay for it themselves?

As we have discussed, all this does is compell people to spend more for housing in good school districts. Wouldn't it be more efficient to pay for schools directly?

Mcwop writes:

Ok, on average students do not do better. What about the students that can attend better school than they otherwise might have access to through the voucher? The report does not seem to address changes in these students as a sub-group. Maybe these students see a safer school, better facilities, and more activities. Maybe they do not do significantly better, but rather stay away from the negatives that can inhibit one's growth as a student, and have the guidance to continue their education versus dropping out.

Eric Krieg writes:

From one of Arnold's links:

>>But sometimes I wonder how much schooling, in the formal sense, matters at all. The United States has mediocre schooling, by international standards, but still produces highly productive individuals. Maybe a school is really just a collection of kids, in which case you can only get so far by reshuffling the mix.

All the more reason to go to cut the link between public schools and their local funding. It's a feedback loop. People buy into "good" neighborhoods with "good" schools because they want a good education for their kids, which in and of itself is an indicator for school success. People are self-segregating in ways that make the schools good (or bad).

The goodness of the schools is a function of who goes there, which is a function of who is willing to pay the big bucks to buy the expensive homes in the right neighborhoods, and who is not.

Why not just cut to the chase? Cut the link between schools and localities. charge tuition. Give vouchers on a sliding scale, so that school is affordable for all.

Isn't that how higher education is organized?

Peter Gallagher writes:

Hi Arnold,

I think that the purpose of a school voucher system is to make the schools more responsive to the needs of the voucher beneficiaries (the students, their parents etc). I've long considered it one of those ideas that I'd like to see work better. But there's a disconnect in the next step in the reasoning about vouchers which troubles me.

It's a disconnect that I'm sure you've observed in other voucher systems, too: such as the 'food stamps' program that the US government makes available to needy people.

The assumption is that people who receive the vouchers need (want) what the people who give them the vouchers suppose that they need (want). Frequently, it turns out, that people want something else: gratification and entertainment rather than nutrition or education.

Who can blame them?

I doubt that a tweak such as a progressive scale of vouchers changes this problem much. What beneficiaries want may NOT be a better education: and even if they do, they may not be very discriminating in deciding what 'better' looks like when it comes to education.

A choice of education is something that most people do very infrequently in their life and often with the benefit of advice that is no better informed about outcomes or objective results than they are themselves. Giving them market power as consumers is good: but it doesn't FORM their choices or make their choices as consumers any more discriminating.

And the schools? Well, vouchers will probably lead them to respond more accurately to the revealed preferences of the students for e.g. entertainment or a certificate or the right sort of companions. In fact, you would predict (as I would) that the schools will soon be MUCH better at understanding the needs (wants) of their market than the designers of the voucher program.

I've only to look at business school behavior (one of the first market-funded education sectors in Australia) to see that school administrators quickly learn that the market is not particularly discriminating. They put their efforts more and more into marketing and less and less into the quality of production, because it's a more profitable opportunity.

Frankly, I can sympathize with the Chilean schools.

Best wishes,

Peter

Boonton writes:

"All the more reason to go to cut the link between public schools and their local funding. It's a feedback loop. People buy into "good" neighborhoods with "good" schools because they want a good education for their kids, which in and of itself is an indicator for school success. People are self-segregating in ways that make the schools good (or bad)."

I assume you don't mean to socialize education. Parents who wanted to supplement the voucher with their own money would still be able to do so. That means you will have good districts & bad ones...just like you have places where you can get good Italian food and places you can't.

Is the results of the experiment that schools are competiting for rich students or quality students? If it is quality then the only way to counter it is to make the voucher scalable to how good the kid is. In other words, if your kid is dumb your voucher is bigger. Guess what parents will have their kids do when they start testing?

Eric Krieg writes:

>>Is the results of the experiment that schools are competiting for rich students or quality students?

The quality student. Not necessarily the rich student, although there is a lot of overlap between those two sets.

Vouchers are probably a more fair system than today. The voucher would be set up to allow quality students without the financial means to buy into high end housing to go to "good" schools.

Of course, maybe we need to think about all the "Bell Curve" impliations of vouchers. We already have a pretty extreme meritocricy in this country, which is the cause of inequality. A voucher sysytem is going to make inequality worse.

That's not a problem for me, but it may be for quite a lot of people, especially Democrats. Inequality if their bread and butter issue.

Boonton writes:

Unless you are talking about vouchers in the range of $10-$20K, how are they going to allow the smart but poor student to 'buy into' housing in areas with great public schools?

Eric Krieg writes:

>>Unless you are talking about vouchers in the range of $10-$20K, how are they going to allow the smart but poor student to 'buy into' housing in areas with great public schools?

Even if awful schools like DC, the amount spent per student is pretty high, on the order of $10k. If we are really going to get rid of locally funded public schools, and put all that money into vouchers, then the voucher is going to be pretty high. Like $10,000.

Jim Glass writes:

As a Chilean has posted on DeLong's blog, the situation in Chile is so different from that in the US that it's dubious it has any relevance for the US at all.

But regarding the US, the new day brings another happy example of the priorities of our urban public school educators...

~~~~
Detroit Schools Pushed Away $200M Gift

October 12, 2003 -- Thanks to the poisonous atmosphere created by a hostile Detroit public school establishment, philanthropist Robert Thompson has decided, with deep regret, that it is impossible for him to donate a $200 million gift to the city's schoolchildren.

The gift would have come in the form of 15 new charter high schools ....

After seeking legislative authorization for his schools for almost a year, Thompson threw in the towel after the Detroit teachers union threw what can only be described as a tantrum at the prospect of having to compete with charter schools ...

In response to this pressure from the public school establishment, both the governor and Detroit Mayor Kwame Kilpatrick walked away from the Thompson gift ....

Would the Thompson academies have helped? ... Mackinac Center research shows that, despite efforts of the public school establishment to undercut them, performance of students in Michigan's charter schools on the state's MEAP achievement test is improving at a rate dramatically faster than in traditional schools.

By their actions, the defenders of this failed system could not have made their scale of priorities more clear. Low on that scale are the prospects and well-being of the children ...

http://theoaklandpress.com/site/news.cfm?newsid=10307607&BRD=982&PAG=461&dept_id=511388&

Thomas Blankenhorn writes:

Did anybody notice that the paper exclusively focuses on performance changes (or the lack thereof), but says nothing about the _costs_ of schooling? As I understand it, economists who come out in favor of vouchers say the found evidence of increased schooling _productivity_, measured in performance per unit cost. Hsieh and Urguiola provide reason to believe that vouchers didn't improve schooling in Chile. But it is still possible that they did reduce schooling costs enough to make the excercise worthwile.

Does anyone know of any evidence about that? Thanks!

Patrick R. Sullivan writes:

This post from DeLong's comments section indicates that there is little, if anything, to be learned from Chile's system, for the U.S.:

___________quote--------
My wife is Chilean and attended private school in Chile so I asked her what she thought.

What I found is that while the Chilean experience is interesting, it's not very comparable to the United States.

First a couple of things about Chile. Chile is one of the most class-stratified societies in the world, and the school system is one of the primary ways that class distinctions are maintained. The way that you make sure that your kids marry into the "right" families is to send them to private schools where they will only associate with kids from the "right" families. The elite private schools (which are NOT part of the voucher program) are very exclusive, and there is intense competition from the upper-middle class and those who are not the very elite to get their children into the best private schools. In some ways it sounds a lot like Manhattan in that the best pre-schools have long waiting lists and there is cutthroat competition to get your kids into the best preschools so that they can mix with the "right" kids and move on into the best elementary schools and so on. "Marrying Up" is one of the main ways to advance in Chilean society. So for parents with some money, getting your kid into the right private school is really more about giving your kid the chance to associate with the "right" families than getting a good education. This is ESPECIALLY true for girls. The elite private schools would never in a million years take vouchers or voucher students because it would defeat their whole purpose of exclusivity to bring in middle and lower class students.

Second, Chile is perhaps the most conservative Catholic country in Latin America. This is a country where divorce is still illegal. There isn't the same concept of separation of Church and State and most of the country is nominally Catholic. Many of the subsidized private schools are Catholic, and the voucher program is an easy way for Catholic politicians to funnel government money to the Catholic church.

Finally, with respect to the paper itself. I asked my wife about the subsidized private schools. She laughed and said they are mostly horrible. That there are probably not more than three good subsidized schools in the entire country and those would be good regardless of whether they had vouchers or not. She said the subsidized schools (as they are called in Chile) are known for having huge class sizes and little resources. They rarely turn students away because each additional student brings in more money. The article suggests that they are more selective. My wife says that what really happens is they take anyone they can get but then flush out the bad apples later. The kids with dicipline problems and the poor learners are simply not asked to come back the next year.

Remember, Chile is very much about class. What the subsidized schools do for lower-middle class parents is give them the ability to *SAY* that they are sending their kids to private school, which is a status thing. If you admit to sending your kid to public school it is the same as admitting that you are low class. Many of the subsidized schools pick names that are very close to elite private schools and they have uniforms that look like those of the best private schools. It would be as if a bunch of new fly-by-night private schools popped up in the California with names like West Harvard Prep, West Princeton Prep, Berkeley Prep, etc. And, for the Catholic schools, the allow parents to give their kids a Catholic education. The subsidized schools give parents the outward appearance of upward mobility without really providing it.

In any event, my wife says that it's not surprising that the vouchers haven't really improved educational outcome. Because she says that frankly almost all of the public and subsidized private schools are considered pretty bad. Bad in the sense that they all are plagued by excessively high class sizes and limited resources. Many of the teachers are dedicated and talented individuals who just make the best of a bad situation. Chile just doesn't have the resources to provide all children with a top quality education. Although it doesn't do a bad job compared to the rest of Latin America.

As for what kind of lessons the Chilean system has for the US. There probably aren't any in the sense that none of the voucher programs proposed in the US would remotely resemble the Chilean system where vouchers can only be used at subsidized private schools that cannot charge tuition.

Posted by Kent Lind at October 13, 2003 10:55 AM
--------endquote___________

Steve writes:

Well don't forget that there is also the parental factor. A child whose parents do not care about school performance, let him watch television vs. doing homework, and don't stress the importance of education is likely to do badly no matter the quality of the school.

Boonton writes:

While Chilie is certainy different than the US I think the above description demonstrates one of my concerns about vouchers: Quality education is not the only motive of parents, just like it is not the only motive of teachers, principals, school boards etc.

While Eric is very eager to trash the local system, it needs to be recognized that such a policy would be a radical change. In fact, taken to their logical conclusion Eric's policy would be a radical expansion of Federal power over both the states and localities. To really make it work the Federal Gov't would have to ban local communities from funding their own schools with property taxes (or supplementing the all powerful Federal Voucher with their own local vouchers).

Before we accept such a change it needs to be shown that:

1. The current system is really bad, not just examples of bad schools or districts but that the entire system is fundamentally unworkable.

2. It should be proven beyond reasonable doubt that this new system will not create problems that are equal to or worse than the current system. A new type of car or drug is not introduced to the market with anything less than extensive testing.

Ray Gardner writes:

The bottom line is accountability. A system of choice is needed to institute real world accountability.

There are too many differences between the U.S. and Chile to make a valid, across the board comparison, especially lacking more real numbers and details.

What is universal however is that spending the entire GDP on education still would not make some students care enough to try nor their parents care enough to make them try.

But, instead of looking towards an elusive and unrealistic “outcome approach,” the right thing to do is give parents who do care a choice. The fact that not everyone will correctly or efficiently use that choice shouldn’t be a reason to take that choice away.

The closest analogy that comes to mind would be if we lost our right to vote because so few people actually vote on a regular basis.

Ray Gardner writes:

I meant to say in the previous post:

"An elusive and unrealistic 'equal outcome' approach."

Essentially meaning that simply because not all people are going to make the right decisions, we don't take away from everyone the ability to make those decisions.

Boonton writes:

"But, instead of looking towards an elusive and unrealistic “outcome approach,” the right thing to do is give parents who do care a choice. The fact that not everyone will correctly or efficiently use that choice shouldn’t be a reason to take that choice away.
The closest analogy that comes to mind would be if we lost our right to vote because so few people actually vote on a regular basis. "

No one is denied a choice. I'm unaware of any state or locality where it is illegal to send your kid to a private school. Are you defining 'denied a choice' as meaning 'not given enough money by the taxpayers to afford a certain choice?' If so then you should acknowledge how radical the idea that you are presenting really is.

Ray Gardner writes:

Boon

Because of the monopoly that government schools enjoy, those with limited income do suffer from a lack of choices.

As it is, Senator Landrieu from Louisiana can afford to send her children to Georgetown Day for around 20K per annum. Those lacking these kind of funds however do not have the ability to “choose” a better school because of the existing monopoly and union controlled public education process.

Please note that this is not saying everyone has the right to choose a Lexus over taking the bus.

Our tax money is being spent in increasingly large amounts on an increasingly failing public school system. Given the choice, many parents would choose to spend this tax money of ours on a better education for their children.

One could argue that education is not a public good I suppose but that doesn’t carry too much water anymore. Accepting that education is a public good, it is denying a person “a choice” to lock up tax payer money in a monopolistic system accountable to no one.

Eric Krieg writes:

>>One could argue that education is not a public good I suppose but that doesn’t carry too much water anymore.

Education is not as much a public good as people make it out to be. For one thing, the returns to education are so high that the individual benefits far in excess of the society that pays for that education.

You know, one reason that people can't pay for their own kids education is that they're being taxed so much to provide education for other people. And while their kids are being educated for 12 years, or whatever, they pay a lifetime of education taxes.

So is anyone really coming out ahead in the education racket? The rich could afford to pay for their kids education themselves. And the poor, who cannot, perhaps in part because of taxes, are forced into inferior public schools.

The only people making out by this system are the teachers. No wonder they are so rabidly against vouchers.

Eric Krieg writes:

>>While Eric is very eager to trash the local system

Damn, B, what can I say? You bring out the utopian in me. I'm ready to liquidate the HMOs, the public school system, and the greedy geezer lobby.

That's not very conservative!

Jim Glass writes:

"Did anybody notice that the paper exclusively focuses on performance changes (or the lack thereof), but says nothing about the _costs_ of schooling? As I understand it, economists who come out in favor of vouchers say the found evidence of increased schooling _productivity_, measured in performance per unit cost."

Yes, indeed. This is the elephant in the living room of the voucher debate that nobody ever talks about.

The argument always is about whether voucher schools spending a fraction of the money of public schools provide better educational results or not. Nobody argues that the private schools perform worse. But while the argument about results is endless, the words "fraction of the money" somehow drop from sight.

E.g. Here in NYC for 20 years there has been study after study showing Catholic schools get better results than public schools with the same students, and rebuttal after rebuttal by the public school defenders saying 'taint so. Of course, *nobody* contends the Catholic schools get worse results.

Yet somehow nobody ever mentions, average public school per student expenditure $12,000 (up 50% in the last eight years!), average Catholic school per-student expenditure $3,500.

To quote the then Budget Director of the NYC Board of Education:

"The paradox of urban school reform is the steady increase in education cost per pupil with no increase in student outcomes. ... Over the past 20 years, factoring for inflation, (NYC) per-student spending has risen 80% while graduation rates, SAT performances, and Regents results have declined and indicators of student misbehavior,
violence and drop-out rates worsened."

-- "Resource Allocation And Productivity: A Financial Analysis Of The New York City High Schools" By Robert Sarrel, Ed. D., UMI Dissertation Services.

One would think that in *economics forums* at least this would be noticed.

Jim Glass writes:

"Well don't forget that there is also the parental factor."

The single most important factor by far, and well-proven to be.

But that is no excuse whatsoever for urban schools that combine dreadful conditions, dreadful results and dreadful expense
~~~

"No one is denied a choice. I'm unaware of any state or locality where it is illegal to send your kid to a private school."

Of course, the poor are perfectly free to choose to spend all the money they want -- on this and countless other things too!

However, being that the urban public schools spend *multiples* of what private schools have shown to be necessary to give a comparable-to- much better education (for the truly disastrous Washington DC schools the last figure I saw was $15,000 per student -- although that system is so grossly mismanaged it admits it doesn't even know how much it is spending!) it might be nice to give taxpayers more value for their tax dollars.

If this accidentally helped the poor as a side effect perhaps we could live with ourselves, in spite of our lack of interest is assisting them.
~~~

"Are you defining 'denied a choice' as meaning 'not given enough money by the taxpayers to afford a certain choice?' If so then you should acknowledge how radical the idea that you are presenting really is."

Fully as *radical* as Section 8 housing vouchers, Pell grants, Medicaid, Medicare, food stamps ... (the list goes on and on) all of which make transfers or net transfers to presumably needy recipients while letting the recipients spend the transfers as they choose. And some of theose aren't even means tested -- they go to the rich!

Not so radical really, is it?
~~~

"Please note that this is not saying everyone has the right to choose a Lexus over taking the bus."

No, it's not -- but when people pay enough in taxes to cover the cost of a Lexus only to be provided a Trabant they may be justifiably upset.

Boonton writes:

"Because of the monopoly that government schools enjoy, those with limited income do suffer from a lack of choices."

There is no government monopoly on schools. Are you aware of the definition of monopoly?

"As it is, Senator Landrieu from Louisiana can afford to send her children to Georgetown Day for around 20K per annum. Those lacking these kind of funds however do not have the ability to “choose” a better school because of the existing monopoly and union controlled public education process."

No, those lacking in funds do not have the ability to choose a $20K per year public school because they do not have $20K per year to spend. I am not denied the choice to buy a Honda Accord because the gov't refuses to give me $20K to put towards a down payment.

"Our tax money is being spent in increasingly large amounts on an increasingly failing public school system. Given the choice, many parents would choose to spend this tax money of ours on a better education for their children."

Of course, given the choice I would also choose to spend your tax dollars buying things for me and my family.

"One could argue that education is not a public good I suppose but that doesn’t carry too much water anymore. Accepting that education is a public good, it is denying a person “a choice” to lock up tax payer money in a monopolistic system accountable to no one."

If it is a public good then by definition society (as represented by an elected government) should decide how the money is spent. Public parks are a classic example of a public good...is it a denial of choice that gov't isn't issuing vouchers for people to join private country clubs?

Boonton writes:

"Yet somehow nobody ever mentions, average public school per student expenditure $12,000 (up 50% in the last eight years!), average Catholic school per-student expenditure $3,500."

The missing element here is why. What are the dynamics that make such a drastic difference possible? I notice that many non-religious private schools have high tuitions. The example of the Senator's children was a private school for $20K per year. Most colleges are closer to $12K per year than $3.5K.

This leads me to suspect that the cost differential is due to:

1. A smaller, cherry picked, population. Yes the easiest 5% of the population may cost $3.5K per year. Not so 80% or 90%.

2. The true cost is really higher but is disguised by explicit and implicit subsidies. If the cost of education is being reduced by tax breaks for the property, donations by non-parent Church members, unpaid volunteer work etc. then this should be factored in. Just because we send 20 times more kids to the Catholic schools doesn't mean the number of Catholics willing to donate resources will instantly multiply by 20x.

3. How to account for the fact that most areas seem to be pleased with their schools? Especially considering that most schools are funded with a regressive property tax often voted on at the local level. If the system was really broken why does it seem like 90% of the population thinks 'the schools' are bad but their school is ok? Perhaps the problem is really one centered in *some* large urban areas & not in the idea of public education.

Washington DC, for example, is not only known for bad public schools but bad public everything. From plowing the streets in snow storms to the local police & so on, DC has a bad rep. This, to me, is really evidence that DC's local gov't is broken....not the idea of schools or snow plows etc.

dsquared writes:

Arnold, in your "post of the week" comment, you say:

I agree that families may make choices regarding education that are not the choices that I would approve. However, I would prefer to let them make their own mistakes rather than have mistakes forced on them paternalistically.

Does this mean that you want children to choose their education without interference from their parents? (or perhaps more specifically from their fathers; there is no actual condemnation here of having mistakes forced on one maternalistically).

Comments for this entry have been closed
Return to top