Anticipating three arguments against public school tuition

Previously on this subject:
Charging tuition at public schools
More on the Conservatism of the Left
Absent from school: Pigou

My desire is to preserve the best of both public and private schooling and create an environment in which most schooling is privately funded and all schooling is privately administered. The idea is that eventually all people will have choices about where to send their children, with public financing (perhaps, eventually, even private financing by nontraditional means) for the least well-off. The result will be better quality and lower costs of schooling.

The first argument will probably come in the form of the Perversity Thesis, that my proposal will achieve the opposite of what is desired. More specifically, it says that schools will become worse and more expensive, that more rules and regulations will be required, that eventually the state will have to ban private schools. It is hard to anticipate all of the means by which this will happen, but we could speculate.

As public schools start collecting tuition, private schools become a better-looking alternative than they are now, the more so as public school tuition approaches that of private schools. When that happens, parents will exit public schools en masse, moving their children into increasingly balkanized schools. The public schools stuck with very poor children will creak under the weight as they will have lost the strongest advocates for better quality even as the wealthier parents begin tax revolutions to stop the flow of funds to public schools. The students will all be turned out to the private schools, some of which will be shoddy, fly-by-night operations. Poor students will be increasingly victimized by these. A few large corporations will begin buying the facilities up and charging massive tuition. In the end, only the wealthiest will be able to afford education, while the poorest will simply drop out and become burdens to society.

The second argument will be from the Futility Thesis, that my proposal will not achieve what I desire. The reason for this is that there are social, legal, and other institutional foundations for the current system. Any attempt to change one thing will be superficial at best. Thus, my proposal will result in moving students to a new system which provides roughly the same outcomes as the current system.

The final argument will be from the Jeopardy Thesis, that this proposal will jeopardize what we already have. In a story similar to the one above, it says that not only will it not get me where I want to go, but it will cut the supports out from under the existing system so that when my proposal fails, we won't even have the existing system to fall back on.

These theses are, of course, from Albert Hirschman's Rhetoric of Reaction. An essay outlining them can be found here (hattip: Crooked Timber, I think, but I can't find the reference now). These were arguments that Hirschman distilled from the arguments of Reactionaries to the great shifts in political, social, and economic order in the last 300 years. They are now the arguments of (mostly) left-wing, "Progressive" Democrats to any and all attempts to reform an institution which they arguably now control, and which most Americans arguably believe is failing us.

I don't believe the speculation above: I offer it because I am trying to anticipate the counterarguments to this proposal. I think it breaks down at several points:

• Public schools will respond to competitive pressures despite the resistance of teachers unions. They will also be well-funded because there will be resistance to decreasing the supporting tax revenues while the number of students left in them will decrease significantly. Thus, the per-student funds available to them will increase significantly.
  
• Some private schools may indeed be shoddy, fly-by-night operations. Their shoddiness will not be much different from the presently existing public schools. However, parents will share information about them the same way they do about grocery stores or any other institution.
• Maintaining high tuition prices for all students is only possible if there is a monopoly in education. It is interesting that people would be concerned about a private monopoly in education, but not about the existing public monopoly which seems far less responsive to its customers than to its employees. In any case, I deem the creation of a private monopoly extremely unlikely: startup costs are very low, so entry is easy, and there are limits to returns to scale. In countries such as India, private education is an extremely competitive industry. The more likely outcome is that there will be a range of educational institutions that provide choices in cost, quality, structure, content, etc., the same way there is for groceries. The poor will be able to find good quality, no frills, low cost education, and there will be both public and private support to help them pay for it.

Charging tuition at public schools

A quick Google search reveals very little on the topic line outside of plans to waive or charge tuition to students from outside the community (outside the school district or foreign students).

I'm referring primarily to primary and secondary education, not college. Still, there's this article in The Nation which notes that the nation's colleges, which Americans attend in higher proportion than most other nations including those in Western Europe (see this and this for example), are increasingly private. The article is slightly misleading, as it mentions that free college began in 1862 and continued until the 1960s, but has been getting costlier since, without once mentioning the fact that attendance is up significantly since both of those periods. Why? I'm guessing greater returns to education, especially in the last 15-30 years.

My point is that charging tuition at public schools isn't as radical as it sounds: it was taken for granted by Jefferson, perhaps the earliest strong advocate of public schooling. In "The Diffusion of Knowledge", Jefferson says,

Sec. XIV. A visiter [sic] from each county ... shall be appointed ... who shall call and preside at future meetings, to employ from time to time a master, and if necessary, an usher, for the said school, to remove them at their will, and to settle the price of tuition to be paid by the scholars [emphasis added].

Sec. XV. A steward shall be employed [to cook and clean, etc.] ; the expense of which, together with the steward's wages, shall be divided equally among the scholars boarding either on the public or private expence [sic]. And the part of those who are on private expence [sic], and also the price of their tuitions due to the master or usher, shall be paid quarterly by the respective scholars.

(This is from Jefferson: Public and Private Papers, but many other interesting education-related Jefferson quotes may be found here) I have stated before a preference for this type of public schooling for which primary schooling is free for all or at least tuition is means-tested, secondary schooling tuition is also means-tested, and advance to further schooling depends on ability (with tuition picked up for those who can't afford it).

Think of it as a co-payment. Or as a revenue-boosting program. In my earlier post, I slightly-tongue-in-cheek referred to it as a Pigouvian tax on a system that fails to provide a public good (analogous to the normal Pigouvian tax on transactions which yield actual public bads).

Tuition at public school is no panacea. People rightly protest about the wanton spending of public money, pointing out that allocating tax money without stipulating outcomes will not automatically fix problems. Others will ridicule such protests because they believe that all public school problems are the result of underspending, but they would be wise to remember that spending is an input, not an output. However, the additional revenue from tuition may address some problems, and it will also give some impetus to people seeking a cause for their voice. They can, in effect, go to the board and demand something for their tuition dollar which they only pay as a result of sending their child to that school, something which they can take away by exiting that school and going to another. They cannot plausibly threaten to take away their tax dollars, which are usually levied against property regardless of whether you have children in a school or not. It is funny that citation of Hirschman is usually done by the left in order to argue against vouchers, but people doing so fail to recognize how Hirschman revised his own argument and fail to recognize that payment through taxation alone does not give them a plausible threat of exit.

I'm certainly not the first person to have this idea, as evidenced by these letters to the editor (see this and this). However, I'm surprised that it hasn't occurred to more people. According to the recent Wall Street Journal ($) article, "Parents Rebel Against School Fund-Raisers", school fundraising has become so necessary that it has spawned both a cottage industry and a rebellion. According to the article, "The association of principals survey shows that 76% of schools expect to hold between one and five fund-raisers during the 2007-08 school year -- and 3% of schools could hold as many as 15." The cottage industry is there to reduce the agony of fund-raising with pre-packaged promotional offers:

Charles Best, a former high-school social-studies teacher from the Bronx, N.Y., created donorschoose.com, a not-for-profit Web site that teachers in more than 5,900 schools have used to raise nearly $12 million in the past few years. The site lets teachers, students and schools alert parents, friends, families and businesses via email and word-of-mouth to the individual projects that teachers need help funding. The philanthropically minded go online, read about the project and its needs, and then contribute. The site currently works with schools in just 10 states. It will roll out nationally in the fall.

Even still, parents are trying to get out of selling crap nobody wants to people who really only buy because we all believe in supporting education, so they are apparently trying to figure out how to write checks directly to the schools, teachers, and coaches.

At Dutch Neck Elementary School in West Windsor, N.J., more than two-thirds of the school's families voted to do away with traditional fund-raisers two years ago and to instead rely on a "Just Write a Check" campaign. The school raises about $15,000 annually with the new campaign, slightly less than a traditional fund-raiser, "but parents like this program so much better because they don't have to solicit donations," says Anita Grueneberg, the school's PTA president.

Why not simply call it tuition and formalize it?

More on the Conservatism of the Left

What are the hallmarks of conservatism?

• Defense of traditional methods and beliefs
• Resistance to change
• Disregard of methods and beliefs used or held by foreigners as un-American

Nothing is more cliched than the defense of public schools by the American Left.

They defend the status quo to the death. No matter that real, per-pupil spending doubles every few years, no matter that test scores are flat or falling, no matter that students are signing up for remedial classes in college or at their employer's in droves. The answer is always the same: there is nothing wrong with it except that it is underfunded.

Suggest any change -- smaller districts, privatization, year-round schooling, changes in the way curriculum is selected, homeschooling, vouchers, or apprenticeships -- and they oppose it.No matter that private schools are all the rage in India, Japan, Eastern Europe, they don't want anything to do with it. Public schooling is an American institution.

For example, see this. It contains several very specious arguments about access to specific public school programs for people who otherwise homeschool. For example,

"One argument that home-schoolers make in favor of access to extracurricular offerings is that they pay taxes that finance the public school enterprise. Therefore, they claim, they are entitled to take advantage of the school's offerings to the extent that they and their children are interested in doing so.

[...]

"This argument may sound persuasive, but it is based on a faulty premise about a taxpayer's entitlements. Paying taxes is not the equivalent of paying tuition for public school. If it were, then people who have no children, or whose children are grown, would not have any obligation or reason to pay. Yet we all pay taxes, regardless of whether we have children and of how many we have."

Note that this whole part of the argument is misdirection. It begins with a statement about whether taxes are tuition, but never actually refutes the claim. What the author is essentially arguing is that public schools are a take-it-or-leave-it package deal: you can either attend and get all of the goodies, or homeschool and get none of them. The underlying truth about school funding is that funds are doled out by many states on the basis of school population, and allowing homeschoolers to participate in some programs might lead to undercounting and therefore underfunding. True enough, but that is not something the author gets around to discussing.

The entire argument, however, rests on a faulty premise itself: that other people should pay taxes to support schools, regardless of whether they have children or not, on the basis that there is a public good generated as an externality. The author makes only passing mention of the private benefits (when pointing out that public schools provide them without charging a fee, as if they were charitable institutions full of volunteer teachers). On average, people's income is related to their educational attainment. The existence of a positive externality is not everywhere and always a reason for moving a transaction totally into the public sphere, nor is it any reason why the institution must be both publicly funded and administered. The author simply assumes away any question of whether the student or his/her parents ought to contribute some percentage of the cost of education, and this is easy to do because the author analyzes only on the basis of the externality, the side-benefit, while hardly acknowledging the main benefit which is private.

This is vulgar Second Best theory at work, yet again, and it proves as bankrupt as in other venues. To review, Second Best consists of a three step process:

1. Identify a market failure (in this case, a positive externality)
2. Recommend a policy solution ("free" public schools)
3. Declare victory (ignore any negative externalities resulting from the policy, ignore any potential market-based alternative solutions)
   

I carry a cell phone, and I can use it to call the police if I see a crime - should the government pay for and operate cellular phone service on that basis alone?

Absent from school: Pigou

Let's just stipulate that the use of fossil fuels creates negative externalities. Whether these are supporting terrorists and despots, pollution, AGW, or simply underwriting a lifestyle that cannot be sustained forever, I am indifferent. In any case, this is the case made for public intervention in that market. There are two main responses: One is to subsidize or mandate alternative energy, the other is to raise taxes on fossil fuels with the desired result of shifting investment from fossil fuel exploration and extraction to alternative energy research and development. The former is thought to be the worst approach for a variety of reasons, including the argument that the government is a poor decision maker, that it leads to rent-seeking by interested parties (like ADM lobbying for ethanol from corn), and that it shifts money from taxpayers to political entrepreneurs with lots of deadweight losses in between (spending on lobbyists), whereas the latter merely raises the costs of fossil fuels, possibly buttresses state coffers (allowing us to pay for public goods like defense), and does not put the government in the place of solution-picker.

That sounds plausible, even if you might pick at some of the details and ultimately argue that the best solution would be to get government out of either punishing or rewarding any industry. Let's just say that that is a standard wonk description of the problem and solutions.

So let's look at an industry that has some parallel: public education. In this case, the schools should be providing public goods (literacy, numeracy, critical thinking skills, and other skills that Jefferson would argue are critical for the proper function of a democracy), but are widely regarded as having underproduced them. This is apparent whenever the left demands more spending or when the right demands more accountability. Very few will argue that it is perfect the way it is, and usually when they do, they always come back with, "but it could be improved" (usually with more money). My analogy turns on the idea that overproviding a public bad is similar to underproviding a public good. I also believe, but do not intend to prove here, that public school problems are institutional in character: that they will always underprovide no matter how much money you spend on them in the same way that pollution is always overproduced no matter how expensive oil is in a free market (those of you anti-free marketeers please note, however, that the efficient amount of pollution is not 0 and that the heavily regulated soviet industries tended to pollute even more).

The two responses analogous to those above for fossil fuels would be to subsidize the alternative or to change the price structure on the failing industry to make the alternative look more attractive. The former exists as a policy choice in the form of vouchers, charter schools, and mandatory testing; they are at least as controversial in education policy as alternative energy subsidies and mandates are in energy policy. The Pigovian tax response does not currently exist: it would consist of charging more at the pump schoolhouse. In other words, charging tuition for public schools.

Take a few breaths and allow the shock to wear off before continuing.

I propose that we start charging a moderate, nominal tuition for attendance at public schools. I think this can be made palatable to all political parties except the rabid left and religious right. To the former, who think that all education should be free, I simply point out that it is not and cannot be truly free; someone must pay for it because buildings and equipment must be built and maintained and teachers should be compensated fairly; to the extent that the wealthy are made to pay for it, they will be interested in controlling it; and they have many ways to exert influence in the large, complex state that y'all prefer. The "free" universities in Europe are not as successful as those in the US (see Figure 9 in this, our higher per capita spending results in a better healthcare education system according to this), and the European systems are under tremendous pressure to start charging tuition because the current system is underperforming. The problem for the religious right will come up below. To everyone else, I think you'll find that I can address your concerns.

First, I have no more problem with making the tuition means-tested than I do with providing relief for the poor on fuel taxes (ah, funny how you forgot to be indignant on *that* point, eh?). In fact, it's easier to bring in your 1040EZ and get a waiver at the schoolhouse than at the pump, where it must be done with something like the EITC. Many public school systems already have a mechanism for charging tuition for out-of-district students (not alway to the student: sometimes to the sending district).

Second, I'm talking about nominal tuition: $100 - $500 per year per child. Not the full value, just a little. And I'm not advocating reducing the public expenditure on public schools. A school with 1000 students (the average high school in the US has about 752) will suddenly find itself with an additional $100-500k per year for facilities and teachers. Neat, huh?

Third, schools could charge whatever they decide locally. Schools in wealthy districts would charge more than schools in other districts. This would decrease the pressure on legislatures to equalize spending (more tends to get spent in wealthier neighborhoods now), and poorer schools would therefore find it easier to get more public spending sent their way. And to the extent that the wealthier schools would charge and spend more, and therefore be better, poor enrollees would be better off at those schools (remember, I favor making the tuition means-tested, so they can free ride, same as now).

Fourth, I believe that we should be more circumspect about the subsidies we provide for breeding. If you are concerned about fossil fuels and the environment, you should be at least as concerned about the pressure of population on the environment. And for those of you on the right: how far are you willing to take this "personal responsibility" idea? Isn't it a little unfair that people with no children have to subsidize those of you who do? Why should married couples enjoy both income tax breaks and subsidies for raising their children? You don't want to pay for the pill, condoms, or abortions for other people? Fine, but why should we pay for your refusal to use the pill, condoms, or abortions? Let's face it: the two-income family depends on the public school-as-babysitter as much as an institution of learning, and that in turn subsidizes the culture of conspicuous consumption. Incidentally, contrary to popular myth, people on public assistance tend to have fewer than the average number of children, while people who vote Republican are known to have more children than those who don't. "Free" public schools are an example of a public institution that disproportionately supports those who complain most about the taxation that pays for it.

Fifth, to the extent that schools charge some tuition, they make it possible for alternatives to rise up in a similar way a carbon or fossil fuel tax directs makes research into alternatives possible. If a school charges $500 per student per year, and a competing private school or tutor charges $1000 per year, the difference is more palatable than between the "free" public school and the private school. You get similar results to voucher and charter school plans, but

• without the vouchers that "divert" resources away from public schools (they don't in per pupil terms, if you do the math),
  
• without setting up situations where the state may be sending money to religious schools and charlatans,
  
• without therefore putting the state in the position of certifying schools to receive vouchers,
  
• without having to repeal the Blaine Amendments,
• and without giving the state an opening to start killing the independence of private schools with onerous regulations.

Questions?

UPDATE: Apparently Tyler Cowen has reached the same conclusion:

I would be happier with vouchers if we were starting from scratch in designing educational institutions. And while I agree with Jane that children have a positive right to an education, I think the out-and-out laissez-faire option doesn't get enough attention. Keep the public schools we have, but make them charge tuition.

Hirschman and Olson

In this article, I briefly mentioned a problem that Hirschman had overlooked with respect to the use of voice in improving school quality. Upon starting to read Olson, two more occur to me.

The first has little to do with Olson. It has to do with Hirschman's oversight. In Exit, Voice and Loyalty, he makes the case that as quality declines, consumers differ in their responses to that decline. The work overall is brilliant and addresses a set of issues that economists have inadequately treated: the mechanisms by which consumers respond to declines and by which those responses affect the declining organization. But it seems to me that Hirschman similarly has overlooked a key point: how does decline occur in the first place? Is it due to a set of decisions carefully crafted and acted upon, or to outside factors, or to gradual shifts in the organization's attention?

The implications are serious. If, for example, the organization has decided to take a new direction, those decisions could be reversed when the customers' exercise of exit or voice has the intended effect; the more explicit the link, the easier to reverse course. On the other hand, if the decline is due to outside factors, then management may not know how to respond: there is no decision to reverse, and the outside factors could be causing the problem in a non-obvious way. For example, it could be that society itself is creating more difficult conditions in which to conduct education, with less emphasis on study and discipline and more emphasis on distractions. Or the quality decline could be a perceived decline in comparison to other, similar products, in which case the organization needs to research the underlying comparison and formulate a responding change.

Finally, the decline in quality may be due to a shift in the organization's attention (their mission) toward, for example, greater employee morale rather than toward greater productive outcome. If that is the case, it could be difficult to reverse because the shift was probably determined by other real problems such as high turnover, increasing costs of retention, and increasing costs of recruitment. Reversing course here is in fact a requirement for more resources so that, in the common example, both more guns and more butter be produced.

Quality changes - real or perceived - that result from poor decisions, new environments, and new organizational motivations all require a different set of management responses to the resulting expressions of dissatisfaction. I suggest that exit and voice may interact with each of them with differential success. For example, voice would seem to interact very simply with poor decisions, possibly even preemptively. On the other hand, voice may inform decision-making in the new environment problem, but exit would inform experimental entrepreneurs searching for a better answer. Neither exit nor voice would seem to help in the third case (mission shift), since the organization might discount the voice as uninformed about the true nature of the problem, while exit would only serve to starve the organization of the "required" new resources and thwart the proof of what they have already decided is the correct answer. Please note that this list of possible causes and responses is not exhaustive or even representative or realistic. These were merely the first three I thought of; a true set of causes for decline might be much larger, but perhaps only a few dominate the history of failure.

The second problem has much to do with Olson. In the introductory material to The Logic of Collective Action, Olson makes the case that small groups are more likely to achieve the desired outcomes, and that groups will tend to sustain costs until the outcome desired by the participant with the largest share is achieved. Hence, group members will tend to exploit the largest member. This seems to support Hirschman, since it supports his thesis that the rest of the community will rely on the parents with the greatest interest in quality to lobby for the restoration of quality. However, the problem is that once that person gets what they want, the group will tend to fall apart because that person will then stop contributing. We might find, for example, that one parent insists on getting AP classes in the high school, and stops lobbying when they get them. That works great for that person, but what does it do for parents with special needs children? The school, with limited resources, cannot choose both without an increase in resources, but the most effective lobbyist has now exited just as effectively as if they had gone to another school. If they had *actually* gone to another school, the additional resources would now be available for the parents of the special needs student - oops!

Voice in schools

I'm about halfway through Albert O. Hirschman's Exit, Voice, and Loyalty and noticed something that I've seen to be true whenever his work is invoked against voucher programs -- or any privatization, for that matter. Since there is still another half left to go, I'm holding out hope that he will address it, but I didn't find anything promising in the index, so hope is running out for him ... and his little dog, too.

The main emphasis of the 1970 book, so far, is to (1) describe two mechanisms which consumers have recourse to when the quality of a firm or industry deteriorates, and (2) emphasize how little attention has been paid to these, by political scientists and especially by economists who had begun to colonize the other social sciences. The two mechanisms are exit (as in leaving one business and going to another) and voice (as in expressing your displeasure). Hirschman correctly notes that exit is not so automatic as economists have tended to assume, nor is voice so silly an alternative.

Hirschman levels the charge of bias against economists with a quote of Milton Friedman's original voucher proposal from 1955. When he returns to education later in the book, I got the impression that he was intentionally trying to create a tool for defending public schooling. In any case, having created the tool, I think he forgot to investigate whether it worked as well as he thought for that particular application.

The anti-voucher argument can be summed up in this statement.

... those customers who care most about the quality of the product and who, therefore, are those who would be most active, reliable, and creative agents of voice, are for that very reason also those who are apparently likely to exit first in case of deterioration.

But in applying this, both Hirschman and those who draw on him to argue against school privatization take this as revealed truth. Their argument relies on several assumptions that are mistaken in my estimate:

Poor parents are less effective at voice. First, poor does not mean stupid or incapable, and this is frequently the implication of this claim! Poor parents are sometimes the most vocal because they understand first-hand the value of an education. Second, it is true that voice is more expensive to poor parents because of the time required to research, meet, debate, and follow up on ideas, time which they don't have which wealthier parents do (professionals can take time off, one spouse may not have a full time job, etc.). Parents who stay in the public schools will thus have more resort to voice the wealthier they are. Forcing them all to stay ensures that the school will be more responsive to wealthy parents than to poorer parents. If the wealthier parents had the option of exiting, the poorer parents remaining would have greater effective voice as their share of representation goes up. Which leads to ...

The only parents staying in the public system are poor parents. This assumes that voice is less expensive than exit. But Hirschman himself points out the opposite only a few pages before:

In fact, in addition to this opportunity cost [that of foregoing exit], account must be taken of the direct cost of voice which is incurred as buyers of a product or members of an organization spend time and money in the attempt to achieve changes in the policies and practices of the firm from which they buy or the organization to which they belong. Not nearly so high a cost is likely to be attached to the exercise of the exit option in the case of products bought in the market - although some allowance should be made for the possible loss of loyalty discounts and for the cost of obtaining information about substitutes to which one intends to switch.

Hence, in comparison to the exit option, voice is costly and conditioned on the influence and bargaining power customers and members can bring to bear within the firm from which they buy or the organization to which they belong.

In fact, it is in the status quo, where "free" public schools appear to be less expensive than non-free private ones and exit has a cost on that account, that poor families have no recourse except voice and that wealthier ones have both more recourse to voice and exit. In a system in which "free" public schools are no different than voucher-subsidized private schools, exit would be a lower cost means of responding to lower quality than is voice, if Hirschman's claims immediately above are correct. Either he is right in saying that the wealthy will leave them behind, in which case their comparative voice increases, or he is right in saying that exit is the lower cost option, in which case vouchers give an additional option to the less wealthy.

Voice will be effective at effecting change in public schools. But the people making the voice/exit argument against vouchers are usually the same group of people insisting that there is no problem with public schools - the quadrupling of real, per student expenditures is worthwhile, the flat standardized test scores mean nothing, criticisms of public schools are groundless, all they need is more money and a new alphabet soup of programs to plug other gaps. Consider, for example, Rothstein: "Despite careful admissions restrictions, magnet programs attract the most highly motivated students, draining neighborhood schools of students and parents who could spur higher achievement levels." They are offering a false double bind: claim there is no problem, in which case you don't need those with an effective voice, and claim that the system will collapse if it is changed because problems will not be addressed. Note that Rothstein craftily avoids saying "correction" and instead says, "higher achievement levels;" other voucher critics aren't this clever.

Parents voice will be equal to or greater than that of teachers, administrators, and patronage granters. This is a surprising oversight by Hirschman. He even cites Mancur Olson's The Logic of Collective Action, so he cannot be unaware of the problem. Parents are interested in their children's education for at most 18 years (after which they get to access the public/private arena of college, where competition has apparently not had the deleterious effect), while teachers are looking at careers of 35 years or so. Parents' interest are varied (math, literacy, warehousing or baby-sitting, athletics, arts, and so on) and contradictory (I don't care about football, I want math and science, while someone else wants football, and someone else wants money spent on arts), and parents' knowledge is limited (they gather most of it second hand from mostly non-cooperative agents, which anyone who has ever been or raised a teenager will know), while teachers interest is focused (secure, high-paying jobs in a low-stress environment) and their knowledge direct. Administrators of the schools and patronage systems likewise have an advantage. Moreover, the parents' gains are likely to be shared and spread among others, while teachers' gains are going to come directly back to themselves. As a matter of fact, there is proof that teachers -- or more precisely, teachers' unions -- think the voice option works in their favor: they oppose vouchers and pay to lobby against them even though this would increase their exit options as employees!

Furthermore, Hirschman says "voice plays a more important role with respect to organizations of which an individual is a member than with respect to firms whose products he buys..." - I think even Hirschman would admit that the increasingly centralized, almost national school district has eliminated the possibility that parents are a member of the organization.

Incidentally, I am not a strong fan of vouchers. I fear that voucherization would be a Trojan Horse for regulation of the few remaining private schools. A much better solution, I think, would be to charge means-tested tuition for public schools, the same way you pay park fees. For one thing, the free breeders out there would have to carry more of their share of the load. For another, private schools would spring up as an alternative, no subsidies required. For a third, public schools would have increased budgets (or the state and local governments could offer tax breaks or redirect tax moneys).

[UPDATE] Haha, sure enough, he provides an argument on the very next page, and takes it up again later in the book. I swear I looked in the index.

Anyhow, Hirschman provides an elegant graphical proof of what happens when two consumers, one sensitive to quality and the other to price, encounter a deterioration of quality in a market which has both a higher priced, higher quality alternative as well as a lower priced, lower quality alternative. Sure enough, the quality sensitive consumer bails early. Unfortunately, he doesn't explore the situation that arises when the alternatives are both lower priced and higher quality, which is a common case with private schools. The outcome of that very distinctly depends on the assumptions, since both types of consumers stand to benefit.

Another problem I have with the analysis is that he begins with a discussion of a consumer who has greater consumer surplus and then treats this person as if they were the fastidious consumer (the implication that wealthy = quality minded, not wealthy = don't care about school quality). I think that is a mistake (a mistake I propagated above), the same mistake in fact of justifying patient waits because the doctor's time is more valuable than the person in the waiting room. Somewhere in Marshall's Principles (drat, can't find it right now), I remember that he warns that a demand schedule is not precisely the same as utility. Given the anti-voucher PR machine that operates in urban environments, graphs like the one below tell me that poor education consumers have a strong utility for education (stolen from Creative Destruction after finding it searching on something unrelated).

The Education President?

I just finished a superficial analysis of the historical budget information for the Federal Department of Education. The results? The President Bushes were the most profligate spenders since 1980.

• On average, Ronald Reagan requested increases of 8% per year, and the Congress (consisting of a Democrat House and a vacillating Senate) appropriated 6% per year.
• George H. W. Bush requested increases of 10% per year, and the Democrat Congress appropriated 9% per year.
  
• Bill Clinton requested increases of 3% per year, and the mostly Republican Congress appropriated 4% per year. That includes two years in which he requested no change in the budget (1996 and 2000) and three years (1994, 1997, and 1999) when the president's budget actually requested decreases in funding, one of which the last Democrat Congress gave (a whopping 17% decrease, which was restored (19%) the next year by the Republican Congress).
• George W. Bush requested increases of 9% per year, and the mostly Republican Congress (always Republican House, temporarily Democrat Senate) appropriated 17% per year. His last two budget requests (2006 and 2007) have been for decreases. Congress responded to the 5% reduction request in 2006 with a 20% increase. The Department spent $88 B in 2006, more than twice the 2001 appropriation of $42 B.
  

(These are current, not inflation-adjusted figures, and they don't account for the increases in students. I've written about this before, and found that real, per student federal spending was flat from 1972 through 1988, and has been increasing since then (through 2002, anyhow).)

Note: the Y-axis in the chart is the fractional increase over the prior year (delta); 0.1 represents a 10% increase, -0.1 is a 10% decrease. Blue is the President's request, maroon is the appropriation. Click the chart for a large version. The data start in 1981, no request was available for 1981, no appropriation numbers were available for 2007. All other 0's are a 0% increase.

Democrats refuse to give Bush any credit for education, but when you look at these numbers, and apply their stated standards, Bush is much better on paper than Clinton. Republicans, on the other hand, refused to give Clinton any credit for holding the line on federal spending, while continuing to believe that the current Bush believes in small government and federalism. Now that's "faith-based".

Federal spending on education, in fact, is but a drop in the bucket compared to state and local spending. In 2003 (the last year in the latest Statistical Abstract), Federal spending was $35.6 B, state was $212.9 B, and local was $185.3 B (that's "billion" with a "B") for elementary and secondary education, or just 8.2% of the total. But those federal dollars always come with a string, and the string may not make any sense in the social and cultural context of the local system. I once read that 50% of all federal dollars went to administrative requirements rather than to classrooms, and No Child Left Behind has been just the latest string.

So what?

The public education system spends too much money on too little result. Based on the spending of private and parochial schools in places like New York, our public system spends about 3x more than is necessary to provide a good education ($10k/pupil public vs. $3k/pupil private in 2000). The fact that the real, per student spending is going up with no improvement in outcome should be an indicator that something is wrong with the status quo.

Let's separate two things that desperately need it, eh? Funding is an input. Student/teacher ratios are an input. Teacher pay is an input. All of these things are indirectly related to the outcome and all have been increasing at the federal, state, and local levels. The outcome - education, ability, understanding, knowledge, however you want to measure it - has been flat or declining for years. Inputs up, outputs down: there is a disconnect that nobody on the Left wants to acknowledge. Their response, "more money", has been tried and found wanting.

If more spending is not the answer, then what is? I see two possibilities: one is to inject some competition and the other is to test outcomes. There may be others. Simply calling for "better administration" is naive; who is to say that we aren't getting exactly what we actually want, as revealed in our voting patterns?

The American Left is stridently against private schooling. Despite a cosmopolitan view on almost every other topic, on this they seem unaware that other parts of the world have well-developed private school systems. Something like 17% of all schools in India are private, 2/3 of poor students attend them, and the concurrent growth in India's tech sector has been impressive. The Japanese rely on private schools to prepare their children for the intensely competitive university exams. With the collapse of communism in East Europe, many of those countries rely on a predominantly private school system.

In this country, there are three alternatives to public schooling. One is to send your child to a private or parochial school, an option available mainly to wealthy parents (or to parents lucky enough to find a private school with a sizable endowment to allow them to offer means-tested grants, like the Albuquerque Academy has). Public school teachers consistently say in polls that they would prefer to send their children to private schools. Another alternative is to be lucky enough to live in the two or three districts in the country with a voucher program. Surprisingly, urban blacks are the most pro-voucher voters in the country despite being pro-Democrat on almost every other issue. Finally, parents unlucky enough to live somewhere else, unable to afford the existing private schools, or unwilling to tolerate the religious doctrines of the existing parochial or religious schools have increasingly turned to homeschooling. In 2003, 2.2% of all children were home-schooled. Home school parents, on average, have lower incomes than public school parents, but higher educational attainment.

I'm not even totally opposed to local and state involvement in education; I think Thomas Jefferson's proposal was pretty good:

Jefferson proposed a system under which land for the buildings of the schools would be provided by the local government. The first three years of school would be provided gratis. The next level - grammar school - would be provided free only to those too poor to afford it themselves. Furthermore, not every student would be allowed to attend: only "some one of the best and most promising genius and disposition, [would] proceed to the grammar school of his district." They would then be required to appear before the Aldermen, who might or might not appoint them to go to grammar school. All those would then attend grammar school for one or two years, "save one only the best in genius and disposition, who shall be at liberty to continue there four years longer on the public foundation, and shall thence be deemed a senior."

What I like about this is that it provides an education for all, and incentives to excel, but requires payment of those who can afford it. The status quo is a system in which the wealthy send their kids to public school, where they make sure that the best facilities are in their neighborhoods, while the poor, who have no more power to influence politicians than they do to influence other aspects of their lives, get what's left over. I don't like Jefferson's idea of creating a multi-tiered system based on testing: that's what they have in Germany, where it is subject to gaming (the wealthy make sure their less bright students get into college prep despite their test results) and punishes those whose early disposition doesn't give a full accounting of their potential. I do, however, like the idea that good performance has immediate benefits, like a scholarship to the next year of school. Means-testing for all, and merit-based scholarships means that good, poor students might actually make money going to school!

However, we have to admit that schools have become warehouses for our children largely because there is no accountability. Parents are frequently unable to determine the quality of education their children are getting, especially if they had poor educations themselves. An excellent indication of the quality of outcomes is that the fastest growing classes at the college level have been remedial math and writing classes. Despite its drawbacks, No Child Left Behind attempts to introduce some accountability into a system that has none.

I oppose federal funding for schools, in part because it brings such ill-advised strings with it. The Left opposes No Child Left Behind because it was signed by George Bush; they don't acknowledge its pedigree (it was written by Ted Kennedy), the problem it attempted to solve (the lack of accountability), and they offer no alternatives. Their most coherent complaint is that Bush, whose average budget increase requests have been 3x Clinton's, has not spent enough money on it.

The response from the Right has not been much more helpful. Creation science is not science. We don't need prayer in schools, we need school in schools. Vouchers are not the solution, they are but one possible means to a solution. Likewise, testing is a means, not a solution; while it introduces accountability, the question is - accountability of whom to what? Teachers not only teach the test, but are having "ethical lapses" (helping students to cheat or outright falsification of results).

These are difficult problems that the tribalism between the two dominant political camps is not helping to solve.

Bush Administration Spending (1 in a series)

One of the most frustrating things to me is that partisans will not be consistent. There arguments are frequently opportunistic. For example, do Democrats want Bush to be fiscally responsible, or do they want him to spend more? (Or, do they want both, and believe that this can be achieved by raising marginal tax rates?)

I have been gathering figures on spending on various programs. In this post, I have the figures for Head Start spending. The three columns are Year, enrollment, and appropriations in current $Million.

1965 561,000 96.4
1966 733,000 198.9
1967 681,400 349.2
1968 693,900 316.2
1969 663,600 333.9
1970 477,400 325.7
1971 397,500 360
1972 379,000 376.3
1973 379,000 400.7
1974 352,800 403.9
1975 349,000 403.9
1976 349,000 441
1977 333,000 475
1978 391,400 625
1979 387,500 680
1980 376,300 735
1981 387,300 818.7
1982 395,800 911.7
1983 414,950 912
1984 442,140 995.8
1985 452,080 1075
1986 451,732 1040
1987 446,523 1130.5
1988 448,464 1206.3
1989 450,970 1235
1990 548,470 1552.01
1991 583,471 1951.8
1992 621,078 2201.8
1993 713,903 2776.3
1994 740,493 3325.7
1995 750,696 3534.1
1996 752,077 3569.3
1997 793,809 3980.5
1998 822,316 4347.4
1999 835,365 4658.2
2000 857,664 5266.2
2001 905,235 6199.1
2002 912,345 6536.6


A few points from this:

• Real, per student spending was flat throughout the 70s, decreased slightly through the 80s, but has exploded since then. The apparent reduction in real, per student spending under the mostly-Democrat Congress and Ronnie was due to increasing enrollment and inflationary erosion, not actual cuts.
• The first president Bush began the prolific spending. It subsided slightly at the beginning of the Republican Revolution in 1995/96, but picked up again afterwards.
• The current enrollment and spending levels are higher than they were at the end of the Clinton Administration and nearly twice as high as they were in the first years. Are there really that many more poor children? If so, then the Clinton miracle economy must not have treated children that well. Or is means-testing becoming less popular or less feasible?