The Structure of School Choice

Wednesday, October 30, 2002

School Choice

Will school choice, as its critics charge, worsen problems of class and race, benefiting society’s better-off at the expense of the underprivileged—or is school choice a tool to promote social equity, opening more educational opportunities to poor and minority students ill-served in poor-performing public schools today?

The answer to this critical question depends on how choice programs are designed. The simple way to frame the issue is to realize that school choice always operates within a structure—a framework of rules—which in turn affects the kinds of outcomes choice ultimately generates. In some structures, choice will lead to equity problems. In others, it will not. In still others, it will tilt the playing field in favor of the disadvantaged and aggressively promote the cause of social equity.

My aim here is to highlight two factors that lie at the heart of the choice issue but that are often overlooked in the contentious school choice debate. The first has to do with the structure of choice, the second with the choice of structure. Each is key to understanding the role of choice in American education and its impact on educational equity.

The Structure of Choice

From its modern origins in the early 1900s, America’s public education system was designed to be a purely governmental system in which markets play no role at all. The idea was that government-run schools would act as local monopolies within their own geographic areas, with children assigned to schools by the education bureaucracy.

Nearly a century later, though many details have changed, the same top-down system remains in place. For all the debate on school choice, there are now just four public voucher programs in operation (all of them small) and some 2,000 charter schools, enrolling more than 500,000 kids nationwide. Even counting magnet schools and other programs of public school choice, such reforms are a small drop in a very large bucket.

Even a purely governmental system, however, does not eliminate all forms of choice. In effect, an implicit choice system—not consciously designed by anyone but present nonetheless—exists alongside the formal governmental system. Although its outcomes for society are accidental and unplanned, they are hugely important. They can also be perverse, generating widespread equity problems that have worsened class and racial divides in American society.

Consider two factors in the present, unplanned system:

• Public school parents may typically not be able to choose their children’s school, but they are free, within economic constraints, to choose where they live. Many families “choose” a good public school by buying or renting a house in the right school district. Not surprisingly, parents who exercise residential choice tend to have higher incomes and, because earning power tracks education levels, be more educated than the average. The upshot is that residential choice injects a serious social bias into the current education system: The best schools tend to attract the educational “cream of the crop,” whereas the worst schools are filled with disadvantaged children. This is perhaps the most fundamental skimming problem in American education today, and the most socially destructive.

• Public schooling, of course, is not compulsory. Parents can choose to send their kids to private schools, but private schools are costly. In practice, therefore, this equal right to a private education is sharply conditioned by earning ability. As a result, the current system promotes a class bias in the types of parents who go private. This factor, reinforced by the impact of residential choice, tends to concentrate poor and minority kids in our worst-performing urban schools. Here again, the existing system of unplanned choice produces a skimming effect that adds to social inequities.

Defenders of the public schools tend to attribute these effects to choice per se, arguing that if more choice is introduced into the current system—through vouchers, say—equity problems will only get worse. Today’s equity problems, however, are actually due to the constraints on choice—or, more accurately, to the way the current system of top-down control creates cost barriers for parents who wish to exercise their choice to go private. The irony is that by trying to keep markets out of education, the system’s designers unwittingly created a structure that nevertheless allows a form of choice—but one that perversely undermines the system’s most fundamental goals.

The ideals Progressive educators brought to the public education project 100 years ago—visions of common schooling, of social equity, of democratic governance—are inspiring. But they are poorly met in practice. A century of theory, research, and experience has taught us that top-down governmental structures are extreme forms of social organization that are often overly costly and unproductive.

Under the current top-down arrangement, public schools are guaranteed students and resources regardless of how they perform. As a result, they have few incentives to produce high-quality education, to respond to parents, to allocate funds efficiently, or to innovate in socially productive ways—for nothing bad happens to them if they don’t, and nothing good happens to them if they do.

When parents are allowed to choose, the situation is very different. No longer a captive clientele, parents can remove their children from schools they consider undesirable and seek out schools they believe are better. When schools have to compete with one another for parental support, this is a strong incentive to be responsive and efficient—an incentive that leads, over time, to a more effective, more innovative population of schools.

The Choice of Structure

It is tempting to imagine that the free market is without structure, unconstrained by any overarching set of rules. Even libertarians, however, don’t see it this way. They recognize that, for markets to work properly, a society requires well-defined property rights backed by the rule of law. It is not markets alone, therefore, but markets embedded in a governmental structure that yield such wondrous results. When economists and other market advocates think about education, they recognize the great power of markets to promote incentives and efficiency—but they also understand that, if markets are to promote desired social values in the most effective ways, they must be constrained and channeled by a set of social rules chosen with those values in mind.

The idea of choice opens up new vistas once we recognize that a choice system can depart rather substantially from a free market in ways designed to promote social equity. To get a sense of what is possible, consider the following questions that need to be considered in constructing a choice system.

Who should qualify for a voucher? The free market ideal is a universal voucher system, open to all children. Such a system would presumably generate beneficial competition on the part of schools, both public and private. It may also promote social equity by providing a choice option to poor and minority families stuck in low-performing schools.

Two objections arise. First, one might ask why society should offer vouchers to families that demonstrate no urgent economic exigencies. To promote equity with certainty, according to this argument, a voucher system might best be targeted on those most in need.

The second objection—which may persuade even some people who favor universalism as a long-run goal—is the risk in adopting a universal system through a massive, all-at-once shift that could involve vast upheavals and uncertainties. Better to start with small pilot programs, see how they work in practice, and move incrementally from there. Here, too, it makes perfect sense to focus initial efforts on the neediest kids in society. This is where society clearly gets the most benefit, and where the risk of failure—because things are currently so bad for these kids—is exceedingly small and well worth bearing.

What should the amount of the voucher be? The standard free market solution is to give all kids vouchers of the same amount. Yet equality is not always equity. We know that disadvantaged kids are more costly to educate than other kids are; the same is true for kids with learning disabilities and behavior problems. It may make sense to give bigger vouchers to these kids than to others—both to compensate schools for the true costs of educating them and to make these children more attractive as clients so that schools will actually compete to serve them. It might make sense, for instance, to institute a sliding scale, with the value of the vouchers being quite high for the poor, dropping steadily until at some level of family income they become zero.

In any case, the absolute amount of the voucher is critical. The bigger the voucher, the greater the incentive for new schools to emerge, and the greater the choices open to parents.

Should parents be allowed to add on? The free market solution is to allow parents to add on to their vouchers, as it gives parents more freedom and more choice. The downside is obvious: More-affluent parents would be better able to “top off” their voucher to buy their way into expensive schools, leaving the poor behind in a two-tiered system that would reinforce class cleavages.

In practice, programs in Milwaukee, Cleveland, and Florida have chosen to prohibit add-ons. The equity argument in so doing is clear enough but not without its own costs: Prohibiting add-ons would prevent some parents, perhaps many, from choosing a school they really want for their children, and this in itself is a negative. It would also put an upper bound on the tuition that can be charged by schools participating in the voucher program; and unless the amount of the voucher is high, the danger is that only schools offering a basic, low-cost education could participate. The larger the voucher, the less often such problems would arise.

Should private schools control their own admissions? In the free market ideal, private schools should be allowed full autonomy to admit students on their own criteria. The danger, of course, is that private schools might favor children who are easier to educate, who have more-affluent parents, or who come from the same religious or social group, undercutting access for poor and minority children.

Although worth recognizing, this danger tends to be overstated. The experience in Milwaukee, for instance, where low-income kids use vouchers to attend nearly 100 different private schools, is that there have been few parental complaints of discrimination. Still, if the possibility exists, what design options might help deal with the danger of discrimination? Designers could require participating private schools to select some portion of their students by lottery or reserve a certain percentage of each school’s slots for low-income kids. There are costs, however, to aggressively regulating private school admissions. One of the reasons many private schools are so successful is that they have the autonomy to define their own missions and programs as they see fit, and selecting appropriate students is an integral part of that process. By imposing a randomly selected student body on private schools, an important foundation of their strength may be lost. Any restrictions, therefore, should be carefully considered.

Should private schools be held accountable by government? In a free market, private schools would not be regulated. The idea is that in the educational marketplace, private schools are automatically held accountable from below—by parents who leave bad schools and seek out better ones.

Not everyone has so much faith in markets, however. This is true not just for many liberals but also for many people who have no ideological take on markets at all and are moved by very practical concerns. For instance, taxpayers who foot the bill for education may desire some concrete accounting that shows their money is being put to good use: hence requirements for curriculum and standards, teacher qualifications, annual audits of finances, periodic testing of students, and public information on school performance, to name just a few measures. Yet because the autonomy of private schools is pivotal to their strength, there are good reasons for keeping accountability regulations simple and basic.

Should religious schools be included in a voucher system? The free market answer is that religious schools of all types should be included, as part of the general aim of providing families as much choice and diversity as possible. Others argue that religion should be kept out of education, particularly when funded with taxpayer money, even though the Supreme Court, in the recent Cleveland voucher case, held such programs to be constitutional.

As a practical matter, however, the exclusion of religious schools has enormous consequences. Under the current system, the vast majority of private schools are religious. The reason is simple: Public schools offer a nonreligious education for free, and nonreligious private schools have a hard time competing with that. With a full-blown voucher system, this would presumably change, as there would be a greater demand for nonreligious private schools; over the long run, assuming the vouchers were big enough, there would be an increase in their supply. In the short run, however, if religious schools are excluded from a voucher system, there will be few private schools to choose from. In effect, to exclude religious schools is to eliminate most choice.

How should government funds be divided between public and private schools? The free market answer is that money should follow
children to their schools of choice—public or private—the gain or loss of funding providing the incentive effect of competition. Not everyone is persuaded by this line of reasoning. Indeed, the main argument of voucher opponents is that vouchers would “drain money” out of the public schools and make it impossible for them to improve. This argument is misleading because it ignores the fact that the public schools would still have as much (or more) money per child—they would simply be losing money for the children they no longer have to educate. It is true, however, that public schools have fixed costs—and especially if only a small number of kids go private, districts may find it difficult to achieve many cost savings, particularly in the short run.

One response is to design a system that would hold back a portion of the funding that would otherwise leave the public schools when a voucher student departs. In one sense, public schools would thereby get paid for children no longer enrolled there, but the money would cover fixed costs the districts continue to incur after the children leave.

The Politics of Structural Choice

Even this brief discussion of design options should make it clear that there is no single free market approach. The challenge is to strike the right balance—to unleash what choice and competition can contribute, while channeling those forces in socially desirable directions.

In principle, this objective is straightforward enough. In practice, however, as in any exercise in public policy, virtually every decision is determined through the political process—which is heavily shaped by power and self-interest. In education, choice reform collides with the extraordinary political power of the teacher unions, which find the present purely governmental system of top-down control a safe, noncompetitive environment in which to organize teachers and engage in collective bargaining.

School choice, and especially vouchers, would change all this. A voucher system would allow children and money to flow from public to private schools, creating a more decentralized system in which teacher unions would have less power and control.

Teacher union opposition has little to do with how vouchers affect the opportunities of children, the quality of the schools, or any other basic social values. If it could be shown with 100 percent certainty that vouchers benefit students and social equity, the unions would still oppose them. Indeed, the unions oppose vouchers even when limited to the poorest students in our nation’s worst schools.

As a result, real-world choice systems are but a pale reflection of what is possible. The product of politics, they are subject to influence not just by school choice advocates but also by their political enemies—who want those programs to be limited and even to fail. Union opposition holds true for charter school experiments as well. When teacher unions have failed to stop the adoption of a charter plan—or when they have “supported” charters as a means of heading off vouchers (which they fear even more)—they have consistently pressured for structures that limit the program’s scope and impact. Among other things, they lobby for low ceilings on the number of charters that can be created, for requirements that charter teachers be unionized and part of the district bargaining agreement, for low levels of funding, for no assistance with building or set-up costs, and for the extension of as many district regulations and controls as possible. Charters are on the rise nationwide. But due to the power of the unions and their allies, most are constrained by charter laws that sharply restrict how much choice and competition the new schools can really bring.

Nonetheless, even the imperfect choice systems that result are preferable to a purely governmental system that makes no use of markets whatever. And as the choice movement grows in political power, which—given the Supreme Court’s recent decision upholding the Cleveland voucher program—seems a good bet, these imperfections can be addressed through design changes that allow choice and competition to realize their full potential.


Even if one puts almost exclusive emphasis on social equity, it is difficult to argue that American education should not move toward a greater reliance on choice and competition. The top-down system we have now is a legacy of the last century and traps “modern” education in an antiquated iron cage. Whatever may have been the case in the early twentieth century, there is today a mountain of social science evidence demonstrating that market forces are powerful engines of efficiency, innovation, and social welfare. To make no use at all of market forces is simply a mistake.

It is also a mistake, however, to think that the only alternative to top-down government control is a pure free market, in which the entire education system is privatized and families are thrown into the marketplace to fend for themselves. What choice advocates almost always have in mind, and what choice critics would be wise to consider (for their own social purposes), are forms of organization that lie between the extremes and involve important elements of both government and markets.

Seen in this light, two factors are fundamental. The first is that there is always a structure to any choice system—a specific framework of governmental rules—the details of which determine how choice and competition operate, and what social values the system promotes.

The second factor is that, just as there is a structure of choice, so too there is a choice of structures—one that allows a balancing of equity issues against considerations of efficiency and incentives.

The challenge for American education at this juncture in history is to move beyond ideological battles over markets versus government and find ways to harness the social force of choice in the service of social values we all share.