FROM ISSUE NUMBER 5 ~ FALL 2010 GO TO TABLE OF CONTENTS

TEXT SIZE A A A

Does School Choice "Work"?

FREDERICK M. HESS

These would seem to be dark days for the school-choice movement, as several early champions of choice have publicly expressed their disillusionment. A few years ago, the Manhattan Institute's Sol Stern — author of Breaking Free: Public School Lessons and the Imperative of School Choice — caused a stir when he backed away from his once-ardent support. Howard Fuller, an architect of Milwaukee's school-voucher plan and the godfather of the school-choice movement, has wryly observed, "I think that any honest assessment would have to say that there hasn't been the deep, wholesale improvement in [Milwaukee Public Schools] that we would have thought." Earlier this year, historian Diane Ravitch made waves when she retracted her once staunch support for school choice in The Death and Life of the Great American School System. "I just wish that choice proponents would stop promising that charters and vouchers will bring us closer to that date when 100 percent of all children reach proficiency," she opined in her blog. "If evidence mattered, they would tone down their rhetoric." Harvard professor and iconic school-voucher proponent Paul Peterson has characterized the voucher movement as "stalled," in part by the fact that many "new voucher schools were badly run, both fiscally and educationally," and in part because results in Milwaukee were not "as startlingly positive as advocates originally hoped." Likewise, Peterson argues, "the jury on charter schools is still out."

To many who hold out hope that choice can help fix what ails America's schools, these hedges and reversals have been startling. And yet, looking back, it is hard to see how they were not inevitable. For decades, school-choice advocates have seemed bent on producing this hour of disappointment.

There has been, for instance, a tendency to vastly overpromise. In 1990, the same year that Milwaukee's tiny voucher program launched the school-choice debate, scholars John Chubb and Terry Moe argued in their seminal volume, Politics, Markets, and America's Schools: "Without being too literal about it, we think reformers would do well to entertain the notion that choice is a panacea...It has the capacity all by itself to bring about the kind of transformation that, for years, reformers have been seeking to engineer in myriad other ways." Chubb and Moe are gifted thinkers, and their book was a tour de force, but this may have been some of the worst advice that school reformers ever got.

The search for that panacea, and the insistence that it must be just around the corner, have been destructive distractions. They have led champions of market-oriented reforms — and so also allowed skeptics — to adopt a ludicrous standard for judging whether school choice "works." Since reformers have suggested that the mere presence of choice will bring about dramatic improvement in schools, the expectation has been that the simple fact of having an alternative — even inadequately funded vouchers, or charter schools hog-tied by regulation — should yield demonstrable gains in academic achievement. And so, for the past 20 years, the question of whether school choice "works" has been understood to mean simply whether a school-choice program boosted reading and math test scores in a given year.

The need to answer this question with an unequivocal "yes" has forced choice advocates into bizarre contortions and short-sighted thinking. The same can be said of opponents, whose insistence, in the face of all evidence, that school choice is harmful has led them to ignore its real achievements.

Particularly problematic is how this way of thinking has caused school-choice proponents to ignore crucial questions of market design and implementation — especially the extent to which reforms have, or have not, created a real market dynamic in education. The chief promise of choice, after all, was that it would displace ossified, monopolistic school bureaucracies or at least inject into them a degree of flexibility, competition, and quality control. The question education reformers should be asking, then, is not simply whether choice "works" — because choice is neither the sole end of nor a sufficient means for bringing about successful market-based reform.

The questions to focus on are when, how, and why deregulation and monopoly-busting improve the quality and cost effectiveness of goods and services — and whether they can do the same for K-12 schooling. What would a vibrant market in K-12 education look like? To what degree has it really been tried? What needs to change in order to bring about such a market, and how would we assess whether it is in fact improving the education received by children in America's schools?

Before answering these crucial questions, however, it is important to understand how reformers came to paint themselves into a corner.

A LEGACY OF OVERPROMISING

Educational choice is hardly a modern innovation. In some ways, it dates back at least as far as ancient Athens with its marketplace of sophists and philosophers. In the Anglo-American context, explicit proposals for state-funded arrangements that would let parents choose how their children would be educated can be traced to the writings of Thomas Paine in the late 18th century, and of John Stuart Mill in the 19th century. Both men thought it appropriate for the state to ensure that young people were given at least a basic level of education — but both also felt that this aim should be advanced through private arrangements, rejecting the notion of state educational monopolies.

School choice as we think of it today originated with an essay penned in 1955 by economist Milton Friedman. Friedman's argument was that a voucher system of education — one in which the government's role would be limited to providing funding and setting basic standards for "approved" educational institutions, while parents would retain the right to determine which of these institutions could best educate their children — would promote both equitable and efficient schooling. The first substantial effort to translate the concept into policy took shape in President Lyndon Johnson's Office of Economic Opportunity, where a cluster of social scientists toyed with vouchers as a politically viable and promising alternative to school busing. The efforts of the OEO wonks resulted in a tiny, dead-end "school voucher" pilot program in Alum Rock, California; the experiment, which ultimately looked a lot more like a magnet-school program than anything we would today call a "voucher" system, resulted in little besides a mammoth-yet-banal study by the RAND Corporation.

The modern school-choice movement did not begin in earnest until the 1980s, when it grew out of the overlapping efforts of four distinct champions. One was Ronald Reagan, who had reached out in 1979 and 1980 to disaffected Catholics who were dismayed by judicial assaults on school prayer, and were seeking assistance with parochial-school tuition. Another was an ascendant Republican Party, seeking a proactive, market-friendly agenda. The third was legendary American Federation of Teachers president Al Shanker — who was pitching "charter" schools as a way for teachers to establish schools in which they would have heightened autonomy, and in which they could insulate themselves from district bureaucracy. And the fourth was a group of frustrated African-American leaders seeking good, safe schools for urban children stuck in horrendously mismanaged districts.

The most politically marketable of these advocates were the urban black leaders — and so, before long, the various parties found common ground in emphasizing the language of empathy and rights. The case for school choice was thus not argued in terms of efficiency or deregulation, but instead presented as a moral imperative — an obligation to give poor, black inner-city parents the kinds of educational choices taken for granted by suburban home owners. This "social justice" rhetoric was the mantra of the school-choice movement when Wisconsin enacted the Milwaukee voucher program in 1990; it has been the reigning justification ever since.

This approach helps explain why choice advocates — inclined to approach choice-based reform not as a regulatory question, but as one of justice and rights — have spent so much less time considering the dynamics of deregulation than have pro-market reformers in sectors like transportation, telecommunications, and cable television. Because education reformers have approached choice not as a matter of political economy but as a moral crusade, they have favored grand, sweeping claims over empirical reality.

For example, in a celebrated 1999 article in the New Republic, author and former Al Gore advisor David Osborne boldly declared, "Those who invented charter schools...wanted to improve all 88,000 public schools in the country [and]...empirical studies have demonstrated that, indeed, competition works just as the reformers predicted." In remarks that proved unduly optimistic, Wisconsin governor Tommy Thompson declared in his 2001 "State of the State" address: "Nowhere in America does a parent have more choices than in Milwaukee, Wisconsin. And it's making all the difference...There is no doubt in my mind that Milwaukee will become the national model for renewing urban education in America within a few years." Such rosy assessments ensured that more realistic appraisals would inevitably disappoint.

Another consequence of the empathy-and-rights approach has been that the few education economists sympathetic to market-based reforms have felt compelled to devote their energies to demonstrating the superiority of choice-based systems, rather than to exploring and explaining the complexities of market-based reform. They have devoted limited attention to studying the political economy of K-12 education or potential sources of market failure.

Meanwhile, with a few exceptions — most notably the NewSchools Venture Fund and the Center for Reinventing Public Education — earnest reform advocates focus on selling the appealing promise that choice "works" rather than on the more arduous task of tackling K-12 education as a serious deregulatory project.

As a result, crucial questions have received scant attention — including regulatory and licensure chokepoints; the tendency of successful non-profit charter organizations to grow slowly; the dearth of information regarding the quality of providers; a third-party financing system that gives consumers no reason to weigh cost considerations; when and why private schools add new capacity in response to voucher programs; and the way in which statutes and collective-bargaining agreements limit how school districts respond to competition.

In any other deregulation project, such questions would be front and center. But in the effort to establish a genuine marketplace in education, they have been largely ignored.

THE DATA ON CHOICE

This is not to say, of course, that straightforward evaluations of student achievement should not play a central role in assessing market-oriented reforms. But it is unwise to interpret these data without a broader appreciation of how markets work.

To a frustrating degree, the conclusions one draws from the educational-performance evidence depend on which experts one trusts. And different credentialed, respected scholars have offered very different takes. For instance, Jay Greene — chair of the University of Arkansas's Department of Education Reform, and a widely recognized authority on school choice — argues that research shows unambiguously that "vouchers have positive effects for students who receive them." The only question, Greene adds, "is in regard to the magnitude of vouchers' benefits." On the other hand, Princeton economist Cecilia Rouse and Chicago Federal Reserve economist Lisa Barrow characterize the evidence rather differently; last year, they concluded that most of the small gains made by voucher students "are not statistically significant from zero."

More recent studies have mostly added to the ambiguity. In 2009, the Center for Research on Educational Outcomes (CREDO) — generally regarded as a pro-school-choice organization — issued a controversial study of charter-school performance in 15 states and the District of Columbia. The study found that 17% of charter schools outperformed local district schools, 46% performed similarly, and 37% performed worse than local district schools. CREDO's conclusion was that the overall picture shows "wide variation in performance."

Earlier this year, University of Arkansas professor Patrick Wolf — principal investigator on a major study of school-voucher effects in Milwaukee — reported that there were no significant differences in achievement between students who received vouchers and those who did not. Wolf summarized, "At this point the voucher students are showing average rates of achievement gain similar to their public school peers."

And this summer, two long-awaited studies on school vouchers and charter schooling issued their final analyses. In July, the Institute of Education Sciences released the multi-year "Evaluation of Charter School Impacts" study, which examined student performance in 36 charter middle schools across 15 states. The study found that, on average, the charter schools were "neither more nor less successful than traditional public schools in improving student achievement, behavior, and school progress" (though admission to a charter did "consistently improve both students' and parents' satisfaction with school"). The study also found that "charter schools serving more low income or low achieving students had statistically significant positive effects on math test scores, while charter schools serving more advantaged students — those with higher income and prior achievement — had significant negative effects on math test scores." It is worth noting, too, that in order to participate in the study, the charter schools needed to have enough excess demand to require an admissions lottery — meaning that the charters evaluated were those that parents most wanted their children to attend. If oversubscribed schools are typically better than charters with available seats — which seems a perfectly plausible assumption — then the study may actually overstate charter-school quality.

In June, the federally mandated study of the Washington, D.C., Opportunity Scholarship Program — another evaluation led by Patrick Wolf — also issued its final report. Established in 2004, the Washington scholarship program provided vouchers for up to $7,500 per child per year, which could be applied to tuition, transportation, and other fees required to attend the participating private school of a family's choice. Supporting between 1,700 and 2,000 low-income D.C. children a year, this was the first federally funded voucher program in the nation.

Wolf's team tracked educational outcomes over four or five years for 2,300 public-school students who applied for the scholarships, which were awarded by lottery. The researchers found that the reading and math scores of lottery winners were not statistically different from those of the control group at the conventionally recognized 95% confidence level (though their reading scores were higher at a 94% confidence level). As Wolf has explained, "A reasonable person would conclude that the voucher students made small gains in reading due to the program...[even if the] gains were modest and somewhat fragile." On a more upbeat note, students who won the lottery and used the vouchers to attend private schools were more likely — by 21 percentage points — to receive a high-school diploma than were students who lost the lottery.

The mixed findings suggest that simply legislating "school choice" programs, or enrolling a child in a charter school, will have no obvious short-term impact on achievement. But choice does consistently increase parental satisfaction, and there is evidence that carefully designed choice programs — like Washington's voucher program — may modestly bolster achievement and substantially boost graduation rates. Indeed, a fair-minded observer could read the middle-school and D.C. studies and conclude that choice "works" — almost by definition — if the goal is to get low-income children out of terrible urban school systems and into high-quality private schools where they can learn safely and increase their odds of getting high-school diplomas. Even the most determined choice skeptic should be able to acknowledge this as an improvement over the status quo.

This case was laid out most forcefully by Paul Peterson, the University of Chicago's William Howell, and two colleagues in their invaluable 2002 book The Education Gap. Reporting findings from randomized-control trials in New York City, Washington, D.C., and Dayton, Ohio, the authors found that attending a private school through a voucher program had a significant positive effect on the achievement scores of African-American students (though not on those of other students). Families who won the lotteries to attend these schools were also much more satisfied with their schools, and found them far safer, than families who lost the lotteries. While Princeton economist Alan Krueger later made the case that the achievement effects in New York City could be washed away with enough data manipulation, the straightforward and plausible conclusion is that students stuck in failing urban schools often benefit from moving to high-quality private schools.

More recent research has similarly found clear academic benefits for students attending choice schools. In 2009, Stanford University economist Caroline Hoxby examined the gains made by New York City students who won admission to charter schools from 2000 to 2007, and compared them to those of students who did not. Hoxby found that a student who attended a charter school from kindergarten through eighth grade would substantially outperform his district-school counterparts; on average, African-American students who enrolled in charters closed 86% of the black-white achievement gap in math and 66% in reading. And in another 2009 study, Harvard University economist Thomas Kane found "large positive effects for charter schools" for Boston students in both middle school and high school. Charter students, Kane discovered, had larger gains in reading and much larger gains in math than their peers in district schools. The largest observed gains were in middle-school math, where the effects of charter-school enrollment amounted to half of a standard deviation. These effects were large enough to lift a student from the 50th percentile of performance to the 69th percentile in a single year.

In a 2010 study, Mathematica Policy Research examined student achievement in the nation's most prominent charter-school network: the Knowledge Is Power Program, or KIPP. A non-profit founded in 1994, KIPP operates a national network of 99 schools serving mostly low-income children. KIPP schools feature strict discipline, high expectations, a longer school year, and a school day that runs from 7:30 a.m. until 5 p.m. (typically including Saturday classes). KIPP middle schools begin in fifth grade — and the Mathematica study showed that, by seventh grade, half of the KIPP schools evaluated showed growth in math scores equal to an additional 1.2 years of schooling. KIPP reading gains reflected an additional three-quarters of a year of growth. In general, Mathematica reported that middle-school students in KIPP academies significantly outperformed similar public-school students in both reading and math.

The positive findings from New York, Boston, and KIPP should not be too surprising. Students who switch from troubled schools to high-quality charter alternatives are likely to benefit — especially in cities like Boston and New York, where caps on charter schooling and an abundance of talented charter operators have produced a rich crop of terrific schools. By the same token, however, proponents of market-based school reform should not be surprised that the results may look very different in other environments. Only by stepping back from the notion that "choice" itself is a panacea, and instead embracing the contingent nature of choice's impact, can we make sense of when and why choice "works."

THE COMPETITION QUESTION

A key feature of genuine markets is, of course, competition. In evaluating whether choice "works," what matters is not only whether escape routes to private or charter schools offer some students better educational alternatives, but also whether school-choice programs make traditional district schools better.

Any observer who takes market theory seriously would probably respond, "Of course they do." But because of the peculiarities of American education, the answer is hardly so obvious. After all, market dynamics depend upon consumer behavior, regulatory frameworks, labor-market considerations, and incentives and consequences for producers and consumers. Competition matters only when it pinches, and the reality is that competition in K-12 education has not yet been given a robust test.

Still, there is some evidence that districts and schools may respond to even the mild competitive pressures that choice currently exerts. In 2003, Jay Greene examined Florida's A+ voucher program and reported that those low-performing schools that risked having their students granted vouchers to attend private alternatives were improving "in direct proportion to the challenge they face[d] from voucher competition." In 2005, scholars Paul Peterson and Martin West reported similar findings. Under the accountability standards imposed by Florida in 2002, students at public schools that received "F" performance grades became subject to the threat of vouchers if they continued to perform poorly; Peterson and West concluded that students at schools put under the gun "performed at a higher level in the subsequent year than did students at similar schools not so threatened."

More recent research confirms the same patterns. After the Florida Supreme Court ruled in 2006 that the portion of the A+ program that funded private-school vouchers was unconstitutional, the state created an alternative program (which relies on private funding incentivized by tax credits to corporations). In June of this year, economist David Figlio released a study of the new voucher program reporting that students in Florida public schools with a diverse array of private schools in close proximity showed slightly larger achievement gains than students in public schools with fewer nearby alternatives. Figlio also determined that student gains were larger in those schools at risk of losing state funds tied to the proportion of low-income students they enroll. Figlio concluded that the public schools' response to competition was real, although limited: "What we find is certainly positive and statistically strong," he explained, "but it's not like public schools are revolutionizing overnight because of this."

One challenge in interpreting these results is gauging whether the market-induced improvements reflect attempts to fundamentally rethink or re-engineer a school or district, or merely a re-allocation of effort and resources from untested activities to tested ones. This matters a lot, because quick-fix measures — like, say, shifting time from science or art to reading instruction — may improve student test performance (and thus answer a competitive threat) without signifying any attempt to boost productivity or overhaul cost structures, staffing, operations, or management. Moreover, in most cases, district responses to choice-induced competition have primarily been changes in marketing and outreach — such as the distribution of t-shirts and ads on local billboards intended to persuade parents to keep their children in their local public schools. Choice advocates have historically erred in reading these developments as signs of bigger changes to come.

It has been a mistake, in other words, to expect public schools to behave like the private sector — where competition, investor demand, and personal consequences for success or failure drive executives to press on productivity and the bottom line, and where executives have substantial leeway to remove, reward, and otherwise recognize employees based on their contributions to organizational improvement. In systems choked by politics, bureaucracy, collective-bargaining agreements, and institutionalized timidity, there is little incentive or opportunity to react to competition in these ways.

To get schools to respond more meaningfully to competitive pressure, incentives and rules must be changed in order to ensure that the competitive pressure is actually felt. Consider that today's charter schools get about 75 cents for every dollar that district schools receive, and that the per-pupil funding levels of the voucher programs in Washington, D.C., and Milwaukee amount to less than 50% of district per-pupil spending. This funding disparity prevents public-school alternatives from mounting serious challenges to traditional district schools.

Moreover, the D.C. voucher program capped enrollment at about 3% of the District of Columbia's student population, and there was no risk of monetary loss to the school district if students departed for private schools. Indeed, the compromise that allowed the voucher-program legislation to pass required that D.C. public schools receive additional funding, even as they would no longer bear the expense of educating the voucher students. The initial sum was an extra $13 million a year; this figure was eventually boosted to $40 million per year after Democrats took control of Congress in 2007.

Milwaukee's public schools have been similarly insulated from the consequences of losing students to the Milwaukee Parental Choice Program: Even as the tiny pilot grew from 337 participants in 1990 into a program that enrolls 20,000 students today, the Milwaukee public-school system has remained largely unscathed. Since 1990, while enrollment has dipped, the district has boosted per-pupil spending by more than 80% (from $6,200 to more than $11,700), and increased the teacher work force (from 5,554 to 5,768). This is choice without consequences — competition as soft political slogan rather than hard economic reality.

As a result, despite hopes that school choice could "all by itself" bring about the other changes reformers have battled for, choice has not necessarily changed incentives or dynamics. Imagine a private-sector manager who knows that gaining or losing customers will have little or no impact on his salary, performance evaluations, or job security. Bizarre as it seems, this is exactly how "competition" generally works in K-12 education today. When a principal loses dozens of students, her evaluations, job prospects, and salary remain unaffected. And a principal who competes successfully is typically rewarded with nothing more than the joys of a more crowded cafeteria.

Of course, none of this is an indictment of school choice or of market-oriented reforms; rather, it suggests that much of our policymaking to date has tended to reflect impassioned hopes instead of cool calculations. If every dollar spent on a student followed him when he changed schools — a state of affairs that exists nowhere in this country today — the verdict on choice-inspired competition would likely be quite different. That difference would be sharper still if the laws and contracts that protect teacher and administrator jobs and salaries, and that handcuff managers, were changed — or if changes in school enrollment became a significant criterion for evaluating superintendents and principals.

The fiscal crunch in which many state and local governments now find themselves, or the moves in some jurisdictions to reform teacher tenure and pay, may make it possible to remove some of the insulation that has protected educators for so long. And because school districts are political entities, an exodus of students has the potential to spur useful change by altering the local political calculus. The departure of one-third of students to charter schools over the previous decade, for instance, helped create the conditions that led Washington, D.C., mayor Adrian Fenty to appoint the hard-charging Michelle Rhee as schools chancellor in 2007. For now, however, competition does not appear to much perturb most public-school administrators — especially as many superintendents and school boards seem perfectly content to run slightly smaller districts with proportionally fewer dollars.

The biggest mistake pro-market school reformers have made can thus be put simply: They have mistaken choice for competition. The conviction that school choice constitutes, by itself, a market solution has too often led reformers to skip past the hard work necessary to take advantage of the opportunities that choice-based reform can provide. Choice is merely part of the market equation; equally crucial are the requirements that market conditions permit high-quality or cost-effective suppliers to flourish, that regulation not smother new entrants, and that rules not require inefficient practices or subsidize also-rans.

Note that reformers rarely focus on "choice" when promoting market-based improvements to other sectors; in earlier decades, reformers didn't speak of "telecommunications choice" or "airline choice." Rather, they talked of "deregulation." Implicit was the understanding that deregulation involves more than the mere proliferation of options, that dynamic markets require much more than customers' choosing among government-operated programs and a handful of non-profits, and that vacuums in a particular sector will not naturally or necessarily be filled by competent or virtuous actors. Whether dealing with nascent markets in Eastern Europe in the 1990s or the vagaries of energy deregulation, reformers have struggled to nurture the institutions, incentives, and practices that characterize healthy markets. Markets are a product of laws, norms, talent, information, and capital, and the absence of these can readily yield market failures — not because markets do not work, but because markets are not a magical salve.

Just as school improvement does not miraculously happen without attention to instruction, curriculum, and school leadership, so a rule-laden, risk-averse sector dominated by entrenched bureaucracies, industrial-style collective-bargaining agreements, and hoary colleges of education will not casually remake itself just because students have the right to switch schools. Happily, in recent years, a growing number of thoughtful scholars — like Andrew Coulson, John Merrifield, Terry Moe, Jay Greene, Patrick Wolf, and Paul Hill — have paid increasing attention to these questions of market structure and design. But such thinking remains the exception, not the rule.

MAKING CHOICE WORK

So, taking account of all of this, does school choice "work"? The question needs to be answered in three parts. First, for poor parents trapped in dangerous and underperforming urban school systems, it is pretty clear that school choice works. The evidence is reasonably persuasive that access to private schools and charter schools increases the likelihood that their children will fare well on reading and math tests or graduate from high school. And even if those results do not materialize, the parents are more likely to be satisfied with their children's schools and to regard them as safe.

Second, school choice can help make possible more coherent, focused schools. When families and teachers are assigned to schools based upon geography or bureaucratic formulas, it becomes difficult to forge the kind of agreement needed to establish strong discipline or clear expectations. The opportunities that choice creates for school leaders to recruit like-minded teachers and families — and then to set clear norms around conduct, learning, and pedagogy — can be a powerful tool. Still, their impact ultimately depends on effective use by savvy school leaders — as these opportunities in themselves surely will not automatically yield better schools.

Third, it is far from clear that school choice will necessarily offer broad, systemic benefits. Choice has not inspired hordes of charter-school operators to develop outstanding alternatives; there is no evidence that charter schools, on average across the nation, are better than district schools. Moreover, there is (at best) only very modest evidence that choice programs, in and of themselves, prompt school districts to become more productive or cost effective. There is, however, fairly clear evidence that school districts do respond under sufficient duress and that high-quality charter schools will emerge under the right conditions.

The path forward requires that choice advocates overcome the legacy of their inflated expectations and promises. The insistence that school choice simply "works" helped put a saleable, amiable face on the tough medicine that champions of school reforms sought to deliver — but often at the cost of silencing discussion about how to make choice-based reform work well. In fact, to even question the claim that "choice works" has frequently been deemed a betrayal by choice advocates; this has left the field to a coterie of enthusiasts eager to talk about moral urgency, but disinclined to address incentives or market dynamics.

On one level, the benefits of such smiley-face advocacy are plain to see. One need only look at the raft of strong-willed, pro-charter-school Democrats — figures like New York City schools chancellor Joel Klein or Colorado state senator Michael Johnston — to see how the choice mantra has helped to broaden and deepen the support for transformative change. It is also true that there has not been a major pull-back in any place where choice has gained a foothold. Outside of the Obama administration's move to end the D.C. voucher program — a change imposed on the school district from the outside — nowhere have charter schooling, school-voucher programs, or tuition tax credits been implemented and then lost favor.

At the same time, however, there has been little attention paid to the innate limitations of the "social justice" case for choice, even as a political strategy. For one thing, this approach immediately signals to the three-fourths of American parents whose children are not enrolled in inner-city schools that this debate is not about them. And given that only about one household in five even contains school-age children, choice proponents are pushing an agenda sure to seem disconnected from or even threatening to the vast majority of Americans.

Like the architects of the Great Society nearly half a century ago, choice advocates have an unfortunate habit of dismissing or denigrating middle-class voters who do not share their moral zeal. They ignore the genuine, practical worry that choice-based measures may adversely affect the property values of suburbanites who paid a premium to purchase homes in good districts or school zones, and the concerns of these home owners that their children may find themselves crowded out of popular schools.

Perhaps not coincidentally, in roughly two dozen referenda across the country over the past few decades, voucher advocates have yet to record a single win. In fact, the annual poll in the Hoover Institution's choice-friendly journal Education Next has shown that popular support for vouchers declined by a third, from 45% to 31%, between 2007 and 2010.

Proponents can (and do) rightfully place much of the blame for this track record on ferocious opposition from teachers' unions, but they have also blithely ignored basic political reality and prudence. They need to stop hectoring suburbanites, ease up on the moral indignation, and start promoting reform that will credibly improve the quality and cost effectiveness of American education for more than a small slice of households.

If advocates of market-oriented school reform accept this diagnosis, they can take a number of steps to improve their practical and political prospects.

First, they should get serious about markets as a way to promote cost efficiency. Given the fiscal straits school systems now face — and given that the country has just been through a monumental health-care debate that focused on the problems with third-party purchasing and the lack of incentives for consumers to think about costs — it is peculiar that the power of markets to engender price competition remains so unexplored in education. School spending entails no direct contribution from parents, and parents currently gain nothing from choosing a more cost-effective school; as a result, administrators in charter, district, and private schools have less reason to take tough steps to adopt cost-saving technologies or practices. And yet the choice agenda neglects mechanisms that could reward price-conscious parents by permitting them to save dollars for other educational expenditures (such as college or tutoring) if they chose lower-cost school options.

Second, reformers should broaden the educational-choice discussion beyond "school" choice. The narrow vocabulary of school choice made more sense 20 years ago, when online tutoring and virtual schooling were the stuff of science fiction, and when home schooling was still a curiosity. But in 2010, this language is profoundly limiting. In the health-care debate, even the most ardent single-payer enthusiasts believed that patients should be free to make a series of choices among physicians and providers of care. Yet in education, the most expansive vision of choice asks parents to decide among schools A, B, and C. This kind of choice may appeal to urban parents eager to escape awful schools; it does little, however, for suburban parents who generally like their schools but would like to take advantage of customized or higher-quality math or foreign-language instruction. A promising solution would be to permit families to redirect a portion of the dollars spent on their children through the educational equivalent of a health savings account. Such a mechanism would help families address children's unmet needs (such as extra tutoring in difficult subjects, or advanced instruction in areas of particular aptitude); it would also allow niche providers to emerge, would foster price competition for particular services, and would make educational choice relevant to many more families.

Third, champions of market-based reform should stop downplaying the role of for-profit educators. The Obama administration has been particularly guilty on this count, enthusiastically championing charter-school expansion even as its Department of Education radiates hostility toward for-profits in K-12 and higher education. The result is entrenched funding arrangements, policies, and political currents that stifle for-profit operators — organizations such as National Heritage Academies, which operates 67 charter schools in eight states, or EdisonLearning, which operates schools and provides supplemental education services across the United States and overseas. If choice-based reform is to yield more than boutique solutions, for-profits are a critical piece of the puzzle.

Consider that it has taken the celebrated KIPP program — an organization lauded for its aggressive expansion — 16 years to grow to 99 schools serving fewer than 27,000 students. This is longer than it took Microsoft, Subway, and Amazon to grow from start-ups to global brands. For-profits find it easier to tap private equity; they have self-interested reasons to aggressively seek cost efficiencies and to grow rapidly; and their focus on the bottom line can make them more willing to re-allocate resources when circumstances warrant a change. Of course, these same incentives can translate into corner-cutting and compromising quality; still, no one should imagine that non-profits can readily match the dexterity, capacity for rapid growth and massive scale, and aggressive cost-cutting that are hallmarks of the for-profit sector.

Fourth, reformers should foster genuine competition by arranging markets so that there are real consequences for competitive failure or success. One simple step would be to ensure that all of the dollars spent on students follow children when they change schools (the notion implicit in efforts to promote "weighted student funding" systems). Such a reform would entail stripping school districts of their hefty subsidies and of their monopolies over local school facilities. It would mean overhauling contracts and statutes that protect teacher jobs and seniority-driven pay scales — practices that leave school and district leaders without the tools needed to reward good teachers and penalize mediocrity. Real consequences for enrollment loss could help push educational leaders to start taking enrollment and parental preferences seriously when evaluating employees and doling out bonuses. And, because school districts are politically governed entities, it would enable reformers to leverage student flight — as they have in Washington, D.C. — to create the pressure and political cover that public officials need to pursue painful, but essential, reforms.

Fifth, markets are predicated on the assumption that consumers have the ability to make informed choices. It is not essential for every single consumer to have the knowledge or inclination to make savvy decisions — but providers do need to expect that the quality of their performance will be known, and will matter. Today, unfortunately, it is enormously difficult for parents in most communities to get useful information on school quality. Simple test scores generally tell parents at least as much about the students attending the school in question as they do about the quality of instruction. Reliable measures of how much students learn during a year (i.e., the school's "value added") are infinitely more useful, but they are as yet available in only a handful of places for a limited number of schools, grades, and subjects. Similarly, it is difficult for parents to find comparable or trustworthy data on school safety, arts instruction, programs for high achievers, or the fate of former students. There is a gaping need for third parties to step up and play the role of a Zagat's guide or Consumer Reports, providing accessible, independent information on K-12 schools. As these examples make clear, there is absolutely value in having multiple providers, perhaps focusing on different educational concerns or kinds of schools. This area presents a vast opportunity for philanthropists or civic-minded enterprises, especially as promising but primitive information-distribution efforts already exist in cities like New Orleans, Milwaukee, and New Haven, Connecticut.

Finally, reformers should recognize that dynamic markets require vibrant entrepreneurial ecosystems. What has made Silicon Valley a locus of entrepreneurship is not that it has a "freer" marketplace than other American cities, but that it has attracted over decades the investors, researchers, and networked expertise necessary to develop and sustain high-quality ventures. Experience has made clear that such ecosystems don't necessarily spring into being unbidden, and that they sometimes need to be consciously cultivated. Even in choice hotbeds like Milwaukee and Washington, we still do not see many growth-oriented providers or savvy investors screening potential new entrants and nurturing those with the most promise. Meanwhile, too little is being done to help new education providers find facilities, negotiate political obstacles, or leverage labor-saving technologies. Ventures like New Schools for New Orleans and The Mind Trust in Indianapolis represent pioneering efforts to clear bureaucratic obstacles, attract talent, and cultivate networks. Such efforts are multiplying across the land, spurred by supporters like the Gates Foundation and the NewSchools Venture Fund, and aided by federal policies like the Race to the Top program. These are promising developments — and they deserve more attention and care from reformers.

MAKING MARKETS

It would seem, then, that school choice "works" in some respects and in some instances — but that choice alone could never work as well as many of its champions have expected, and promised. It is time for those who would like to transform America's schools to let go of the dream that choice by itself is any kind of "solution." The goal ought to be a much more serious agenda of school deregulation and re-invention.

Choice advocates still routinely invoke iconic market thinkers, particularly Milton Friedman, in asserting that "school choice works." It might be time for them to take another look at their Friedman, and their Friedrich Hayek. It was Friedman who admonished that the market "is not a cow to be milked." And it was Hayek who, in collecting his Nobel Prize, encouraged policymakers to think of themselves as gardeners — creating the conditions in which enterprise could flourish. Neither Friedman nor Hayek believed that markets were self-sustaining or failsafe. Their approach to market-based reform was not the enthusiastic cheerleading of the choice movement; it was a far sterner, grittier charge. And as school choice now enters its third decade, its champions would be wise to take the counsel of Friedman and Hayek to heart.

Frederick M. Hess is director of educational-policy studies at the American Enterprise Institute. His next book, The Same Thing Over and Over, will be published in November.