- school-choice policies have failed to create a vibrant marketplace for private educational options
- proponents of school choice need to look beyond public policy to the significant part that society must play in order for school choice to succeed
- any successful school-choice program will rely on the capacity of society to generate viable alternatives to what the gov does now
Conservatives have long championed school choice as a way to improve America's education system. Ronald Reagan called for vouchers in his first term as president, and many other Republican leaders have taken up the mantle since then, with choice — both of private and charter schools — becoming a key plank of the party's education platform. Speaker of the House John Boehner has argued, "When parents have the ability to select the best learning environment for their kids, they thrive and so do their communities." Senator Ted Cruz has even gone so far as to describe school choice as "the civil rights issue of the 21st century."
These calls for greater school choice have not gone unheeded. Since 2000, the total enrollment in voucher, tuition-tax-credit, and education-savings-account programs has grown from just under 30,000 students (with a large proportion clustered in Milwaukee and Cleveland) to over 300,000 students all across the country today. In 2013, there were 49 total private school-choice programs in 24 states, up from only 12 programs in 2001. In just the last half-decade, several states have taken great steps to encourage school choice. Led by then-Governor Mitch Daniels, Indiana created a statewide voucher program in 2011 that today enrolls almost 20,000 students. In Louisiana, Governor Bobby Jindal has overseen the creation of two separate voucher programs (over objections from the Obama administration) which now enroll about 7,000 students. And in Wisconsin, Governor Scott Walker has made expanding his state's existing voucher program a priority, leading to the program's enrollment rising by almost 30%, from just under 21,000 students to over 27,000 students during his time in office.
These school-choice policies promise to give students access to the education that best fits their needs. Primary and secondary education in America has traditionally been monopolized by the government. If parents did not have the means to send their children to a private school, their children had no choice but to attend their local public school, even if the school offered a poor education or did not meet their specific needs. This system led to massive variation in quality across the country and left many students trapped in failing schools. School-choice programs offer these students a way out by giving them money to attend another school, often a private one, that they can choose for themselves. Policies promoting school choice seek to create an education marketplace that has the potential, as Milton Friedman put it, to "change the character of education."
And yet, despite the potential of these programs, private school choice has not led to the sea change in educational quality that many of its early backers predicted. Research on choice programs tends to find small positive gains for both students who participate in school-choice programs and students who are left behind in public schools, but the improvements are not anywhere near what choice advocates had hoped. School choice has just not lived up to the hype.
Why is this the case? In theory, school-choice programs can cause three things to occur in education systems. They can fill empty seats in existing high-quality private schools, encourage existing private schools to grow to serve more students, or drive the creation of new high-quality schools. But a gap has opened up between the theory of school choice and policymakers' ability to implement it. School-choice programs have been great at filling in existing space in private schools, but they have not helped these schools expand or encouraged the creation of new schools. As a result, there has been a stubborn upper bound on the amount of change any of these programs can effect in the American education system. In short, school-choice policies have failed to create a vibrant marketplace for private educational options, so the idea's potential to dramatically revolutionize K-12 schooling has so far gone unfulfilled.
It seems unlikely that Republicans are going to abandon their commitment to private school choice, and rightly so. The moral and conceptual cases are strong. But if the actual outcomes from school-choice programs fail to live up to the rhetoric or the theory, the disappointment risks derailing the pursuit of greater educational opportunities for millions of American children. Rather than declaring that choice is the surest way to break the cycle of poverty, policymakers serious about making school choice work should instead set about creating the conditions necessary for a thriving marketplace to develop.
Proponents of school choice face several challenges to expanding access to good private schools. Choice programs too often do not account for either the capital that schools require to expand or the extra expenses private schools face in educating certain students; moreover, the traditional models of funding school choice do not reflect the real cost of educating students in general. Further, school-choice policies have led some to call for greater regulation of private schools, which could stifle their ability to innovate. And proponents of school choice need to look beyond public policy to the significant part that society must play in order for school choice to succeed. To strengthen school-choice programs, each of these challenges should be addressed.
STRUGGLING TO EXPAND
One of the primary reasons school-choice policies have floundered is that private schools do not have the physical capacity to accommodate the many new students with access to public funding, and the dollars that private schools receive from the state are simply not enough to finance expansions.
School-choice programs tend to fund students in one of two ways. Under most voucher models, students receive some percentage of the amount the state would have contributed to the public school the student was slated to attend. In the Indiana Choice Scholarship Program, for example, students with household incomes up to a certain level can receive 90% of the state's per-pupil spending (usually around $5,000) to put toward private-school tuition, while the students' public schools keep the funding from local property taxes and the federal government. The second mechanism, tuition tax credits, encourages private support of students' tuition by giving a tax credit to individuals or businesses that contribute to private scholarship-granting organizations. These organizations then give students voucher-like scholarships to attend private schools. In some states, the amount each student can receive through the scholarship is capped. In Florida, for example, a student could receive only $4,880 for the 2013-2014 school year through the tax-credit program.
Both of these funding mechanisms are designed to cover the marginal cost of educating an additional student at an existing school, but not the cost of accommodating all of the students who want to attend the school. If a school wanted a new building to welcome 100 new students with vouchers, for example, or a new campus to welcome 1,000 new students, the funding to build the space to seat these students would not come with the vouchers. Instead, the private school would have to raise more money from non-public sources to accommodate the increased demand. As a result, vouchers and scholarships do not help private schools take in more students than their current capacities permit.
Public schools do not face the same capital barriers to expansion. School districts all across the country use bonds to raise start-up funds for new schools or to finance major capital expenses. For instance, the Miami-Dade school district in 2012 asked voters for a $1.2 billion bond measure to fund the renovation of old school buildings and the construction of new ones. Beyond that, public schools and many charter schools have access to facilities funding from the government. According to Census Bureau reports, of the $600 billion spent by local, state, and federal governments on K-12 education each year, a full $52 billion — nearly 9% — was spent on capital outlays. This funding comes out of separate budgetary line items and is not included in most published calculations of per-pupil expenditures, giving public schools a distinct advantage over private schools that most school-choice programs do not account for.
Because private schools typically cannot leverage public bonds for capital expenses, they are left with two options for financing new buildings or renovations: fundraising or loans. Both options are problematic for schools serving low-income students or subsisting off of vouchers or tax credits. Schools from historically poor areas simply don't have a donor base from which to raise money. And private schools that rely on politically tenuous voucher programs for funding — money that could end with the next election cycle — are not in a strong position to receive favorable interest rates from banks that want to ensure they will get their money back.
One possible solution to this funding imbalance is simply to boost the amount of funding for voucher programs or allow private schools to access some chunk of the facilities monies made available to public schools. This approach might alleviate the problem in the short term, but it would likely be unpalatable politically. The leaders of traditional public schools argue that even if students take some money to private schools through vouchers, the public schools still have to pay capital costs for buildings and certain fixed costs for janitorial staff, administration, and support services. The supporters of public schools would almost certainly resist further cuts into public schools' budgets. Given the contention that already surrounds voucher programs in state legislatures, a move to bolster vouchers with facilities funds would likely not pass, even in Republican-controlled legislatures.
One state has offered another way for private schools to receive public aid for large capital-improvement projects. More than 30 years ago, Colorado created the Colorado Educational and Cultural Facilities Authority, a body with the ability to provide tax-exempt bond financing for a variety of educational and cultural non-profit organizations, both inside and outside of Colorado. Over the years it has offered more than $5.5 billion in bonds to educational and cultural institutions, including the National Jewish Federation Bond Program and the Association of Christian Schools International. In 2012, the Catholic Educational Capital Corporation accessed $9 million in funds from the CECFA and offered them to Iona Prep, an all-boys high school operated by the Christian Brothers in New Rochelle, New York, to purchase an elementary school. According to the National Catholic Register, over the lifetime of the bonds the school is expected to pay 2.09% in interest, instead of a standard rate of more than 5%. This favorable rate stands to save the school $1.8 million.
CECFA is not a state agency and does not receive any funding from the state of Colorado. It simply has the imprimatur of the state to offer tax-free capital financing, provided that the efforts it funds fall into the realm of arts, culture, and education. It is regulated by a board of directors who are appointed by the governor and confirmed by the Colorado Senate. It finances itself through loan fees and sells the bonds through investment banks, just like any other bond issuer. While CECFA can sell bonds for organizations outside of Colorado, provided that they have some Colorado connection, other states could set up organizations that fund projects only within their own borders. Such organizations could channel some state aid to private schools for large-scale projects that could help them accommodate more students with vouchers.
"Social-impact bonds" give states another option to help private schools finance large projects. These bonds give private investors an opportunity to provide the start-up capital for a program that, if successful, will ultimately provide a social good at a lower cost than what the government would pay. If the program is successful, the investors not only get their money back, but they are also rewarded with a share of the government's savings. The more successful the program is, the higher the return. Social-impact bonds have been used to fight recidivism, homelessness, and unemployment, and they could be used in education as well.
The government savings from education social-impact bonds would come not just from the reduced cost to the state posed by private schools. Patrick Wolf of the University of Arkansas and I have calculated that the public saves about $90,000 for each student that graduates from high school, since high-school graduates pay more in taxes, commit fewer crimes, and rely less on social services than high-school dropouts. If a private school gets more students to graduate from high school than the local public schools do, the government and taxpayers will ultimately save money. A social-impact bond could allow investors to share in that savings by financing a private school with the goal of getting more students through to graduation.
There are lots of particular questions and difficulties to work out before applying such a model, of course. For one, what if one state finances a student's high-school graduation but the student moves, gifting the increases in revenue to another state? And what if the value of a high-school diploma drops precipitously? State and district leaders would have to consider these possibilities, but the underlying pay-for-performance logic could marshal private dollars for the public good and allow private schools to expand enough to accept new voucher-funded students.
AVOIDING ADVERSE SELECTION
Another problem with the funding structure of school-choice programs is that they give private schools only the money needed to educate the average student — not the funds to cover more expensive students. By funding private schools in generic per-student lump sums, these programs encourage private schools to accept only those students who are cheapest to educate. The resulting adverse selection undermines the central premise of school-choice programs: that students should be able to seek out the schools that best fit their unique needs and should not be limited by either geography or cost.
Milwaukee's voucher program illustrates the risks here. Local public school districts in Wisconsin are eligible for 26 different "categorical aid" programs that provide supplementary state funding for everything from bilingual education to gifted-and-talented programs to educating students with special needs. On top of these funds, the state provides aid for low-income districts as a part of state "equalization" funds. As a result, the Milwaukee Public Schools spend an average of over $14,000 per student, with students with special needs receiving even more. By contrast, through the Milwaukee Parental Choice Program, each student received only $6,442 for the 2013-2014 school year, regardless of the true cost of educating him.
This funding disparity depressed the number of more-expensive students who attended private schools with vouchers. Political scientists Patrick Wolf, David Fleming, and John Witte found that only 7.5% to 14.6% of students (depending on estimation strategy) participating in the Milwaukee Parental Choice Program were identified as having special needs, compared to 19% of students in traditional public schools.
A 2010 evaluation of the D.C. Opportunity Scholarship Program bore out similar results. Voucher-lottery winners were less likely than voucher-lottery losers to attend schools with special services for English-language learners (32% versus 57%). Students with special needs using the scholarship program similarly attended schools with programs for them at a lower rate than students who lost the lottery (75% versus 90%). Money again was likely a contributing factor as the scholarship was capped at $7,500, compared to the over $16,000 spent per pupil by District of Columbia Public Schools at the time of the study. (The scholarship amount has since been increased to over $12,000 for high-school students, but this still falls far short of the more than $18,000 that DCPS now spends per pupil.) The private schools just didn't receive enough money to educate more expensive special-needs students at the same rate as public schools.
The popular policy prescription to solve this problem is to create blind lottery systems that require schools to take any student that applies. But this solution is bad for both students and schools. What makes private schools unique is the variety of particular approaches to education they offer. Prohibiting them from setting any kind of restrictions on whom they admit, from religious preferences for religious schools to behavioral expectations for "no excuses" schools, could erode the educational programs they are attempting to implement. This approach also fails to provide any help for students who are legitimately more expensive to educate, since the schools do not receive any extra money.
A more sensible solution is to bump up voucher or scholarship amounts commensurate with the increased cost of educating particular students. If, for example, states offer public schools an extra $1,000 for English-language learners, vouchers should include the same. If public schools get extra funds for low-income students, students with special needs, or gifted students, those dollars should follow the students into the schools they attend. This would provide an incentive to serve harder-to-educate students, because the amount of funding requisite to their academic needs would come with them.
Several school-choice programs around the country are specifically designed for students with special needs, and the vast majority of dollars that would follow those students into public schools go to the private schools instead. In Florida, for example, in the 2013-2014 school year, 27,772 students in 1,240 schools participated in the McKay Scholarship program for students with disabilities. And while other school-choice programs are stagnating, Florida's program for special-needs students is flourishing: The program has grown by over 65% in the last eight years.
The Florida program's success shows that, with the proper funding, private schools will serve harder-to-educate students. A voucher or scholarship program that gives each student the true amount necessary to educate him would help all students, not just average ones, have a real choice between public and private schools.
LEANING ON CIVIL SOCIETY
More robust and tailored funding mechanisms would help give private schools the financial means to expand and accommodate more students, but these changes would not help these private schools expand and improve in other less tangible — but equally important — ways.
One of the great strengths of private-school programs is that they leverage the diverse and unique interests, passions, and dispositions of willing individuals throughout the country to defray some of the cost of delivering instruction. Many teachers at religious schools, for example, accept a lower salary than their public-school counterparts receive because they are committed to the school's unique mission and the religious culture matches their own values. Likewise, such schools are subsidized by the religious organizations that sponsor them.
When looking to improve school-choice programs, then, public policy cannot be the only focus. Equally important is enlisting the contributions of religious organizations, philanthropists, and individual citizens to build up and strengthen those schools.
Private citizens have invested in higher education in this country's past. After the Civil War, when titans of American industry found to their frustration that the ministerial model of education delivered by the Ivy League was not preparing the scientists and engineers necessary to build a modern economy, they simply founded their own universities. Their namesakes — Carnegie Mellon, Johns Hopkins, Stanford, and others — have become some of the most elite institutions in the world.
The same thing could happen in K-12 education. It takes time and money to develop a plan for a school, find a staff, secure a building, and get a new school off the ground. Most private schools lack all of the above. When it comes to selecting staff, in particular, they also have a uniquely challenging task. Teaching and leading in a private school is different from teaching and leading in a traditional public school. For that matter, teaching and leading in a private school in the kind of vibrant marketplace created by a robust school-choice program is different from running either a traditional public or traditional private school. Unfortunately, there are very few pipelines for developing future school leaders and administrators with the mix of legal, financial, marketing, and instructional acumen to create a successful choice-market school, and very few programs preparing teachers to teach in one.
Clearing these hurdles starts with people. One solution could be integrated pipelines from universities to schools. Catholic universities can partner with dioceses to identify their needs and train teachers and leaders for them. Lutheran universities could partner with synods. Evangelical universities could partner with affiliated schools.
This kind of partnership is already happening on a small scale in several locations. Loyola Marymount University has four teacher- and leader-training programs that have worked with the Archdiocese of Los Angeles to place over 1,500 teachers and leaders in schools serving low-income populations since 2001. The University Consortium for Catholic Education, anchored in the Alliance for Catholic Education program at the University of Notre Dame, is a collection of 13 Catholic universities that places over 400 teachers in Catholic schools in 24 states each year. But given the fact that there are almost 6,600 Catholic schools across the country, this is still a very small endeavor. The Wisconsin Evangelical Lutheran Synod has set up a pipeline from its two affiliated universities (Martin Luther in New Ulm, Minnesota, and Wisconsin Lutheran College in Milwaukee) and several other local universities directly into schools that participate in the Milwaukee Parental Choice Program. But again, this program is relatively small when compared to the staffing problems these schools face.
Beyond ramping up existing programs, private philanthropies could fund the creation of entirely new schools of education for private-school teachers. Teacher-preparation programs are widely derided as having low entry requirements, a lack of intellectual rigor, and inadequate clinical experience for teachers. Secretary of Education Arne Duncan said himself, "By almost any standard, many if not most of the nation's 1,450 schools, colleges, and departments of education are doing a mediocre job of preparing teachers for the realities of the 21st century classroom." The many philanthropists interested in education reform would do well to take note and consider investments in new schools of education.
The charter sector offers examples of such institutions. In 2007, a prominent charter-school operator in California, High Tech High, created its own graduate school to prepare teachers to teach in their unique, technology-driven environment. In 2008, leaders of several high-performing "no excuses" charter schools in New York launched Teacher U, which became the Relay Graduate School of Education, the first independent, non-profit graduate school of education to be credentialed by the state of New York in over 80 years. In 2012, MATCH charter schools founded, and the Massachusetts Board of Higher Education approved, the Sposato Graduate School of Education for teachers in their system of schools.
Many regional universities train large numbers of teachers. Perhaps a philanthropist could fund the creation of a private-school preparation track within these schools. Other universities that do not have schools of education at all could either create schools of education or create teacher-preparation programs in colleges of arts and sciences, professional schools, or business colleges. Rice University created the Rice Education Entrepreneurship Program, which trains aspiring school principals and other leaders in management, accounting, and finance through the Rice University business school. In fact, REEP is the first principal-prep program in the country to be housed in a business school. Still other non-traditional programs could be created, especially in places where private-school teachers do not need to be certified by the state.
Policymakers are not helpless to encourage private investment in private schools. The tax deduction for charitable giving is a good example of a policy that strengthens civil society by encouraging donations to eligible non-profit organizations (like most private schools). Other policies, like social-impact bonds, could further encourage private investment in private schools themselves and in other programs meant to support them. But the fact remains that the lead on private-school innovation and expansion will have to come from the private sector, and policymakers have ample reason to believe that society will, left with enough space, create new schools and innovative ways to educate. The biggest challenge for policymakers might not be crafting new policies to encourage private activity but simply staying out of society's way.
More sensible funding mechanisms and greater involvement from civil society would help expand the capacity of private schools, but they would not ensure that the schools are performing adequately. A virtue of private schools is that they are relatively free of confining regulations because they do not rely on the state for funding, and as a result they have much more flexibility to meet the needs of their students. But the influx of public funds to private schools through voucher and scholarship programs has undermined private schools' independence, as state leaders respond to political pressure to closely monitor participating schools. Taxpayers have a reasonable concern when their dollars go toward sending students to inferior private schools, and the state has a right to demand that these dollars are used well.
A proper regulatory framework can mitigate the risk that the state is sending students to underperforming private schools, but the structure of such regulation is hotly contested. The same regulations that control public schools likely will not work for private schools since the two are so different. For example, private schools in many states are not required to have state-certified teachers. This freedom allows them to pull teachers from a greater range of areas and with a greater range of experiences. What's more, private schools have an additional force of accountability acting on them: market pressure. Clearly an alternative regulatory structure is necessary.
One approach has been to require students at private schools receiving public dollars to take the same tests and be held accountable by the same grading programs as traditional public schools. In Wisconsin, Indiana, and Louisiana, the states with the three most expansive voucher programs, students are required to take the state end-of-course exams, and schools that score poorly risk losing their ability to admit students with vouchers. This approach has the virtue of regulating outcomes and not methods, leaving school leaders in charge of how they will achieve the necessary testing results. Nevertheless, basing regulations on testing does restrict the ability of schools to innovate in big ways that might disrupt the traditional teaching model. And testing does dictate curriculum choices to a certain degree.
Another way to regulate private schools in a voucher system is to treat participating private schools more like charter schools, which are authorized to receive public dollars by independent agencies granted that power by the state. Andy Smarick, a former senior official at the New Jersey Department of Education, pioneered this idea in his 2012 book, The Urban School System of the Future, arguing that chartering is a way to allow organizations to keep their uniqueness but still be accountable for performance. Smarick contends that each state should have a small number of officials whose sole work is authorizing private schools.
This idea could be expanded further by using "performance contracts." If a private school wanted to receive public dollars, it could apply to an authorizer, either an independent state board or a non-profit or university that the state permits to charter schools. The school and the authorizer would then negotiate the evaluation metrics for a given period of time. Designated performance goals would be codified in the charter of the school. After three to five years, the authorizer would look at the school's performance on these metrics and see if it is in alignment with what was promised in the charter. If not, the school would enter some kind of corrective-action plan and risk losing access to public dollars.
The negotiation of the metrics would be the key to making this system work. Today, the leaders of many private schools complain that they are held to irrelevant metrics by the state. This contract model would turn this complaint on its head, as schools would be held to the standards that they themselves have negotiated and agreed to. Schools would get a say in how they are going to be evaluated, but they would then be strictly held to those goals.
Given the diversity of approaches that has sprouted up in the charter sector — from the classical-focused Great Hearts schools in Arizona, to technology-heavy hybrid schools like Carpe Diem and Rocketship, to arts-focused schools like the Benton County School of the Arts in Arkansas, to STEM-focused schools like the Dove Academies, to fully online schools like the Pennsylvania Cyber Charter — it's hard to argue that this approach presents a real threat to educational diversity. If applied to private schools, it could give them the flexibility they need but the accountability that the state demands.
All of these ideas — better voucher-funding mechanisms, more tailored regulations, and greater investment from civil society — tend toward one goal: the decentralization of control over education and the proliferation of education options for students. The fundamental premise of school-choice programs, after all, is that different students have different needs that can be best met in different educational arrangements, and centrally regulated public schools are too often ineffective in providing those different environments. Private schools are thus a great alternative, and the state should encourage them.
One final policy reform promises to push this decentralization even further. An education savings account is a restricted use, flexible-spending account for qualifying students into which the government might put a set amount of money each year and that parents would then be free to spend on qualifying educational expenses. ESAs have been implemented in Arizona, and they are under debate in other states, including Mississippi, Oklahoma, and Iowa. In Arizona, for non-special-needs students, the state provides around $5,300 per year. Qualifying expenses include private-school tuition, textbooks, online coursework, standardized tests (like AP exams, the ACT, or the SAT), tutoring, and supplementary therapy services for students with special needs. Parents can mix and match education services, deciding how best to apportion the money. Unused funds can roll over from year to year and can even be put in a tax-advantaged 529 college savings account.
ESAs present a viable solution to a fundamental challenge in the way that education is delivered in America. Rather than going to one location run by one organization for an entire day, parents can find the combination of programs and providers that best meets their child's needs. Perhaps they love their school but want their child to learn Mandarin; with money from their ESA, they can purchase an online program to teach the language. Perhaps they are in a rural area without access to AP or dual-enrollment courses; they can use the funds to pay for credit-granting online courses provided by universities.
Another advantage of ESAs is that they hold the promise of harnessing one of the newest innovations in education. Many of the country's top colleges and universities are offering Massive Open Online Courses, and many of these MOOCs are free to attend and require only an inexpensive professionally administered exam to grant credit. Higher-education advocates and entrepreneurs are trying to figure out the proper place for MOOCs in colleges and universities, but perhaps their most promising use isn't in higher education but in secondary education. High-school students could use their ESA dollars to access these credit-bearing college courses, regardless of where they live.
ESAs also address another big problem with education in this country: We don't know how much it costs to educate a student. We know how much we spend, but because public schools are funded in lump sums to districts — and even voucher, tax-credit, and charter-school allotments are lump-sum "coupons" to purchase education — we see schools set their costs right at the government subsidy. As competition for state budget dollars increases with expanding Medicaid liabilities and looming pension obligations, cost-effective education becomes more important every year.
"Unbundling" education services through a flexible-spending-account model encourages all providers to compete on price. If parents know that whatever they do not spend on tuition they can spend on tutoring or save for college, they will evaluate schools based on the return on their investment. To date, private schools that participate in school-choice programs are evaluated on one dimension: quality. With fungible dollars, these schools can be evaluated two-dimensionally: on quality and price. This competition should serve to both increase quality and decrease price.
ESAs, especially in tandem with other reforms, thus could help to spur a strong education market where students and their parents can shop for the best education options. If a student finds a traditional public school to be the best option, then he can go there, but if a private school or a mix of programs provides the best education for a student, then he can receive the benefits of those options instead. ESAs are thus a logical extension of the decentralizing trend that school-choice programs are advancing.
THE CHALLENGE OF SCHOOL-CHOICE POLICY
The exponential growth in enrollment in school-choice plans over the last decade is laudable. More families than ever have access to the education programs of their choosing. But the current total of 300,000 students looks pretty small in comparison to the 55 million schoolchildren in America. If we want the marketplace to expand — and for the winnowing forces of markets to work to increase the quality and diversity of educational programs and products while reducing their costs — much more work needs to be put into creating a functional, vibrant, flourishing education marketplace.
Policymakers must approach school choice with a balance of ambition and humility. The goal of developing an education market is indeed ambitious: School choice marks a radical departure from the traditional method of providing our children an education in America, and it presents a possible way to transform the education children receive in many of our poorest and most despondent communities. But at the same time, school-choice policies are trying to harness a force that is, by its nature, beyond any policymaker's ability to control. Any successful school-choice program will rely not on the good intentions of government employees but on the capacity of society to generate viable alternatives to what the government does now. This fact means that, for a policy of school choice to succeed, policymakers will have to step back and relinquish control.
Americans have always had a strong tendency to form associations to meet their common needs, and today there is no better example of civic associations in America than private schools. Education policies that harness this constructive propensity of American society through a wise mix of incentives, regulations, and freedom are likely the best way to improve our education system. School-choice policies have started to grasp this fact, but they have not done nearly enough, and our education system has therefore not been changed nearly as much as many people hoped. We can, and must, do more.