March of the Vouchers - What Should the Left Learn from School Choice Debates?
— Harry Brighouse
IN APRIL, FLORIDA became the first state to adopt a statewide school voucher plan. By a vote of 25-15 the State Senate adopted the absurdly named “A+ Plan for Education” which had previously been passed in the House by a vote of 70-48.
The bill establishes a mechanism for “grading” all public schools in the state, based on test scores and other factors such attendance and graduation rates. Schools that earn A's or show improvement over the years will get incentive rewards of $100 per pupil from the state. But students in schools identified as “failing” for two out of four years will be eligible for $4000 vouchers, which can be used to send them to the public or private schools of their parents' choice.
While less than a handful of schools will qualify as "failing" this fall, it is expected that as many as 169 schools with 156,000 students will be eligible in the 2000-2001 school year. Florida's plan is not the first to involve widespread use of vouchers to send children to private schools, but it is the first statewide plan, and represents a massive legislative victory for the voucher movement.
The two big voucher plans in existence are concentrated in the inner cities of Cleveland and Milwaukee, and are carefully targeted at children from low-income families. Both of them, like the Florida plan, allow the State to pay for children to attend religious, as well as secular, private schools. (The Milwaukee plan was modified in 1995 to allow children to attend religious schools, but the reform was held up in the courts until the 1998-1999 school year).
The only high-profile attempt to win a statewide voucher plan was the disastrous campaign for Proposition 174 in California in 1993, which would have allowed all children to attend private schools with a voucher of only $2,500 and imposed almost no regulation on participating schools. It was defeated by a massive majority and opposed by both political parties as well as the Republican governor, Pete Wilson.
Since the California vote, voucher proponents have gained strength by touting the virtues of the Milwaukee scheme. The Florida success, in tandem with the refusal of the U.S. Supreme Court to hear a case against the Milwaukee scheme last fall, can be expected to give them a great boost. Several other states are considering voucher legislation, and California is likely to face a new state-wide ballot proposition in 2000.
Vouchers are only the most radical of a cluster of reform proposals going under the general rubric of “school choice.”They represent a sharp break with the traditional model of government provision of government-funded schools. But other quasi-market mechanisms have enjoyed more success. Several states have choice schemes within the public schools.
In part, public school choice developed in response to judicial pressure for racial integration: Magnet schools in particular were often developed as a responses to, or to pre-empt, integrationist court orders. Interdistrict transfer programs allowing children from inner-city school districts to choose places in suburban districts have allowed suburban districts to resist pressures to merge with the inner-city districts.
Other public school choice schemes were devised to facilitate parental involvement in the public schools and to improve already integrated facilities. Most famously, East Harlem District No. 4 has twenty-three high schools organized around different themes, application to all of which is available to all parents in the district. Several states have also adopted open-enrollment rules, which allow parents to send their children to schools in different school districts, and force the home district to reimburse the chosen district the cost of schooling the incoming child.
In all, thirty-four states now take advantage of federal support and enabling legislation to establish charter schools. These are usually run by educational entrepreneurs in consultation with local school districts, but are exempt from much of the regulatory burden imposed by the State on public schools. The idea is that educational innovation can be fostered in a less constrained framework, and the schools are funded by and accountable to the State.
Parents may elect to send their children to these charter schools, which may not discriminate except on specified grounds. By 1998 more than 250,000 children attended over 1,000 charter schools. Because they are, in some sense, bipartisan—leaders of the National Education Association, and of the Democratic Party, support charter schools—charter schools represent the form of choice likely to be most extended in the near future.
Vouchers and school choice have an peculiar history. Vouchers were first argued for by libertarian economist Milton Friedman in 1955, and the arguments he made then still pervade the documents of school choice proponents. But the idea was taken up, briefly, by liberals in the Department of Education in the late 1960s as a mechanism for integration. One voucher scheme was instituted under the auspices of the federal government in Alum Rock, California in the 1970s, but lasted only two years and was quickly abandoned for lack of local enthusiasm.
In the early 1980s school choice became an organizing idea in the right wing of the Republican party. Because it fuses the Christian and conservative right's “family values” agenda with the libertarian right's obsession with markets, school choice was a perfect tool for building the New Right coalition. Even without any apparent prospects of legislative success, the idea had an major impact on the shape of politics.
In adopting the school choice mantra the right has tapped into deep disaffection with the most important public service delivered in the United States, and in particular has been able to create a cleavage between the traditional civil rights organizations and many Black voters. Opinion polls consistently show majority support for public school choice, and as early as 1986, a Gallup poll showed that 54% of non-whites supported publicly funded private school choice.
Gallup polls in 1993, 1995, 1996 and 1997, asking “Do you favor or oppose allowing students and parents to choose a private school to attend at public expense?” got a positive response of 24%, 33%, 36% and 44% respectively. When the word “government” was substituted for “public” in 1997, the positive response rose to 48%. In 1998 Gallup showed 51% support for the State paying “all or part of the tuition” in a proposal allowing “parents to send their school-age children to any public, private, or church-related school they choose,” with support among non-whites at 68%.
It should be no surprise that vouchers are more popular among the poor and non-whites than in wealthier groups. Whereas in most other advanced capitalist economies schools receive roughly equal per-pupil funding—so that while the schools do nothing to mitigate background inequalities, they do not powerfully exacerbate them—in the United States, uniquely, the government singles out children from wealthy communities for better treatment in the schools.
The fetish of “local control” has produced some 20,000 school districts, many of which manage a single K-12 school. Local control is extremely inefficient, with the duplication of many district-level functions and loss of economies of scale. It also produces local funding, so that wealthy communities can and do spend two or three times as much on their schools as do poorer communities.
The more radical variants of school choice could not have succeeded but for the dissatisfaction of poor and African-American communities with their public schools. A loose and uncertain right-left coalition between right-wing funding foundations and the so-called Christian right on one side, and Catholic dioceses and urban African-American activists and politicians on the other, has pushed these schemes forward.
The support of African-American politicians in particular has been essential for giving credence to the claims of market ideologues that the poor will benefit most from such schemes. The Milwaukee Public Choice Program (MPCP), developed by the Republican Governor Tommy Thompson in collaboration with African-American state legislator Annette (Polly) Williams, is a prime example.
Williams headed up Jesse Jackson's 1984 presidential campaign in Wisconsin, and was still known as a left-wing legislator when she started advocating vouchers in the late `80s. With a few other African-American luminaries and the support of the right-wing Bradley Foundation, she made vouchers respectable and legislative victory possible.
Advocating choice also gives the right a powerful rhetorical advantage. It is impossible to oppose school choice as anti-egalitarian without suggesting that poor parents are less capable than wealthy parents of making good choices for their children. Indeed school choice has always been there for wealthy parents: They can exercise school choice by choosing to live in the best school districts, the neighborhoods served by the best schools in those districts or, to the limit, by buying private schooling for their own children.
It is this existence of choice for the wealthy that gives the extension of school choice to all its egalitarian bite. The stance the left should take on toward school choice is therefore complicated.
Most opposition to voucher schemes focuses on the use of public money to fund religious schools. But this is an artifice of the legal system in the United States. If the State-church separation case can be made successfully in the judicial system, then left activists can avoid confronting vouchers politically. They can also avoid the issue of whether vouchers actually work to the advantage of children from disadvantaged backgrounds.
In fact, the State/church separation objection to vouchers is on weak ground. The judicial precedents do not augur well for challenges, at least at the federal level. The federal government, furthermore, already subsidizes religion to the tune of billions of dollars annually: in 1995 about 57% of the $144 billion made in tax deductible charities went to churches, which, using conservative estimates of average marginal tax rates implies indirect subsidies of around $16 billion.
Besides that, it is hard to believe that the First Amendment objection could be decisive. Any reform proposal will involve tradeoffs between important values. For the left educational equality is a vital principle, which many will think as important as the prohibition on State funding of religious schools. If being educated at carefully regulated religious schools would improve the education of poor children by, say, 100%, that would count powerfully in its favor.
If the left supports the principle that all children should receive an equal education, it should be careful not to reject moves in that direction, unless it can propose better alternatives that have a realistic chance of being implemented. Because the left cares, first and foremost, about equality in education, the facts matter. Does choice lead toward equality or not?
The answer is complex. The Milwaukee scheme is the only voucher scheme to have been subjected to rigorous scrutiny. Based on this scrutiny, one prominent advocate of choice, Paul Peterson of Harvard University's School of Government, has claimed that MPCP has led to improvements in test scores which, if they “could be achieved for all minority students nationwide it could close the gap separating white and minority test scores by somewhere between one third and one half.“
If this were true, it would be decisive in favor of voucher schemes like MPCP. Whatever the merits of the case against using test scores to determine students' futures, they are so used and will be in the foreseeable future, and it is wrong to deprive those students who are already the least advantaged of the prospects which improved test scores would bring.
It is important, then, to understand the way the MPCP works, and to understand why Peterson's claim for it is false. When it was designed, the MPCP allowed students to attend private schools in the Milwaukee Public Schools area. Vouchers were only for children from households with incomes 1.75 times the federal poverty level or below, and who had not attended a private school in the previous year. Participants were restricted to 1% of the total MPS school population (this went up to 1.5% in 1993).
Though parents were responsible for transportation costs, in some cases the MPS would assist. Schools could not charge additional fees, but could charge reasonable fees for extracurricular activities, towels, school uniforms, etc. The voucher is a check, payable to parents, which goes directly to the school quarterly and must be signed over by the parents to the school. Initially only 49% of a school's attendees could be choice students; this cap was raised to 65% starting 1994-5.
The voucher is set at the value of the state equalization aid per-pupil ($4,894 in 1997-8); the total amount sent to choice schools is deducted from the MPS equalization grant. Participating schools are subject to strict admission requirements: They must not discriminate on the basis of race, special educational needs, past academic performance, or past behavior. Oversubscribed schools must select by a random lottery.
The program was initially small students participated in 1993-4, and as late as 1997-8 there were only 1500 students in twenty-three (secular) schools—but the introduction of religious schools and lifting of the caps has increased the numbers dramatically.
In 1998-9, eighty-seven schools participated, with 6,200 students. Of this total, thirty secular schools enrolled about 2,200 students, while fifty-seven religious schools enrolled about 4,000 students. Around $28.6 million in state aid was therefore diverted from MPS to choice schools.
What is the evidence? MPCP was subject to a study led by John Witte, of the University of Wisconsin, who had access to extensive data that the state required participating private schools to gather. Although major changes to the program were made in 1995 (fully implemented only in 1998-9), these changes do not affect the findings of the studies: One of the changes was the elimination of funding for the collection of evidence and data. The most important change was the extension of the program to include religious schools (although these schools are not permitted to require that Choice children participate in any form of religious service if the children's parents object).
Witte's team used scores on the Iowa Tests of Basic Skills to evaluate the children, comparing the Choice children with similarly situated children in the Milwaukee Public Schools, though hampered by the fact that these tests are taken by most MPS children only in grades 2, 5 and 7.
Correcting for social class and parental educational background they found, consistently, no difference in the math scores between the Choice and MPS students. The slight advantage MPS students displayed in the reading tests is not statistically significant after correcting for the missing years of test data. In other words, on the basic comparison made by the study, the voucher scheme yielded no improvement in educational outcomes.
How, then, can anybody conclude that MPCP is a miracle cure that could dramatically reduce the educational disadvantages faced by minority students? A 1996 secondary study by Paul Peterson's team found remarkable improvements based on a different comparison than that made by Witte. These findings have been taken up by the right-wing think tanks and press, especially the Heritage Foundation and the Wall Street Journal, neither of which see fit to mention the Witte studies at all.
A small group of the children who applied for places in the voucher schools were rejected, because of oversubscription, by a random lottery, and educated in the MPS schools. Peterson compares the Choice students with these Reject students, and finds that although in reading the two groups of students are virtually identical, in the Mathematics score there is a stark contrast.
The Choice students initially do 2% better than the Reject sample. But when a full set of controls is introduced improvements are 3.5 after one year, 4.4 after two years, rising to 8.5 and 10.9 after three and four years. Those Choice students who stay in the Choice schools long enough seem to do dramatically better than the reject sample who remain in the MPS schools.
But the Peterson result is highly dubious. Witte points out that the Reject group is tiny, and that of the twenty-seven students in the Reject group fully five of them got scores of one, a score which usually just reflects failure to fill in the dots in the test. Remove these students from the comparison, and the remarkable improvements disappear. The MPCP students do no better than the MPS students, or the Rejects—hardly the miracle cure for the inequality in our public schools.
Of course, the failure of one scheme does not imply the failure of all: Each voucher proposal has distinctive features, which make it better or worse suited to success. But the MPCP has many egalitarian features, and that makes its failure particularly telling.
MPCP is restricted to low-income families; it allows schools very little discretion over whom to admit; and it disallows significant extra expenses which might deter the less resourceful. It is precisely this kind of scheme which would have to succeed in order for choice to serve equality.
Vouchers are not, as I have said, the only form of school choice. Public school choice schemes are more prevalent, and more likely to grow, as are charter schools. Studies of these phenomena tend to focus on the way that the parents make their choices, rather than on the results achieved by the students.
Many activists oppose all forms of choice on the grounds that those parents who are already more advantaged will make better choices than those who are already less advantaged, and thus equality will be compounded. Within a regime of egalitarian funding (such as that which exists in Great Britain, the country that has been in the forefront of school choice), this criticism has real bite, although even then it reveals a remarkable distrust of the capacities of ordinary people to make good choices and the extent to which they care about their children.
But in a regime like that of the United States of radically unequal school funding, choice schemes, especially among public schools, have important benefits which may be more important than the unequal choice-making abilities of parents. The establishment of choice (with the possible exception of charter schools) creates a political dynamic, unintended by most advocates of choice, toward equal school funding.
In interdistrict choice schemes, for example, the district that a child leaves has to pay as much for the education of that child as would the district the child enters. If the original district cannot afford to do this, the State has to step in. But it is politically indefensible for states to provide, directly, radically unequal funds for each child —the political defensibility of savage inequalities rests on the impression that local communities rather than the state are the bodies authorizing the funding.
Once that impression is dispelled, as it must be when the state provides the funds directly, the path toward equal funding is made politically easier. Ironically, the form of choice that initiates this dynamic most clearly is the voucher: It would be impossible for the state of Wisconsin to make vouchers for children in the suburbs of Milwaukee worth three times as much as those for the children in the inner city.
Even in intradistrict choice schemes, a dynamic toward equality is established, even though districts typically provide roughly equal per-pupil funding to each school. This is because, on the traditional model of the neighborhood school, each school will have a different density of children who are expensive to educate, depending on how wealthy is the neighborhood in which it is located.
While school districts can and often do use busing to achieve racial integration, they are typically not allowed to use it to achieve class integration. The inequality of provision within districts is typically compounded by job transfer rules in collective bargaining agreements, which make seniority a central factor in whether teachers can transfer. Thus vacancies in the more attractive schools are often taken by more experienced teachers, while the less attractive schools are left to hire the new inexperienced teachers.
Other factors must be taken into account when evaluating choice schemes. Privatization, voucher schemes, and to a lesser extent charter schools clearly pose a threat to teachers' wages, and additionally to union density. Milton Friedman, the guru of school choice, openly cites these as advantages of choice.
But public school choice programs less clearly pose such a threat: As funding and decision-making shifts toward the level of the State, unions gain economies of scale, and the levelling of funding can be expected to lead to a levelling of wages that may make unionization easier, not harder.
One of the most hopeful findings of opinion polls on choice, incidentally, is that those who answer “yes” on choice questions also overwhelmingly answer “yes” when asked whether private schools receiving public funds should be regulated by the state like public schools.
The left's commitment to educational equality means that we must refuse to defend the status quo. Three features of the traditional model of public schooling cry out for attention, regardless of whether choice is on the agenda.
That local taxation is the major source of funding both generates massive inequalities among public schools, and helps to mask and justify those inequalities. Reforms that shift the major sources of funding from local to State and federal taxation, distributed more evenly, would assist greatly in implementing the ideal of educational equality.
One of the greatest sources of unequal educational opportunity, and of limited opportunities for autonomy, is the de facto segregation of residential neighborhoods by race and class. In many districts education activists concerned with justice would do well to lobby their county boards for an integrationist housing policy, so that the inevitable pressure for local schooling results in an integrated public schooling. This would, in addition, make it more likely that children of parents who never attended higher educational institutions would nevertheless have access to adults who did, and that children would, in their out-of-school lives, socialize with people from significantly different backgrounds.
Secession of school districts from one another should be made much more difficult. Under the current system whereby the local regressive real estate tax is the major source of school funding, wealthy communities can “take their money and run” by seceding from poorer communities. Ironically, some choice proposals, forcing school districts to pay for parental choice for schools in other districts, would make secession less relevant by making the borders between districts more permeable.
The balance sheet on choice is, then, complicated. While the left will usually be right to oppose voucher plans and privatization, it cannot defend public schooling as it is actually organized. Distrust of markets should not lead the left to oppose choice of all kinds.
For any choice program we have to look at its specific features, understand what exactly is involved, and what dynamic it will set off, and take a position based on those judgments. We should evaluate a plan on its merits: Will it bring equality nearer? Will it depress wages for teachers? Will it improve the quality of education for the least advantaged children?
These things can only be known by looking at the evidence, and in the long term the evaluation on these criteria may be positive for many public school choice schemes. Even when we oppose some scheme, we owe it to those who are least well served by the public schools not to consider them dupes, and also to look for reforms which really will serve the least advantaged better than the outrageous way they are currently served.
ATC 82, September-October 1999