School Vouchers

by
Michael Apple, Ph.D. and Gerald Bracey, Ph.D.

Center for Education Research, Analysis, and Innovation
School of Education
University of Wisconsin-Milwaukee
PO Box 413
Milwaukee WI 53201
414-229-2716

January 24, 2001

CERAI-00-31

An Education Policy ProjectBriefing Paper

School Vouchers

Contributors: Michael Apple, Ph.D.

Gerald Bracey, Ph.D.

In recent years, some education reformers have advocated the use of vouchersas a means of improving education. The idea of vouchers in education is notnew; a voucher proposal occupies a substantial section of John Stuart Mill’s1838 essay, On Liberty.1 Nor are vouchers unknown in other arenas:Food Stamps, Medicare and Medicaid are all voucher programs of a kind.

Mill's ideas, though, did not find popular expression until MiltonFriedman's 1962 book, Capitalism and Freedom. Among those who wereinfluenced by that book was the government of Chile, which installed anationwide voucher system, and Ronald Reagan, who, as president, proposed theuse of vouchers in this country.

To date, school-voucher experiments have been small and usually ininner-city settings in communities such as Milwaukee, Cleveland, Dayton, NewYork, Washington, DC, and San Antonio. Taxpayer dollars support the Milwaukeeand Cleveland programs. Private philanthropists sustain the programs in theother cities.

The nation’s one large, publicly funded voucher program, a statewide systemin Florida, is currently in limbo. A judge ruled it unconstitutional, but inOctober 2000 his ruling was overturned on appeal.  The case has beenappealed to the Florida Supreme Court.  Under the program, Floridastudents become eligible to receive scholarships to attend private schools iftheir neighborhood school receives an “F” grade from the sate two years out offour.  All of the schools in danger of receiving a second “F” in 1999-2000improved their grades sufficiently. Thus, currently, no school’s studentsqualify for vouchers. Earlier, two schools in Pensacola received consecutive“F’s”. Some 57 students in these two schools received vouchers to attendprivate schools, and they may continue to attend private schools until theconstitutionality of the voucher program is resolved. A second voucher programoffering students with disabilities scholarships to private schools if theyhave not made sufficient progress toward their individual education goals beganin the 2000-2001 school year.

In California and Michigan, statewide voucher referenda failed in theelection of 2000.  Both propositions went down by margins of 3 to 1, witheven minorities voting against them.  It is too early to tell what impactthese losses will have on activities of voucher advocates.

Proponents and critics alike of voucher proposals make a number of claims.

  1.  Vouchers will improve student achievement. By allowingstudents to opt out of failing schools, students can attend schools that servethem better.

To date, the evidence for this has been mixed at best. The outcomes inMilwaukee, for instance, have been hotly debated. One researcher found nodifference between the performance of Milwaukee Public School children andthose using vouchers.2 Others found differences favoringvouchers in both reading and mathematics,3 while a thirdfound an advantage only in math.4  This last researcher also observedthat the voucher children were in small schools andsmall classes, conditionsknown to improve achievement.5  Moreover, in evaluating each ofthese divergent conclusions, it should be noted that, even if the choice andpublic school students were the same at the start of the experiment, theycertainly were not after four years:  the choice students included in thesurvey had been in one school all four years, something quite unusual for poor,inner-city children.6

In another recent, widely reported study that purported to show students whoused vouchers to enroll in private schools did better than a control group ofpublic school students, the company that gathered and analyzed the datadisputed the researchers’ public statements about the conclusions. Analysts at Mathematica Research said the researchers’ announcement of suchresults was premature and exaggerated the findings.7

In the study, which examined privately funded voucher programs in New York,Washington, and Dayton, data were extremely mixed. In New York,African-American students showed gains in both years of the study, but otherethnicities showed small, but insignificant, losses. In Washington, D.C.,African-Americans in grades 2 through 5 showed a significant gain in year onein math and a significant loss in reading. They showed gains in both subjectsin year two. No other ethnic groups gained in either year. In grades 6 through8, African-Americans showed no change in math and a significant loss in readingin year one, but significant gains in math and no significant gain in readingin year two.  In Dayton, African-Americans showed no significant gains ineither subject in year one, but did show a significant gain in reading in yeartwo. No other ethnicities showed gains in either subject in either year.8

The researchers have failed so far to explain why vouchers appeared tobenefit only African American students and not those of other ethnic groups.

Different researchers have obtained differing results because they have madedifferent assumptions about the data. That, in turn, led to different analyses.For instance, one researcher who analyzed data from voucher programs inMilwaukee and New York City concluded that if the voucher studentsscored higher, it might well be because they attended smaller schools withsmaller classes.9

Other researches in other places have yielded similarly equivocal results.This should surprise no one because, in spite of advocates’ statements that“vouchers work” or “the market works,” the impact of local conditions affectvouchers outcomes. Voucher advocate Terry Moe has stated the case quite well:

Ideology aside, perhaps the most vexing problem [of voucher research] isthat few researchers who carry out studies of school choice are sensitive toissues of  institutional design or context. They proceed as thought theircase studies reveal something generic about choice or markets when, in fact --as the Milwaukee case graphically testifies -- much of what they observe is dueto the specific rules, restrictions and control mechanisms that shape howchoice and markets happen to operate in a particular setting. As any economistwould be quick to point out, the effects of choice and markets vary, sometimesenormously depending on the institutional context. The empirical literature onschool choice does little to shed light on these contingencies and, indeed, byportraying choice and markets as generic reforms with generic effects, oftenbreeds more confusion than understanding.10

  2.  Vouchers will increase the diversity of schools and giveparents a wider range of choices.

To date, no data really bear on this issue because voucher students havebeen in small experiments.  Only two new schools in Cleveland opened toaccept voucher students, and these promptly shifted to charter school statuswhen that option became available (the vouchers were worth only $2000, but as acharter, the schools receive $4500 for each student).  In San Antonio,children use vouchers at well-established Catholic schools. In Milwaukee, thevast majority of voucher recipients have been established Catholic schools orestablished non-sectarian private schools. Some new schools have sprung up inMilwaukee in the decade since that city’s experiment with vouchers forlow-income students began, but a few of them have turned out to be fly-by-nightoperations that shut down after failing to deliver on promises. Clearly, thevoucher option increases the number of school available, but whether or not itincreases diversity is another question.

One might draw from the more extensive literature about charter schools, forwhom the same claim is made. There, studies have found that the charter schoolsdo not serve as laboratories for innovation. Indeed, one evaluation of chartersin Michigan found that nothing had been tried in a charter school that hadn'talready been tried in public schools.11International evidence of the results of the competitive market on schools alsoclearly shows that, rather than stimulating diversity in schools andcurriculum, exactly the opposite seems to be the case.  Most schoolsbecome even more alike and tend to employ methods and curricula that have notbeen proven to succeed.12

3.  Vouchers will increase accountability. Unhappy parentswill vote with their feet and their pocketbooks, making the schools directlyaccountable.

There is to date no evidence that this will happen. As noted, most schoolsin voucher programs are long established schools.   Furthermore, itis often extremely difficult for parents who do not have flexible jobs and mustoften depend on public transportation to move their children around a city. While a few children may be helped by vouchers, there may be even lessfinancial support for inner city schools in the long run, leading to fewerresources for those parents who “choose” to keep their children in under-fundedschools because, notwithstanding vouchers, they cannot avail themselves ofprivate education.

Perhaps more importantly, accountability measures that are regularly appliedto public schools are not applied to voucher-accepting schools. Florida, forexample, awards public schools letter grades, A through F, based largely ontest scores. The pupils in a school that receives an F for two consecutiveyears become eligible for vouchers. But the private schools that accept thesestudents do not have to administer the tests that were used to grade the publicschools in the first place. Similarly, Catholic schools in Milwaukee haverefused to release test scores, scores that would be a part of the publicrecord in public schools. Free market principles hold that for markets to workconsumers must have access to high quality information about the product. Thisdoes not appear to be forthcoming from voucher schools. 

4.  Vouchers will make public schools more responsive toparents' wishes because, again, parents have the option to leave. 

Again, there is no evidence that this is happening. A study of existingpublic and private schools found that differences in communities overwhelmeddifferences in governance. That is, public and private schools in suburbanneighborhoods resembled each other. Public and private schools in poor urbanneighborhoods resembled each other, but differed from the schools in the moreaffluent communities.13

Public suburban schools were actually more responsive to parents, the 1999study by Richard Rothstein and others found. These parents, the authors noted,thought they had both a right and a responsibility to take an active role intheir children's education. Private schools were more successful at tellingparents that in matters of curriculum and instruction, all decisions restedwith the school.14

In poor areas, both public and private schools struggled to involve theparents. One could say that these schools were trying to hold the parentsaccountable. Private schools were more successful at this than public schoolsbecause they could make involvement a condition of admission. On the otherhand, the involvement usually did not involve academic activities. In poorpublic schools, parental involvement was dominated by complaints about a poorgrade or an “unfair” disciplinary action.15

  5.  Vouchers enable poor people to obtain a good educationfor their children.

Because, as previously noted, achievement outcomes from voucher programs arein doubt, this might better be stated as “vouchers might allow a few poorpeople to get a better education for their children.”

The number of existing private schools in the nation could handle only about4% of the existing public school children in the nation. Free-market theoristswould likely contend that new schools would spring up to handle the demands,but this is questionable. For one thing, the existing for-profits such asEdison and TesseracT have yet to be profitable.16TesseracT, staggering under $50 million in debt in spite of $8,000 tuition,barely avoided total collapse heading into the 2000-2001 school year, but byOctober had filed for Chapter 11 bankruptcy protection.17Nobel Learning Communities is profitable -- and charges $6,500.00 tuition.18

Moreover, as those who wished to start charter schools have found, it isextremely difficult to obtain space or funding to build space for a new school. And while Catholic schools, which have been hemorrhaging pupilsin recent years, would no doubt welcome the vouchers, many other privateschools have shown little or no inclination to expand.

Arguments Against Vouchers

  1.  Vouchers drain funds from public schools.

This clearly occurs. People have not made much over this todate because the voucher population is small and because increasing enrollmentsin many districts offset the losses.  Indeed, Roy Romer, former governor ofColorado and now Superintendent of School in Los Angeles, is supporting thedevelopment of more charter schools to ease the pressure in his overcrowdedbuildings.  In an area with stable or declining enrollments, however,vouchers would siphon funds from the public schools.

Some people have claimed that vouchers will actually increase funding forpublic schools because the voucher is less than the per-pupil-expenditure ofthe public school. This view mistakenly assumes that the public and privateschools divide a fixed sum of money per child. In fact, schools’ state fundstypically are based on enrollment, and as enrollment -- usually based on aparticular day’s attendance -- declines, so do the funds. In the first year ofCleveland’s voucher program, for instance, funding consisted of $5.25 milliontaken from Cleveland’s share of state aid.19

2.  Vouchers will “skim” or “cream” higher achievingstudents, thus leaving public schools with a higher proportion of more difficultto educate children and fewer funds for that education.

In several instances, students using vouchers had higher test scores thantheir peers before they entered the voucher programs.20While most children came from low-income families, parents of voucher-usingchildren tended to be wealthier, better-educated and more involved in theirchildren’s education both at school and at home.21

  3. Vouchers will have negative effects on teachers andadministrators, as well as students.

The data also indicate worrisome effects on teachers andadministrators.   While some might claim that vouchers do not causeadded hardship for schools, an extensive body of international research onplacing schools in a competitive market can lead to exactly the oppositeconclusion.22 Much more time is spent on maintainingthe image of a “good” school, with much less time spent by teachers andadministrators on curricular substance. Since it is comparative test scoresthat determine whether a school is “good” or “bad,” children who do performwell on such reductive tests are seen as welcome. Those who do not are oftendiscouraged or are marginalized. Once again, the vast majority of childrenharmed by such reforms are exactly those whom voucher supporters state they aresupporting. These conditions exist not because teachers do not already workextremely hard or are uncaring. Rather, markets in schools seem to worsen,rather than improve, work load, pressure, and access to resources.

  4.  There will be a loss of accountability.

Although voucher proponents hold out the goal of more accountability, it ispossible that, in fact, there will be less. As noted, in some instances, theschools that receive voucher students do not have to administer the statetests. In the area of finance, private schools can avoid the kind of auditsthat are routine and public in public schools.

This again is an area that has received little attention because the numberof voucher students is small. It certainly seems reasonable though, that if thevoucher movement attains any size, the public will demand an accounting of howthe public dollars are spent and what they accomplished. With “accountability”on the lips of so many people in regards to public schools, it is hard to seehow this could be avoided for private schools.

Indeed, in Europe, where government subsidy of private schools is common,the private schools are often constrained by the same rules and laws thatgovern public schools: they must have teachers with the same certification,they must offer these teachers the same salary, they must follow the samecurriculum and, in some instances, use the same pedagogy.23

  5.  Vouchers will cost private schools their autonomy.

This is perhaps the converse of point No. 3 above. A number of educators inChristian schools, for instance, oppose vouchers because they believe thatvouchers will inevitably lead to control by the government. Home schoolers inCalifornia opposed that state’s November 2000 referendum on vouchers for thesame reason.24

On a large scale, vouchers would seem likely to fragment communities. When afamily commits to a public school, it commits to good education for the entirecommunity. A family using a voucher is acting only in its self-interest. In general, it would seem that vouchers would remove discussion of social issuesfrom the public domain. Jeffrey Henig captured this problem well:

   Rather than simply focusing on the strengths and weaknessesof private vs. public  institutions as service-delivery mechanisms, weneed to focus on the differences  between private and public institutionsand processes as vehicles for deliberation,  debate and decision making.The real danger in market-based proposals for choice is not that they mightallow some students to attend privately run schools at public expense, but thatthey will erode the public forums in which decisions with societal consequencescan democratically be resolved.

   The market orientation considers education as a product ofpublic and private decisions; as such the issues involved are generic onesapplicable to other domestic policies. But education also has a special statusas a producer of values, perspectives, knowledge, and skills that will beapplied in the ongoing enterprise of    collective deliberationand adjustment.

   While the risk of abuse [from inappropriatesocialization] must be acknowledged, public schools have anothercharacteristic that makes this risk potentially manageable. Compared to otherforces of socialization -- the family, religion, the mass media -- the schoolsare more open to public scrutiny and democratic intervention. 25

The public schools are, in Benjamin Barber's phrase, “workshops of ourdemocracy.”26  This space needs to bestrengthened and protected, not turned over to a market.

1 Mill, John Stuart, On Liberty. InGray, John, ed., John Stuart Mill: On Liberty and Other Eassys. NewYork: Oxford University Press. 1991.

2 Witte, John, Sterr, Troy D., and Thorn,Christopher A.,  Fifth Year Report: Milwaukee Parental ChoiceProgram.  Madison, Department of Political Science, University ofWisconsin at Madison, December, 1995

3 Greene, Jay P., Peterson, Paul E., and Du,Jiang Tao,  The Effectiveness of School Choice in Milwaukee: ASecondary Analysis of Data from the Program's Evaluation, August 1996. Available at: http://data.fas.harvard.edu/pepg/op/evaluate.htm

4 Rouse, Cecilia,  “Private SchoolVouchers and Student Achievement: An Evaluation of the Milwaukee ParentalChoice Program.”  Quarterly Journal of Economics, May, 1998, pp. 553-602

5 Ibid.

6 Bracey, Gerald W.  “The Sixth BraceyReport on the Condition of Public Education.”  Phi Delta Kappan,October, 1996, pp.  127-138

7 Zernike, Kate.  “New Doubt Cast Is Caston Study that Backs Voucher Efforts.”  New York Times, September15, 2000, p. A21.

8 Howell, William G., Wolff, Pagtrick J.,Peterson, Paul E., and Campbell, David E., “Test-score Effects of SchoolVouchers in Dayton, Ohio, New York City, and Washington, D.C.: Evidence fromRandomized Field Trials.” Available at: http://fas.harvard.edu.pepg.

9  Rouse, op. Cit.

10 Moe, Terry, Private Vouchers. Stanford, CA: Hoover Institution Press, p. 20.

11 Horn, Jerry, and Miron, Garon.  Evaluationof the Michigan Public School Academy Initiative.  Kalamazoo, MI: TheEvaluation Center, School of Education, Western Michigan University, 1999.

12 Whitty, Geoff, Power, Sally, and Halpin,David, Devolution and Choice in Education. Philadelphia: Open UniversityPress, 1998

     Lauder, Hugh, and Huges, David, Trading inFutures: Why Markets in Education Don’t Work. Philadelphia: Open UniversityPress, 1999

13 Rothstein, Richard, Carnoy, Martin andBenveniste, Luis.  Can Public Schools Learn from Private Schools? Washington, DC: Economic Policy Institute, 1999

14 Ibid.

15 Ibid.

16 Information available at www.edisonschools.com. Click on “information for investors” to see Prospectus filed with SecuritiesExchange Commission

17 Creno, Glen, “TesseracT Group Files Chapter11.”  Arizona Republic, October 10, 2000.

18 Barbara Presiassen, Retired Nobel VicePresident, personal communication, September, 2000.

19 Murphy, Dan, Nelson, F. Howard, andRosenberg, Bella, The Cleveland Voucher Program: Who Chooses, Who GetsChosen, Who Pays? Washington, D.C.: American Federation of Teachers, 2000

20 Godwin, R. Kenneth, Kemerer, Frank R., andMartinez, Valerie J.  Final Report, San Antonio School Choice ResearchProject.  Denton, Texas:  Center for the Study of EducationReform, College of Education, University of North Texas, June, 1997.

21 Godwin, 1997

22 Center on Education Policy, Lessons fromOther Countries About Private School Aid.  Washington, DC, 1999

23 Patrick, Ed., Choice in Education. It Sounds Wonderful, But….  East Moline, Illinois, MacArthurInstitute, undated.

24 Duffy Cathy, “Problems With the CaliforniaVoucher Initiative, Proposition 38.” Available at:http://www.grovepublishing.com.

25 Henig, Jeffrey, Rethinking SchoolChoice: Limits to the Market Metaphor.  Princeton, NJ: PrincetonUniversity Press, 1994, pp. 200-203

26 Barber, Benjamin, “Workshops of OurDemocracy,” Education Week, April 19, 1995, p. 34