School Choice vs. School Choice

Policy Backgrounders | Education

No. 155
Friday, April 27, 2001
by John C. Goodman and Matt Moore


Choice Outside the Housing Market: Effects on Public Schools

Public schools enroll 89 percent of the nation's primary and secondary children and consume 92 percent of the country's education spending.43 Even if we wanted to, we could not move all these children into private or charter schools in a short time. So one of the most important questions to ask about new systems of school choice is: what is the impact on public schools and the children who remain in them?

Do School Choice Programs Cream the Best Students? Critics of school choice fear that vouchers will draw the best students from public schools. The most involved parents and the most motivated students will be most likely to choose an alternative school, they argue, leaving the public schools to educate an increasingly difficult population without the support of informed, engaged parents. This will cause bad schools to become even worse.

Figure III - At-Risk Students in Texas Schools

"More than 100,000 students currently attend private schools with public money."

A problem with this argument is that it ignores one of the most important reasons why parents enroll their children in private schools: private schools are a refuge for students who fail in the public schools. Not only do parents seek alternatives for those children, but public school systems themselves frequently turn to the private sector for help with the most difficult students. According to the U.S. Department of Education, more than 100,000 students currently attend private schools with public money. Students with serious emotional disturbances account for 40 percent of these students.44

"The population of Texas charter schools shows no evidence of creaming. In fact, more than two-thirds of their enrollment are at-risk children."

The population of charter schools shows no evidence of cream skimming. In Texas, 68 percent of the students attending open-enrollment charter schools are classified as at-risk (because of limited English proficiency, poverty, race, or geographic location), compared to 39 percent in Texas' traditional public schools.45 [See Figure III.] In addition, seven of the 19 state charter schools now in operation are specifically for students who have dropped out of other schools.46

Voucher programs do not skim cream either. In a 1999 evaluation of the Cleveland program, Paul Peterson discovered that "[Cleveland] Choice students, on average, have significantly lower family incomes than do Cleveland City public school students ($15,769 vs. $19,948), are significantly more likely to be raised by only their mother (68.2 percent vs. 40 percent) and are significantly more likely to be African-American (68.7 percent vs. 45.9 percent)."47 Further, evaluations of the Milwaukee program and others demonstrate that they are not only enrolling students from low-income families, they are also enrolling students with below-average scores on achievement exams.

San Antonio's HORIZON program offers a scholarship to every student in the entire Edgewood School District. The 545 families in the HORIZON program have a slightly lower average family income than the Edgewood public school families.

What Happens to the Test Scores of Students Who Remain in the Public Schools? Schools that serve low-income communities typically operate as virtual monopolies. Their "customers" are held captive, unable to choose a rival institution. Like monopolies in business, they have little incentive to improve the quality of their product. However, when a school choice program emerges, the incentives quickly change.

Research conducted by Harvard economist Caroline Hoxby shows pronounced academic improvement in areas where public and private schools compete for the same students. Among students transferring from public to private schools, Hoxby found a 12 percent increase in future wage gains and a 12 percent increase in the probability of college graduation. But interestingly, Hoxby also found an 8 percentage point increase in the test scores of the students who remained in public schools.48 This suggests that by forcing public schools to compete in areas where they previously had a monopoly, school choice programs improved the educational outcomes of all students.

"Every student in Giffen Elementary School, a failing school in ALbany, N.Y., was offered a voucher; the public schools responded by making radical changes."

Considerable anecdotal evidence supports Hoxby's conclusion. Giffen was perhaps the worst public elementary school in Albany, N.Y., when philanthropist Virginia Gilder offered private school scholarships to all of its students. Within months of the offer, the public school establishment responded by making radical changes. The school board installed a new principal, hired two new assistant principals, moved 10 teachers to other schools, and set aside $125,000 for books, equipment and teacher training.49

After Milwaukee's choice program was expanded to allow participation of up to 15 percent of public school enrollment, about 15,000 students, the public school system closed its six worst schools and responded in other ways:50

  • More childhood education programs were developed, kindergarten programs for the 5-year-olds developed into all-day schedules and the number of kindergarten programs for 3-year-olds tripled.
  • Before- and after-school programs expanded. In 1995 the public school system had one school with before- and after-school child care and tutoring programs for low-income families. Today, there are 82 programs.
  • Charter schools grew. In 1995, only one charter school was authorized. By the 2000-2001 school year, six additional charter schools were in operation.
  • Access to health care improved. Two public schools had health clinics in 1995. Today there are 47.

Florida's A+ program provides additional evidence.51 Florida's program ranks schools by grades of A through F. Students in schools that receive an F for any two years in a four-year period receive vouchers to attend another public or private school of their choice. Meanwhile, the state takes over and reorganizes the school.

Table II - Test Score Gains in Florida Public Schools by Assigned Rank (A through F)

"Florida's F and D schools took aggressive steps to improve performance."

Because Florida's efforts are targeted at failing schools, we would expect F schools to make more serious reform efforts than passing schools. After all, if F schools fail to improve, they are "voucherized," whereas if A through D schools do not improve they are slapped on the wrist. A detailed analysis of Florida's program by Jay Greene shows that student scores did indeed improve more quickly for students in schools that faced vouchers (F schools). According to Greene, the year-to-year change in test results does not differ among schools that received A, B or C grades. However, schools that received D grades improved a bit more than their superior counterparts. The F schools - those that faced "voucherization" if their poor performance was repeated - registered the greatest gains of all. [See Table II.]

"Florida schools that faced 'voucherization' if they failed registered the greatest student performance gains of all."

Is it possible that F schools just had more room for improvement? Would that account for the greater pace of change? Anticipating these questions, Greene compared higher-scoring F schools to lower-scoring D schools, which ensured that the improvements realized by failing schools were indeed due to the threat of vouchers. Both F and D schools had a history of low performance and faced pressures to avoid repeating low performance. While both "types" of schools were alike in many respects and perhaps were distinguished only by chance, F schools faced a much harsher future if they failed to improve. Green found that the improvement achieved by higher-scoring F schools was greater than the gains realized by lower-scoring D schools.

Another study of Florida's A+ program by Carol Innerst for the Center for Education Reform reveals the steps Florida's F and D schools took to improve performance, including longer school days, additional teacher in-service days for professional development and special programs to improve math and reading skills for at-risk students.52 For example, in Escambia County, location of a number of the schools deemed as failing in 1999, public school officials provided Saturday tutoring, hired new teachers and required parent-teacher conferences each grading period.53

Other examples include Gadsden County School District's development of direct instruction programs, Palm Beach County School District's establishment of classroom libraries and closer observation of teachers in failing schools, Polk County School District's new language arts program and Volusia County School District's reduced class sizes and reading specialists.54

What Are the Effects on Public School Finances? Critics often claim that choice programs will drain resources from the public schools. In an ideal choice system, the critics would be right. Indeed, the loss of funds that follows a loss of students is the primary motivator for improvement among schools that compete with each other. This, after all, is the way competition works in other markets.

However, most public school systems have not lost money as a result of an exodus of students. Rather, the public schools have gained. For example, the Milwaukee Public School system makes a profit on students who participate in the voucher program. In 1996-97, the district sent about $4,400 per student to the private schools the students enrolled in, but collected from the state $7,500 - earning a $3,100 "profit" for each student it no longer educated.55

The Cleveland choice program also benefits the public schools financially. Public schools receive full per pupil funds for each student enrolled in the voucher program and, as in Milwaukee, are subsidized for students they no longer teach. In 1996-97, Cleveland public schools received a net surplus of more than $118,000 because of the voucher program.56


Read Article as PDF