Jump to:Page Content
CAMBRIDGE, MA -- Do school vouchers take the best and the brightest from public schools? Do vouchers allow private schools to skim the "cream" from the top of the public school population? Do low-income families use vouchers to obtain a school they prefer? Are low-income voucher students at risk of suspension and expulsion? A new study, directed by Harvard’s John F. Kennedy School of Government and Mathematica Policy Research, looks at a privately-funded voucher program in San Antonio, Texas, to provide answers to these and other questions.
The research, which shows minimal differences in math scores and modest differences in reading scores between vouchers students and their counterparts in the Edgewood public school district, is of national interest because the program is similar to the voucher program just beginning in Florida. As in Florida, the program offers vouchers to every low-income student attending an eligible public school. However, unlike in Florida, there is no requirement that students come from schools with average test scores below a particular threshold.
The researchers found that voucher students come from varied backgrounds. In some respects, they resemble those remaining in the Edgewood public schools; in others, they are modestly less disadvantaged than their public-school counterparts.
"The voucher program in Edgewood seems to be serving the working poor," said David Myers, senior fellow at Mathematica Policy Research in Washington, DC, and director of the study, which followed students for one year.
"We didn't find much 'creaming,'" added co-principal investigator, Paul E. Peterson, who directs the Kennedy School's Program on Education Policy and Governance. "If you want to use the dairy analogy, it was more like '2 percent milking.'"
Whether or not vouchers attract the best and the brightest has fueled much of the debate over vouchers, even though data to inform the debate have been lacking. The Texas Federation of Teachers said the voucher program in the Edgewood school district would "shorten the honor roll" in public schools. Edgewood school superintendent Dolores Munoz stated: "I guarantee you that at least 80 percent will be the high-achieving students. The private schools . . . their doors are . . . not open for every child."
The study provides the following statistics on the actual degree to which the San Antonio voucher students differed from their public-school counterparts:
Voucher students scored at the 37th national percentile rank on the math test; Edgewood public-school students scored at the 35th. The difference is not significant. Voucher students scored at the 35th percentile on the reading test; public-school students scored at the 28th. Difference is significant. Twenty-three percent of voucher students had been in programs for gifted students; 29 percent of public-school students had been. Difference is not significant. Eight percent of voucher students had participated in special education programs, as compared to 16 percent of public-school students. Difference is significant. Voucher family income averaged $16,000, while public-school family income average $15,900. Ninety-six percent of voucher mothers were Latino, as compared to 93 percent of public-school mothers. Neither difference is significant. Mothers of voucher students had completed, on average, 12 years of education, as compared to 11 years for mothers of public-school students. Fifty percent of voucher mothers were employed full time, as compared to 37 percent of public-school mothers. Both differences are significant.
Four percent of voucher mothers were receiving welfare, compared with 5 percent of public-school mothers. Forty-five percent of voucher parents were living together, compared with 43 percent of public-school families. Neither difference is significant.
Twenty-two percent of voucher mothers were receiving food stamps, as compared to 33 percent of public-school mothers. Difference is significant. Additional findings from the evaluation include:
Does family prefer this school?
Ninety-two percent of voucher families said their child was in a school the family wants the child to attend, as compared to 75 percent of public school families. Difference is significant. Suspension rates?
Five percent of both voucher and public school students had been suspended from school. Who changes schools?
Seven percent of voucher students had changed schools since the beginning of the school year, as compared to 16 percent of public-school students. Difference is significant. Dangerous weapons at school?
Thirty-nine percent of Edgewood public-school parents report that guns and other dangerous weapons are a very serious problem, as compared to 28 percent of voucher parents. Difference is significant. Fighting at school?
Fifty-two percent of Edgewood public school parents report fighting as a serious problem, as compared to 28 percent of voucher parents. Difference is significant.
Voucher parents, on average, report their child is in a class of 21 students, as compared to a class of 20 students in Edgewood public schools. Difference is not significant academic quality?
Sixty-one percent of voucher parents are "very satisfied" with the academic quality of the school, as compared to 35 percent of public school parents. Difference is significant. The San Antonio voucher program, dubbed "the Horizon program," is sponsored by the Children’s Educational Opportunity (CEO) foundation. It offers vouchers to all students from low-income families in grades K-12 attending Edgewood public schools, which serve a predominantly low-income, Latino population. The program was announced in April 1998 for the following school year. Vouchers may be used at a private school in the San Antonio metropolitan area or a public school outside the Edgewood school district. Elementary students receive scholarships worth up to $3,600; high school students receive scholarships worth up to $4,000. Horizon officials report that during the 1998-99 school year 837 students used scholarships to attend 57 schools.
The evaluation team obtained information during the 1998-99 school year from three groups of families with children between the ages of 8 and 17: 1) those receiving a Horizon voucher; 2) a random sample of students and their families residing in the Edgewood district; and 3) a random sample of students and their families in three comparison districts in Texas.
The evaluation asked Horizon and a random sample of Edgewood public-school parents and students to complete a questionnaire concerning the school the student was attending. Students also took the Iowa Test of Basic Skills (ITBS) in reading and mathematics. Tests were scored by the publisher of the ITBS. National Percentile Ranking (NPR) scores were calculated; scores can vary between 1 and 100, with 50 as the national average. Tests were administered during the first school year, 1998-99, thereby providing information on student abilities shortly after they had entered private school.
The full report is available on the web at http://data.fas.harvard.edu/pepg/.
The authors are Paul E. Peterson, Program on Education Policy and Governance, Kennedy School of Government, Harvard University; David Myers, Mathematica Policy Research; and William G. Howell, also of the Program on Education Policy and Governance. The evaluation was funded by the David and Lucile Packard Foundation.
Mathematica, one of the nation’s leading independent research firms, conducts public policy research and surveys for federal and state governments as well as private clients. Its mission is to improve public well-being by bringing the highest standards of quality, objectivity, and excellence to bear on the provision of information collection and analysis to its clients. The Harvard Program on Education Policy and Governance is located in the Taubman Center on State and Local Government in the John F. Kennedy School of Government, Harvard University, Cambridge, Mass.