The Black Informant

Although this press release is roughly 3 years old, the information is quite relevant for today as there has been a continual upswing of black parents making the switch to private or homeschooling.

====

Washington, D.C. (February 19th, 2002)—At the conclusion of a three-year evaluation, the test scores of African American students participating in New York City’s privately funded voucher program were substantially higher than the test scores of comparable students in public schools. The difference in scores—over nine national percentile points—was moderately large, enough to erase almost half of the achievement gap found nationwide between African American and white students. The test scores of students of other ethnic (mainly Hispanic) backgrounds in New York City’s public and private schools, however, did not differ significantly from one another.

These findings are included in a forthcoming Brookings Institution Press book, The Education Gap: Vouchers and Urban Schools by William G. Howell and Paul E. Peterson, to be released on April 15th. Howell and Peterson, together with Patrick J. Wolf and David E. Campbell, report key findings from multiple studies.

Results from Washington, D.C. are less certain. Test scores of African American students were nine national percentile points higher than the public school comparison students after two years, but no significant differences were detected after three years. These fluctuations in the data could have been caused by the presence of the large-scale charter school intervention in D.C.

In both New York and D.C., private school parents remained much more satisfied with their children’s schools after three years than were public school parents.

The evaluations in New York City, Washington, D.C., and Dayton, Ohio were conducted as randomized field trials so that valid comparisons could be made between groups, a practice similar to that used in field trials conducted in medical research.

According to Peterson, the Henry Shattuck Professor of Government at Harvard University, “the striking differences between the two cities may be due to the presence of charter schools in D.C. —but not in New York City—during the study’s evaluation period.” Approximately 16% of D.C. students attended charter schools in 2000-01, whereas the charter school presence in New York was minimal. Among those participating in the Washington, D.C., evaluation, 17% of those offered vouchers—and 24% of the control group—attended charter schools, while none of the students in New York City did. (Charter schools do not charge tuition, while the Washington, D.C. vouchers generally covered about half the tuition at private schools.)

“Such a high utilization of the charter school option in D.C., along with the low acceptance rate among students offered vouchers, probably confounded efforts to estimate voucher impacts,” points out co-author William Howell, an assistant professor at the University of Wisconsin. Among the students participating in the New York City evaluation, 70% of those offered a voucher made use of it for three years, allowing for more precise estimates of the voucher effect. In D.C. only 37% did, sharply reducing the precision of the estimates.

Results for the Dayton evaluation, which was conducted over a two-year period, were reported previously. Data collection in New York City was under the direction of David Myers of Mathematica Policy Research; Myers and Peterson were co-principal investigators for the New York evaluation.

Other findings were quite consistent across different sites:

* In combined results for all cities for which information was available, African American test scores, on average, were three percentile points higher in year one, six percentile points higher in year two and, again, six points higher in year three.

* The nine national percentile point gain reported in New York City represents a 0.45 standard deviation (one full standard deviation is approximately the difference between black and white test scores nationwide). National percentile points range between zero and 100, with a median of 50.

* The scores of non-African American students in public and private schools did not differ significantly from one another. The authors explain the differences in the results for African American and other students by the lesser amount of choice in American public education available to black Americans. They point out that “African American families send their children to the least desirable public schools because of more limited incomes and racial discrimination in the housing market.” As a consequence, they say, “new forms of choice, such as vouchers and charters, are more likely to benefit black students.”

* According to the reports of parents participating in the three-city evaluations, private schools had, on average, 172 fewer students, constituting a 38% reduction in school size. Private school classes had an average of 20 students, as compared to 23 students in public schools.

* Students who attended private schools experienced significantly fewer problems such as fighting, cheating, property destruction, racial conflict and truancy. For instance, only 32% of private school parents reported that fighting was a serious problem at their school, as compared to 63% of public-school parents. While 22% of private school parents claimed that destruction of property was a serious problem, fully 42% did so in public schools.

* Private schools maintained closer communications with parents than public schools did. For instance, private school parents were 20 percentage points more likely to receive a newsletter from their school than were public school parents. Private school parents were also 14 percentage points more likely to receive notes from teachers.

* Students attending private schools reported doing more homework than their peers in public schools.

* Private schools had more limited programs and facilities. For instance, public schools were, on average, 28 percentage points more likely to offer programs for non-English speakers than were private schools. Public schools were also 19 percentage points more likely to have a nurse’s office, and 14 percentage points more likely to have a special education program. Private schools, however, were six percentage points more likely to offer after-school programs and 16 percentage points more likely to provide individual tutors.

* Parents who used vouchers were significantly more satisfied with their child’s school than either (1) parents who would have used a voucher had they been offered one, or (2) a cross-section of public school parents, whether or not they had applied for a voucher. For instance, in Dayton 51% of private school parents were “very satisfied” with the academic program at their child’s school, as compared to 19% of public school parents generally and nine percent of parents whose child would have attended a private school had the family been offered a voucher. While differences in parent satisfaction between public and private school were the largest after one year, they still remained very large after three years.

* While the voucher programs enhanced the frequency of religious observance among students, they had a slightly negative impact on religious observance of their parents.

* The voucher programs did not have strong, consistent effects, either positive or negative, on parents’ involvement in their child’s education, student’s self-esteem, student friendship patterns, racial integration, political tolerance, or the formation of social capital.

* While voucher program participation rates appeared somewhat higher among families with greater incomes, more education, and regular religious practices, the impacts were not consistent across the various programs nor did they appear particularly large in any one city. On the whole, only a modest level of “skimming” was observed in these voucher programs.

The authors recommend further research be conducted on larger, better-funded voucher programs in central cities with high concentrations of African Americans.

The findings contained in this press release come from the randomized field trials in New York City, Dayton, and Washington, D.C. In the three cities, 4,159 students were initially tested, of which 2,756 were African American. In total, 3,142 family questionnaires were completed, of which 2,152 were filled out by African American parents.

The New York City program was funded by the School Choice Scholarships Foundation; the Washington, D.C., program by the Washington Scholarship Fund; the Dayton program by Parents Advancing Choice in Education. In addition, The Education Gap also provides data from a randomized field trial of the national program conducted by the Children’s Scholarship Fund and an evaluation of the Edgewood program in San Antonio, Texas, operated by the CEO-Horizon program.

The New York City evaluation was conducted between spring 1997 and June 2000. The Washington, D.C. evaluation took place between spring 1998 and May 2001. The Dayton evaluation began in spring 1998, and final data collection was completed by May 2000. New York and Edgewood data were collected by Mathematica Policy Research, Inc; Washington, D.C. data were collected under the direction of Paul E. Peterson and Patrick J. Wolf; the Dayton and Children’s Scholarship Fund data were collected under the direction of Peterson and David E. Campbell. Baseline and first year information in Dayton and Washington, D.C. were collected with the assistance of the Public Opinion Laboratory at Northern Illinois University.

William G. Howell is assistant professor of political science at the University of Wisconsin, Madison. Paul E. Peterson is Henry Lee Shattuck Professor of Government and director of the Program on Education Policy and Governance at Harvard University. He is also a senior fellow at the Hoover Institution at Stanford University. Patrick J. Wolf is Assistant Professor in the Georgetown Public Policy Institute at Georgetown University. David E. Campbell is a fellow at the Center for the Study of Democratic Politics at Princeton University.

The evaluations were supported by the following foundations: Achelis Foundation, BASIC Fund, Bodman Foundation, The Lynde and Harry Bradley Foundation, The William Donner Foundation, The Thomas B. Fordham Foundation, Milton and Rose D. Friedman Foundation, Gordon and Llura Gund Foundation, John M. Olin Foundation Inc., David and Lucile Packard Foundation, Smith-Richardson Foundation, Spencer Foundation, and the Walton Family Foundation.

CONTACT:

Nicole Pagano, Brookings Institution Press, (202) 797-6106

[Link]