A series of public remarks and opinion columns by the president of the American Federation of Teachers has helped reignite a long-running debate about the relative merits of public and private schools in fostering academic achievement.
For the past three months, Albert Shanker has repeatedly hammered away on the issue in an effort to explode what he calls the myth of private-school superiority.
In press conferences, testimony before the Congress, and in his paid newspaper advertisements, Mr. Shanker has cited data from the 1990 National Assessment of Education Progress mathematics assessment, which showed that, at the 12th grade, private school students performed only slightly better than those in public schools.
“This is amazing,” he wrote in the column he purchased in the Sept. 8 New York Times, “because youngsters in private schools are a far more advantaged group, so they should be leaving public-school students behind in the dust.”
The data suggest, the union chief argued, that any attempt to provide vouchers to send children to private schools--as the Bush Administration has proposed in its America 2000 plan--would be a mistake. Rather, he said, the answer is to improve education in all schools.
An examination of test-score data and interviews with numerous scholars and educators, however, suggest that Mr. Shanker is only partly right, and that the private school advantage is not completely a myth.
While some tests show that students in the two sectors performed nearly equally, particularly in math and particularly in the upper grades, others show that private-school students substantially outperformed their public-school peers.
Moreover, these differences do not simply reflect the higher socioeconomic status of private-school pupils, a number of researchers say. In many cases, they note, private schools are at least as diverse as public schools, and, in some cases, public schools have distinct advantages.
John E. Chubb, a senior fellow at the Brookings Institution and a leading advocate of private-school choice, accused Mr. Shanker of making selective use of data.
“I tend to view this as politics, not social science,” Mr. Chubb said last week. “If there weren’t a proposal to include private-school-choice plans in America 2000, you wouldn’t hear anything about this from the A.F.T.”
But Bella Rosenberg, a special assistant to Mr. Shanker, responded that the steady stream of arguments by the union president was an attempt to show that choice as an educational strategy is misguided.
“The Bush Administration and others are telling the nation that public aid to private education is a strategy for overcoming the crisis in education,” she said. “This is an empirical question .... The evidence goes squarely against that as a strategy.”
“We have an alternate strategy,” Ms. Rosenberg said. “The Administration would reward schools for attracting customers. Our strategy is to reward schools for improving student achievement.”
‘Nearly Identical’ Results
The issue has taken on an increasing urgency in the past year as the notion of private-school choice has moved up the education agenda. Last year, for example, Mr. Chubb, along with Terry M. Moe, a professor of political science at Stanford University, unleashed a vociferous debate with their book, Politics, Markets, and America’s Schools, which called for a free market in education.
The latest round began in June, when Mr. Shanker, a forceful critic of the Chubb-Moe book, read the results of the 1990 math assessment.
Unlike in prior years, the assessment included a large sample of private-school students, to permit a more detailed analysis of their performance.
The math results showed that, over all, the average proficiency of Catholic- and private-school 12th graders slightly exceeded that of the public school students. On a 500-point scale, the public schoolers’ average score was 295, compared with 302 for those in Catholic school and 301 for those in other private schools.
Moreover, the proportion of seniors who attained the highest level of proficiency on the exam--meaning that they could demonstrate “reasoning and problem solving involving geometric relationships, algebraic equations, and beginning statistics and probability,” according to NAEP--was virtually identical in the three sectors. Some 4 percent to 5 percent of the seniors attained that level, the NAEP report found.
A further analysis, according to Mr. Shanker, showed additional similarities. Examining the performance of students according to the level of parental education, he wrote in the Sept. 8 column that “the children of high-school dropouts score about the same, whether they go to public or private schools--and the same is true for the children of college graduates.”
Mr. Shanker also found that students who took the same courses scored at about the same level. “The results are nearly identical,” he wrote, “even though public-school students come from far less advantaged backgrounds.”
Corroborating Data
Mr. Shanker noted that other test data also suggest that the gap between public- and private-school performance is small.
In 1988, Chester E. Finn Jr., then head of the U.S. Education Department’s research branch, told an audience of private-school educators that unpublished data from the 1986 NAEP assessment revealed that private-school 17-year-olds outperformed those in public-schools by 6 percentage points in history, and that private-school students in grades 3, 7, and 11 only slightly outperformed their public-school peers in reading. (See Education Week, March 9, 1988)
An examination of some other data also showed only small differences in performance. For example:
- On the mathematics portion of the Scholastic Aptitude Test, public-school students outscored those from religious schools by 473 to 472 on a 200-to-800 scale. Those from independent schools vastly outdistanced both, with an average score of 524.
- On the 1990 American College Testing Program examination, public-school students earned a composite score of 20.6 out of 35; those from private schools, 21.0; and those from Catholic schools, 21.1.
- On the 1984-85 norming sample of the Iowa Test of Basic Skills, in language arts, Catholic-school students performed about one month-or one-tenth of a grade--ahead of those in public schools. In math, the scores were about even, according to H.D. Hoover, director of the Iowa Basic Skills Testing Program. In grade 8, the Catholic-school students performed about three months ahead of those in public schools in language arts, Mr. Hoover noted. At all grades, those in other private schools were significantly ahead.
- John F. Witte, a political scientist at the University of Wisconsin at Madison, who last year issued a blistering scholarly critique of the Chubb-Moe book, said the federal High School and Beyond Survey, on which the book was based, found “very small” differences in achievement between the two sectors.
‘Significant Difference’
In his Sept. 8 column, Mr. Shanker also noted that James S. Coleman, the University of Chicago sociologist whose analysis of the High School and Beyond data first printed up the existence of a private-public gap, had stated that “one should not make a mistake: Our estimates for the size of the private-sector effects show them not to be large.”
In an interview last week, however, Mr. Coleman stressed that the difference nevertheless exists.
“It’s a question of whether the glass is half-empty or half-full,” he said. “relative to differences among students, or even relative to differences in family background, the differences in the sector of the school system is not as large as either of those.”
But “it’s a significant difference,” he contended.
In addition to Mr. Coleman’s research, other test data support the conventional wisdom and show appreciable gaps between the public and private sectors. These include:
- The same 1990 NAEP math assessment Mr. Shanker has cited. At the 4th grade, the average proficiency of for the public schools was 214; for Catholic schools, 224; and for private schools, 231. At the 8th grade, public schools averaged 264; Catholic schools, 278; and private schools, 274. . NAEP’s 1988 writing assessment. At the 4th grade, the average proficiency of those in public schools was 189.4; for those in nonpublic schools, 203.4. At the 8th grade, the average public-school score was 206.7; in nonpublic schools, 231.3. At the 12th grade, in public schools, it was 222.1; in nonpublic schools, 236.7.
- The National Educational Longitudinal Survey of 1988, which tested students in math, reading, science, and history as part of its survey of 25,000 8th graders. According to the National Center for Education Statistics, students from private independent schools “perform considerably better on all tests than students from Catholic schools,” and those from Catholic schools have “higher mean scores in all tested areas (except higher-level math problem solving) than do public-school students.”
- The Stanford Achievement Test. Data from the 1988 standardization for the eighth edition of the test indicate that “nonpublic students have higher average performance than students in suburban, rural, or urban schools across subject areas and grade levels,” according to the Psychological Corporation, the firm that publishes the test.
Officials from the company emphasize, however, that the data do not suggest “that the quality of education in nonpublic schools is necessarily better than that in the different types of public schools.”
Scholars who have studied public and private-school differences say that the school populations tested can account, in part, for the apparent discrepancies in the data.
For one thing, said Sister Catherine T. McNamee, president of the National Catholic Educational Association, the 12th-grade students surveyed in the 1990 NAEP assessment excluded those who had already dropped out, presumably mostly poorer performers. The absence of such students, she argued, would raise the national average for public schools, which have a much higher dropout rate than private schools.
“At that point, many public schools had lost their least-talented students,” she said. “In the Catholic schools, they are still there.”
Mr. Chubb or Brookings also suggested that the public-high-school average may also benefit from an infusion of students who had attended private elementary schools. Between a third and a half of those who attend private schools in grades K-8 move to public high schools, he said.
In addition to changes in the public-school population, the diversity of the private-school pool may also explain the relatively narrow difference in performance, according to Bruce S. Cooper, a professor of education administration and policy at Fordham University’s graduate school of education.
While the term “private school” often conjures up the image of elite boarding schools, such institutions represent a small fraction of the sector, he said. Of the 22 types of private schools, 19 are religiously affiliated, few of which are selective in their admission policies.
Anthony S. Bryk, professor of education at the University of Chicago, also said the narrow difference in average math scores may reflect the advantages larger public schools have over many private institutions. Such advantages--including the opportunity to offer more advanced courses--may be why the proportion of students who attained the highest level of performance on the NAEP math exam was similar in the two sectors, he said.
Course Taking, School Climate
While much of the debate has focused on test scores, such national averages mask the true differences between the types of schools, said Paul T. Hill, a researcher at the RAND Corporation.
Mr. Hill, who conducted a 1990 study of high schools in New York City and Washington, found that students with similar family backgrounds fared better in parochial schools than in public schools.
“All the evidence,” he said, “is that, for an urban minority population, there is a large difference [favoring] the parochial schools’ productivity.”
Mr. Hill suggested that one of the major differences between the two sectors is in course taking. Most students in Catholic schools tend to take more academic coursework than their peers in public schools, he said.
He took issue with Mr. Shanker’s claim that the similar performance among students who took the same courses represents a “public-school advantage.” On the contrary, Mr. Hill said, the fact that more students in the private sector take such coursework is an advantage.
“If you factor out the distinguishing difference between the schools, and then say the outcomes don’t differ, you’ve muddied the waters,” he said.
Catholic schools also have an advantage in school climate, said Mr. Bryk of the University of Chicago, by fostering a “strong sense of community that favors academic work on the part of all students.”
Such factors are even more pronounced in independent schools, which surpass public schools on a range of factors, said Joyce McCray, executive director of the Council for American Private Education.
“Education is more than one test,” she said. “There is a tremendous difference in the way the [two types of] schools are organized, and in the mission of the schools.”
Ms. Rosenberg of the A.R.T. said she agreed that course taking and school climate are key factors in achievement, and noted that the research on private schools has brought that to the surface.
“There are two lessons to be learned from private schools,” she said. “One is that academic courses is the way to go .... The other lesson to be drawn is [the importance of] discipline.”
