When parents have more options for their children’s schooling, academic achievement, salaries, and graduation rates in school systems all tend to improve—but not by much.
That conclusion comes in a recent review of studies looking at the effects of competition on education. For the report, researchers from the National Center for the Study of Privatization in Education—based at Teachers College, Columbia University—analyzed 35 studies published over the past 30 years comparing states and regions in what were deemed more competitive educational markets with those with fewer schooling options.
“The Effects of Competition on Educational Outcomes: A Review of U.S. Evidence,” September 2001, from the National Center for the Study of Privatization in Education. (Requires Adobe’s Acrobat Reader.)
“What we found is that choice does seem to improve educational outcomes a bit, but it’s not a revolutionary effect,” said Henry M. Levin, a study author and the director of the 2-year-old study center.
The study is unusual because it focuses on the effects of competition on entire populations, school systems, or public schools. Many evaluations of school choice plans, for example, have concentrated on the impact on the students doing the choosing.
To gauge the degree of educational competitiveness in the areas studied, the reports reviewed used a variety of techniques. The most common one was to measure the proportion of school-age children in a given area who were enrolled in private schools. Other studies used the Herfindahl Index, a rating commonly used in economic-market studies to measure, for example, how much choice consumers have in the soft drinks or cars they buy.
A smaller number of studies compared regions, such as the greater Boston area, where families moving in can choose from half a dozen school districts within an hour’s drive, with areas such as Dade County, Fla., in greater Miami, where there is only one public school district.
Not ‘Persuasive’
The Teachers College researchers used data from each of the studies to estimate what would happen if the degree of educational competitiveness increased by one standard deviation. To get an idea of what a change on that scale means, said Clive R. Belfield, the study’s lead author, imagine a neighborhood in which an average of one in 10 families sends a child to private school. To bring about a change on the order of one standard deviation, he said, the ratio of families choosing private over public schools would have to increase to one in seven.
The largest group of studies—26 in all—examined links between more schooling options and test scores. Among that group, the researchers found that a one-standard-deviation increase in competitiveness correlated with a modest one-tenth of one standard deviation improvement in test scores for public school students.
In comparison, other studies show that reducing average class sizes from 25 students to 15 can bump up test scores by about one-quarter of a standard deviation.
“The results are less persuasive than some of the advocates [of school choice] have tried to maintain,” said Mr. Belfield, who is also the center’s research director. “If you’re listening to that side of the debate, you would expect it to come out more strongly positive.”
Mr. Levin also said the benefits were not consistent from study to study or within studies. Test scores might improve in one grade or one subject, for example, but not in another. Of the 150 achievement-related results the researchers examined across the studies, between one-third and two- thirds showed statistically significant improvements.
Similar patterns emerged for other outcomes studied. A one-standard-deviation increase in educational competitiveness might, for example, be associated with a $400 to $1,000 increase in average annual teachers’ salaries, according to the estimates.
The studies also suggested that public schools in more competitive markets tended to be more efficient. A one-standard-deviation rise in competitiveness was linked, for example, to an estimated 12 percent spending reduction for states or districts.
All of the studies were adjusted to take into account socioeconomic differences that also might explain some of the benefits.
But the report warns against reading too much into its findings, since some of them may essentially be the same benefits. Schools that spend less and get better test scores, for example, are presumably more efficient.
Paul E. Teske, a political science professor at the State University of New York at Stony Brook who has also reviewed the research literature on school choice, said the results seemed to be on target.
“Advocates of choice would probably say, ‘But if you really freed up choices for people, you might see some real differences,’ ” he said. “I’m not saying that I believe that, but ... it can be a half-full or half-empty glass depending on where you’re coming from.”
