One of the things that has most impressed people about the work of social scientists John Chubb and Terry Moe is their scientific expertise. Instead of just asserting that children would get a better education if we sent them to private schools, their book "Politics, Markets and America's Schools" offers many pages of data analysis to "prove" their point.
But real scientists follow the evidence wherever it leads. When they find data that contradict or raise questions about their theories, they acknowledge this fact and modify their theories to be consistent with the new evidence. If we look closely at Chubb and Moe's work--at what they gloss over or omit--it turns out that it is more casuistry than science. They do not develop a theory about school performance based on evidence. Instead, they selectively muster whatever evidence supports their preconceived conclusions about vouchers and
"market schools."
It is often assumed that kids get a better education in private or parochial schools than in public schools. This is mere assumption since private and parochial schools seldom release test scores--and do not necessarily use the same tests as public schools. Recently, however, we were able to make some real comparisons when the National Assessment of Educational Progress (NAEP) released the results of its 1990 math exam. NAEP found that math achievement scores in private and parochial schools were very close to those in public schools, particularly for 17-year-olds.
Indeed, when you take into account the differences between private and public school students, it is amazing how close the scores were. Private and parochial schools can reject whomever they please, the majority require some sort of examination to get in, and they generally don't keep you if you don't do your homework or if you turn out to be a discipline problem. There are very few handicapped youngsters in private and parochial schools, and the parents of private school students are much less likely to be high school dropouts and much more likely to have graduated from college than public school parents.
In a Wall Street Journal article (July, 26, 1991 ), Chubb and Moe discounted the closeness of the NAEP scores. One of the major reasons they offered was the fact that a large number of kids leave private school after eighth grade and go to public high school. This, they said, tends to elevate public school scores.
More recently, at a conference I was attending, Chubb presented a new paper that he and Terry Moe have written. It cites data from the National Education Longitudinal Study of the Eighth Grade Class of 1988 (NELS:88), which, according to Chubb and Moe, supports their contention about private school achievement. NELS:88 tested a large group of students in eighth grade and again in tenth grade and found that the youngsters in private schools had made much greater gains than public school students.
However, this turns out to be only half the story. A whopping 37 percent of students left parochial school for public school after eighth grade. Who were these kids? Nearly half--47 percent-- came from families in the lowest 25 percent in terms of socioeconomic status. Since socioeconomic status (SES) correlates with student achievement, the large number of low SES students leaving parochial school in eighth grade was certain to raise parochial school scores. And the large number of these kids entering public school was almost certain to depress public school scores--exactly the opposite result from the one Chubb and Moe claimed in the Wall Street Journal article.
In their recent paper, Chubb and Moe use NELS:88 data showing greater improvement in private than public school scores between eighth and tenth grade to assert the superiority of private schools. Impossible to miss but totally left out of their discussion was that the transfer of low SES students certainly elevated private school scores and almost certainly lowered public school scores. It's as though they were comparing the health records of a private and public hospital and claiming the private hospital had a better record even though they knew that the private hospital screened its patients in the first place and then sent the ones who had the most serious diseases (and couldn't pay) to the public hospital.
What Chubb and Moe are doing here is clearly not social science. It represents a selective mustering of the evidence to support a conclusion they have already reached and a suppressing or ignoring of evidence that contradicts it. Chubb and Moe are entitled to their opinions, but it's time for them to drop their pretense of objectivity and admit that, when they talk about market schools, they are speaking as partisans for vouchers and not as disinterested social scientists.