"Private religious schools reduce teen sexual activity, arrests, and cocaine use. Contrary to popular belief, private religious schools do not achieve these results by enrolling better-behaved students."
In the heated debate over the effectiveness of private schooling, most of the attention has focused on whether private schooling improves academic outcomes. In Sex, Drugs, and Catholic Schools: Private Schooling and Non-Market Adolescent Behaviors (NBER Working Paper No. 7990), David Figlio and Jens Ludwig examine the relationship between private religious schools and student behavior. They conclude that private religious schools reduce teen sexual activity, arrests, and cocaine use. Contrary to popular belief, private religious schools do not achieve these results by enrolling better-behaved students. In fact, the data suggest that poorly behaved children are more likely to be sent to private religious schools where they derive "substantial benefits" from attendance.
The authors use data from the National Education Longitudinal Survey (NELS) of 1988 to distinguish between the effects of family background and the effects of schooling. The NELS provides unusually detailed information about individual students, families, and individual schools. It includes rich measures of substance use, misbehavior, and sexual activity along with complete information on family structure, socioeconomic status, behavior while in school, religious affiliation, and academic achievement. In many other data sets, information on family background comes from student surveys. The NELS asked parents for those data.
The roughly 10,500 eighth graders that comprise the authors' sample were those in the NELS who were interviewed again in 1990 and 1992, who lived in a metropolitan area, and who attended either a public school or a private religious one. Most of the private religious schools were Catholic. Students attending private non-religious schools, roughly 2 percent of the total, were not included.
Unadjusted for family background, students in public and private religious schools had similar rates of smoking, drinking, gang membership, and marijuana use. Private religious schools had much lower rates of sexual activity, arrest, and cocaine use. The differences persisted even after family characteristics were taken into account. When the authors controlled for the possibility that parents likely to produce better-behaved children might also be more likely to enroll them in a private religious school, they found that parents were more likely to choose religious private schools for children at greatest risk for problem behavior.
At present, these results have decidedly murky implications for public policy, in part because they show that private religious schools affect different groups of students differently. Private religious schools appear to reduce teenage sexual activity among girls, but not boys. They reduce arrests, smoking, and cocaine use among boys but not girls. Finally, their beneficial effects are "concentrated among teens who live in households with two parents or guardians." They have no discernable effect on the behavior of students from single-parent households.
-- Linda Gorman