If you want to have an idea of whether a student will graduate from high school or enroll in a four-year college, look at the socioeconomic status of the student’s classmates. New research in the American Educational Research Journal finds that the socioeconomic status of peers matters more than any other school factors in predicting educational outcomes.

Researcher Gregory J. Palardy from the University of California found that while a student’s individual socioeconomic status, measured by such factors as family income and parental education level, is the strongest predictor of educational outcomes, the socioeconomic status of classmates is among the next highest predictors. The author suggests that this finding is significant at a time when schools are increasingly segregated along socioeconomic lines.

Palardy used a national data set on more than 10,000 high school students which shows that 40 percent of students at affluent schools report having a friend who has dropped out of school compared to 61 percent at schools in poorer communities.

Similarly, 57 percent of students at affluent schools report having friends with plans for college. At schools in poorer communities, the figure is 40 percent.

To see whether it is the greater resources available at affluent schools that make the difference, Palardy used controls to isolate the impact of school resources. He found that regardless of the level of resources, the graduation rate at schools at the higher socioeconomic end remained at 93 percent versus 76 percent at the schools at the lower end. College enrollment rates were also unaffected by school resources, remaining at 76 percent for affluent schools and 28 percent for schools in poorer communities.

In contrast, when controls were used to isolate peer influences, the difference on graduation rates between schools on the opposite ends of the socioeconomic spectrum narrowed from 17 percentage points to 2. Peer influence also narrowed the difference in four-year college enrollment rates from 48 percentage points to 19.

Palardy suggests that policymakers explore ways to integrate schools by socioeconomic status in order to equalize outcomes between schools. In cases where neighborhood boundaries hinder integration, state and federal lawmakers should provide incentives, he says.