In this report, the authors use administrative data from California to estimate the impact of suspensions on high school graduation rates, as well as the broader social costs of suspension.
According to the authors, the graduation rate for suspended students in California is 60 percent versus 83 percent for non-suspended students. However, as the authors note, much of this difference is likely explained by factors other than suspension. After controlling for several such factors, including GPA and low socio-economic status, the authors estimate that being suspended in high school reduces a student’s odds of graduating by 6.5 percentage points.
Unfortunately, like many similar estimates from previous suspension studies, this one most likely suffers from “omitted variable bias” insofar as the authors are unable to control for all of the factors that might make students both more likely to be suspended and less likely to graduate. And as in many of those prior studies, at least one omitted variable is fairly obvious: the same behavioral challenges (like poor impulse control) that make students more likely to misbehave might also make them less likely to graduate. Yet scholars lack access to data that tells them which students struggle with such challenges.
This is a well-known problem. So it’s unfortunate that the authors chose not to acknowledge it in the paper.
Worse, even though their estimate of the impact of suspensions is probably inflated by unobserved differences between kids, they nevertheless use it to estimate the economic costs of suspension.
According to the authors, “a single non-graduate generates $175,120 in fiscal taxpayer losses and $579,820 in social losses over his or her lifetime.” Thus, since their estimates imply that 4,621 students a year fail to graduate from high school as a result of suspensions, they conclude that suspensions in California “result in fiscal losses to taxpayers of $809 million over their working lifetimes (from age 18 to 65) and social losses of $2.679 billion.” These numbers are, at best, speculative. Moreover, since (as the authors admit) the report doesn’t take outcomes for non-suspended peers into account, even if they were credible they would only tell part of the story.
Needless to say, all of this is pretty frustrating if you’re a “no excuses” aficionado (or just a lonely suspension skeptic in search of a fact-based discussion, but I digress). While it is true, as the authors observe, that there is “no research-based justification for the frequent use of suspensions,” the case against them is weaker than advocates have led themselves to believe.
SOURCE: Russell W. Rumberger and Daniel J. Losen, “The Hidden Cost of California's Harsh School Discipline,” Civil Rights Project (March 2017).
