The So-Called School-to-Prison Pipeline: Here's My Take on the Browne-Hoge Study
In this video, Dr. Justin Baeder examines a Harvard/Boston University study on the school-to-prison pipeline and explains why the conclusions may not support the narrative often built around them.
Key Takeaways
- The study's conclusions deserve scrutiny - The Browne-Hoge study from Harvard and Boston University makes claims about suspension and incarceration that warrant careful examination
- Correlation isn't causation - Linking suspension to later incarceration doesn't prove that suspension caused the outcome
- Removing consequences creates different risks - The focus on the school-to-prison pipeline often ignores the school-to-hospital pipeline for victims of unchecked violence
Transcript
So is there any such thing as the school-to-prison pipeline?
I've been asking people to send me studies, and it turns out there is a Harvard, Boston University, and University of Colorado at Boulder study, which we'll take a look at now.
So this study was published in July 2021 in Education Next, and it's called Proving the School-to-Prison Pipeline, Stricter Middle Schools Raise the Risk of Adult Arrests.
And you can see the three authors here.
Now, one important thing to note about the authors is that they are all economists.
They study statistics.
They don't study how schools work.
They have never been teachers or principals.
And that introduces a kind of big methodological problem in this study.
And if you've read the book of Why, two big methodological problems jump out at you from this study.
The first has to do with the counterfactual.
And the counterfactual is the what if something was different question.
So for the school to prison pipeline, the counterfactual is what if we didn't suspend the student?
What if we just found something else like a restorative practices effort?
Or, you know, what if we just did something else instead of exclusionary discipline?
Would the student have better outcomes in terms of being less likely to get arrested?
This is a very important question, but the problem is this study is not actually looking into that question.
What this study is looking into is school strictness as measured by the number of suspensions that the school issues.
And if you're thinking, well, maybe schools suspend more students when they have more behavior problems maybe it's not just you know kind of a random how strict do we want to be maybe it's actually a function of some uh underlying difference in the rate of behavior problems in the school well that brings us to the second methodological problem with this study and that has to do with unobserved variables as the book of y makes very clear if you are not finding the right underlying variable if you are studying something that's basically a proxy for something else you're going to make incorrect conclusions about causality and that's exactly what we see in this study they're not actually measuring school strictness they're measuring just the suspension rate and there's something behind the suspension rate probably multiple things behind the suspension rate And what could have happened is they could have designed their study in such a way as to account for and tease out the multiple contributors to the suspension rate, one of which is principal strictness, but another of which they don't consider at all is student behavior.
So if we're going to consider the counterfactual of what if we didn't suspend this student, we can't just compare schools that suspend a lot of students to schools that don't suspend a lot of students because obviously there are differences in student behavior between those schools, not just in strictness.
So I would love to hear from you if you have other studies or other questions about this study.
But based on what I've seen here and based on my reading of the study and understanding of causality, I have to say this study falls pretty flat in demonstrating the reality of the school-to-prison pipeline.
Let me know what you think.