There’s a growing consensus that suspending students is extremely problematic. The American Education Research Association sums up the research and critiques with two succinct tweets (h/t Keith Benson, who pointed me towards these): 

Researchers have found that being suspended or expelled tripled the risk of involvement with juvenile justice. #AERALectures

— AERA (@AERA_EdResearch) April 12, 2017

Black students are currently 3.5x more likely than white students to be suspended out of school. #AERALectures

— AERA (@AERA_EdResearch) April 12, 2017

The combination of these two factors is devastating. On one hand, we have more and more evidence that suspensions wreak havoc on students. This makes sense — taking students who are already struggling out of the classroom contributes to those students falling even further behind their peers. There is even new research that argues suspensions to black and brown students may be contributing a significant amount to the achievement gap.

So when the New Jersey Department of Education School Performance Reports came out this week, I quickly flipped to the page on school culture to get a sense of what Camden schools were suspending the most students. 

And, wow. I knew that black and brown districts saw heavier suspensions (see Mark Weber’s tweet below) but some of the numbers in Camden were almost unfathomable. The highest suspension rate in Camden was Freedom Prep’s 78.1% suspension rate (that’s 78.1% of students that were suspended at least once in the 2015-2016 academic year).  

@SteveDanley @Okaikor They aren’t alone. Important to note this would NEVER be tolerated in suburban schools. https://t.co/5Yfd1cBbwM & https://t.co/oG9Wo7hoCz pic.twitter.com/ufpTw9qEAq

— Jersey Jazzman (@jerseyjazzman) April 12, 2017

Now, there are some caveats to this data. 1) it is self-reported. That means we should keep in mind Campbell’s Law — the idea that as the stakes go up for a social indicator, it is more likely to be manipulated. There are certainly schools in Camden that are downplaying their suspensions while finding other similarly problematic ways to discipline students — plenty of qualitative data points in the city point to troubling practices at the Uncommon’s Camden Prep (for example) despite a low suspension rate (3.2%). Just because your suspension rate is low doesn’t mean you’re a good actor. But some of this self-reported data is unbelievably incriminating, and clearly points to schools abusing their power over students.

2) schools that serve more challenging populations of students typically have more suspensions. So I wanted to put these suspension rates in context. We know that suspension rates are closely correlated to the population being served in a school, so I sliced the data so that we could compare schools with similar populations. Below is a quick data-dive into that question. It’s a scatter plot with percentage of students suspended on the Y axis, and percentage of special needs students on the X axis (red dots are charter/renaissance schools and black dots are Camden School District schools). 

Now, this is a simplification of what’s happening, and we’ll get to the limitations of such a chart in a moment. But the trend line here gets to the heart of the issue: suspensions go up as schools serve more challenging populations. The trend line is the best estimate for a very simple model of how many students we should expect a school to suspend based upon the number of special education students (which here is a proxy for serving a more challenging populations — it’s not a moral statement that special ed students should be suspended more). 

The simplest way to read this chart is that the further above the trend line a school is, the more students it is suspending based upon the nature of its student population. So, even though Hatch and Forest Hill have relatively high suspension rates, they serve a challenging population (again, we can ask questions about whether we should be suspending students at schools with more challenging rates — but based on the data, we do). Hatch and Forest Hill actually suspend fewer students than we would expect based on their student body. Schools such as Camden Promise and Camden Academy Charter are suspending roughly the same number of students as Hatch and Forest Hill, but with a much less challenging student base. Compared to other schools serving similar populations, they suspend at a high level. 

A few outliers — I think it’s probably fair to call them bad actors — pop off the page here: Brimm Medical Arts (27.9% suspension rate, despite only 3% special education students), KIPP-Cooper Norcross (34.2% suspension rate, despite only 15% special education students) and Freedom Prep (with a whopping 78.1% suspension rate, despite only 16% special education students). That is my wall of shame — schools that are suspending students at much higher rates than comparative Camden schools with similar student bodies. And doing so despite very public and clear evidence that it is detrimental to students.

Each of them deserves more scrutiny. Brimm Medical Arts is a magnet school in Camden but has far more suspensions than its sister schools (Creative Arts: 6.4%, Met East: 7.4%). Schools that both serve a comparatively less challenging population and suspend students need particular attention from administrators. The good news here is that because Brimm is magnet school, the district can put pressure on it to reduce suspensions — an issue the district has been progressive on and one that it deserves credit for. 

Camden High School and Woodrow Wilson — Camden’s two remaining traditional high schools — are much more difficult cases. Camden High has exceptionally high suspension rates (53.1%) but serves a very difficult population (37% special ed). I didn’t include Woodrow Wilson, with similarly high suspension rates (55.3%), on the list of “bad actors” because I imagine a more complex model would show they have a similarly difficult population. For example, though they have lower special ed numbers (24%) they have very high proportion of English Language Learners (20%). To top that off, older kids appear to be more likely to be suspended, and a model that incorporates age of students would probably bring Wilson back into the realm of what’s expected given its population. That said, schools with suspension rates over 50% are still troublingly high. I’d like to see efforts to bring them down. But I also acknowledge that a system which creams the best students to magnet, Renaissance, and charter schools in their later years puts schools such as Camden High and Woodrow Wilson in a difficult position. 

This issue of age of students makes the KIPP numbers puzzling. When KIPP graciously hosted me for a tour, they were quick to point out their music room, new playground, and other facilities that showed they were trying to pivot at least partially from a No Excuses model with an out-dated model of extremely harsh discipline for black and brown students. The data shows how much of a struggle that’s been. KIPP has challenged the idea that they serve fewer special education students than other schools — arguing (in the past) that young students are less likely to be diagnosed as special education and data needs to incorporate age and that they were actually serving a representative population. While that’s worth noting, in this case the same should be true of suspensions — there should be fewer suspension at younger ages. But there aren’t at KIPP. Quite simply, KIPP is claiming to be a progressive choice in Camden while suspending an extremely high number of students (34.2%, the second highest number for a non-high school and more than twice as many students as its peer institutions in the district). Worse still, there is very little anyone outside of KIPP can do about it — one of the downsides to outsourcing education in the city to nonprofits with little oversight beyond test scores.

And finally, we get to Freedom Prep. Frankly, these numbers (self-reported) are so heinous that I reached out to numerous folks within Freedom Prep to try to get an explanation (they also offered to meet with me — something I hope to do. I’ll make sure to do a follow-up post if warranted). What I received from the school made little sense and confirmed that the school is pursuing an education strategy that suspends students at high rates. The school claimed that 1) the state numbers did not match the school’s internal numbers (despite these suspension rates being self-reported) and 2) that Freedom Prep likely suspends students at the highest rate in the city and that 3) they, like any school, are trying to lower the number of suspensions. Given that Freedom Prep’s student population is in the middle of the pack on every measure of how challenging its students are, the 78.1% suspension rate is extremely upsetting, particularly compared to peer schools which suspend at much lower rates. In Freedom Prep, we have a Camden school that, despite research that shows students who are suspended are more likely to drop-out and that suspensions contributes to further struggles in school, are suspending virtually every student (78.1%). It’s an ugly, calloused and out-dated model of schooling — the definition of a bad actor. Frankly, there is “no excuse” for an organization that claims to be data-driven to use this model. We know what it leads to and how it harms students.