Do Newer Discipline Practices Erode School Safety?

Max Eden of the Manhattan Institute think-tank has written something of an exposé for The74Million, a website devoted to news about American public schools. Eden writes that he had been doing research into “the unintended consequences and dangers of discipline reform” in public schools when he read about a knife attack at a public high school in New York. The details of the attack are gruesome: it was allegedly committed by a student who said he was being bullied; it occurred during class; it critically injured a 16-year-old student; and it killed his 15-year old friend. After interviewing current and former students and teachers at the school, Urban Assembly School for Wildlife Conservation, Eden concludes that it was “a once safe and supportive school that fell into chaos as new administrators implemented a supposedly more positive approach to school discipline.”

A NY Times article posted two days after the tragedy tentatively places the blame on frequent leadership changes at the school and on neighborhood conflicts that “seeped into the school building.” But Eden spent several months investigating the story and has a different perspective. Quoting from interviews and describing a rapid loss of disciplinary control over the course of three school years, Eden presents evidence that New York City’s misguided discipline policies deserve much of the blame for the student’s death. Former teachers and students told him that the school was a very safe and peaceful place until around the time Bill DeBlasio was elected Mayor. Eden believes that at DeBlasio’s direction, the use of suspensions for major disciplinary situations was severely restricted, and layers of bureaucracy replaced a straightforward discipline process whose authority students understood and respected.

“Under de Blasio,” Eden writes, “the NYC Department of Education began requiring that teachers provide full documentation of a range of nonpunitive interventions to at least three mid-level infractions before asking their principal to issue a suspension. Principals, in turn, had to apply in writing to a central office that often rejected their requests.” According to Eden, the additional layers were worse than ineffective. He documents multiple instances in which logs of major discipline violations and even violent acts at the school were not only never submitted to the district, but were destroyed, noting that superintendents who insisted that schools reduce suspensions did not seem to care that the result was near-anarchy in some schools.

For his part, DeBlasio was responding to some of the more recently developed, but widely adopted, approaches to school discipline. These “restorative” approaches seek to bring equity to discipline policies, with a goal of mitigating the “school-to-prison pipeline” that disproportionately afflicts minority communities. Some of the UA Wildlife teachers and former teachers who were interviewed for Eden’s story described how Positive Behavioral Intervention and Supports (PBIS) and Restorative Practices were implemented at the school over the past three years as replacements to more traditional disciplinary measures and accountability. Eden believes that these practices are ineffective deterrents to major disciplinary infractions.

Preferring rewards for good behavior over punishments for infractions, the PBIS framework is described as “a multi-tiered approach to social, emotional and behavior support.” It seeks to integrate social and emotional learning (SEL) with discipline. A research paper from 2014 describes the main components of a PBIS implementation as follows:

(a) clearly defining behavioral expectations valued by the school community, (b) proactively teaching what those expected behaviors look like in various school settings at least once a year, (c) frequently rewarding students who comply with behavioral expectations, (d) administering a clearly defined continuum of consequences for behavioral violations, and (e) continuously collecting and analyzing data to assess students’ responsiveness to the behavioral support provided.

Nationwide, PBIS is becoming more widely implemented each year. A report from the Texas Juvenile Justice Dept. in 2010 estimated that more than 14,000 US public schools had already adopted the framework. (PBIS is considered a “framework” rather than a “curriculum” or “program” because it consists of loose guidelines that should be customized for each school.)

According to the teachers whom Eden interviewed, in addition to PBIS, UA Wildlife deployed “restorative justice” practices as part of its attempted school-discipline reforms. In the 2016 US Dept. of Education School Survey on Crime and Safety, which was administered to a representative, random sample of 3,553 public schools near the end of the 2015-16 school year, 33.5% of respondents said that their school used “restorative circles” as a discipline strategy (which was defined for respondents as “a formal mediation process led by a facilitator that brings affected parties of a problem together to explore what happened, reflect on their roles, find a solution, and ultimately restore harmony to individual relationships”).

Another US DOE publication focusing on school discipline noted that by the spring of 2015, 22 states and Washington, DC had used legislation to reduce suspensions in public schools, to deploy alternatives to traditional disciplinary practices, and to provide more support aimed at improving student behavior.

A lack of accountability, a power vacuum, and a gap in discipline undoubtedly contributed to the disaster at UA Wildlife, but Eden also faults PBIS and restorative practices. Quite possibly the implementation of PBIS at UA Wildlife in New York was fatally flawed. Eden describes it as “offering students tickets redeemable for prizes in exchange for good behavior” and quotes a teacher as saying that it “failed” because “students did not value” the tickets or prizes. It’s worth questioning whether tickets and prizes are appropriate incentives at the high-school level at all.

The collection and analysis of data is always included in descriptions of the core strategies that compose the PBIS framework, yet Eden cites evidence that two principals at UA Wildlife either stopped logging incidents or actively destroyed existing discipline records.

So, what were they thinking?

Research has indicated that PBIS, which is often referred to as “School-Wide Positive Behavior Supports,” or SWPBS, can be effective. A longitudinal study was conducted at 37 elementary schools over five years and “found that implementation of an SWPBS model resulted in fewer student suspensions and office discipline referrals.” That study was cited in a 2014 research paper in Research in Middle Level Education whose findings, after studying a single middle school, included the following findings:

…infractions for problem behavior declined an average of 40% over a two-year period. Reductions were seen in all eight of the most frequently occurring infractions, including fighting, insubordination, class disruption, inappropriate behavior, skipping detention, cutting class, tardiness, and disrespect to staff. By counting both the number of infractions as well as the number of office discipline referrals, the conclusion can be drawn that the decline in office discipline referrals was not the result of staff reluctance to refer and suspend. With the exception of cutting class, the greatest improvement was seen in lower level infractions that are most responsive to … interventions.

Similarly, restorative practices have been shown to be effective by education research. For example, a study from researcher Anne Gregory of Rutgers found that “Classrooms with a high level of restorative practices implementation (High RP) had fewer disciplinary referrals for defiance and misconduct compared to classrooms with a low level of implementation (Low RP), over the 2011–2012 school year.”

Eden exhibits several signs of bias against what he elsewhere refers to as “leftist ideology” in school discipline policies. He has conducted additional research into the effects of newer policies on school climate and safety. Expressing concern about survey results from districts in multiple regions of the country indicating that teachers feel less safe at school now than they did a few years ago, he blames “top-down mandates to reduce disciplinary statistics” that prompt principals and superintendents to underreport disciplinary incidents and to fail to respond appropriately to potentially dangerous student behavior.

In his analysis of what went wrong at UA Wildlife, Eden goes farther up the chain to place blame at the door of Arne Duncan and the Obama administration:

But it’s not just New York City or New York state. The Obama administration’s 2014 “Dear Colleague” letter on suspensions, motivated by federal Department of Education data showing that black students were much more likely to be suspended and expelled than white students, applied the pressure experienced by UA Wildlife on schools across America: Get your discipline numbers down, or else. This created incentives to produce ever-lower suspension numbers, to prioritize statistics over student safety.

Eden mentions one other instance in which school administrators systematically concealed or destroyed data about negative behavior and its consequences: Washington, DC public schools (DCPS). He cites an article from the Washington Post that details an alleged practice of concealing or mischaracterizing suspensions in an attempt to reduce them (by 40% in two years). The NYC and DC public school systems might be outliers in this regard. However, the incentives to reduce suspensions in response to the Obama-era guidance must be seen to cast doubt on the validity of all US public school systems’ statistics that show recent reductions in suspension rates. It’s entirely possible that those systems are simply underreporting suspensions, miscategorizing them, or not assessing suspensions for significant infractions, jeopardizing school safety.

Although Eden is alarmed by the possibility that schools are less safe under the federal guidance to reduce suspensions, the Washington Post’s investigation did not support that conclusion. Their research included evidence from three high schools in January 2017 and from those same high schools, plus three more, in January 2016. (They requested data for those two months from all DC high schools and were told that most schools could not provide it.) The data indicated that those schools were not under-suspending students but were instead grossly underreporting suspensions. Rather than receiving formal, documented suspensions, many students at those schools were simply told not to come to school after a behavioral incident. When data from all of the seven schools that provided records from January 2016 was aggregated, only 15% of the total days that students were de facto suspended were reported through official channels.

Have newer discipline practices compromised school safety in Wake County? We’ll consider that question in our next post.