A review of documents and interviews reveals a lack of record keeping that reflects a system struggling to comply with state law and reckon with the use of physical restraint
On Jan 24, 2017, at David Ellis School in Roxbury, a teacher physically restrained an emotionally impaired and distraught 9-year-old boy.
There is no consensus or complete account of what happened. Was the boy alone with the teacher, or in class? How was his neck bruised, or was it scratched? “Choke-slammed,” claimed the mother. But the teacher told another story. (Both requested anonymity to protect their privacy.)
In class, the boy raced about the room screaming. After de-escalation tactics failed, the teacher employed a basket hold, wrapping arms around the boy, who “was wiggling around,” he said. Then, “we both fell to the floor.”
The Department of Children and Families (DCF) investigated, which does not mean the allegation of “choke-slamming” is true. The mother has “no clue,” she said, what DCF or BPS, who both declined to comment, discovered. “There’s not much evidence,” noted Elizabeth McIntyre, the boy’s lawyer.
A public information request filed by this reporter showed there was no record of the boy’s incident in the Boston Public Schools data either. But, the teacher claimed, “I reported it.”
The absence of such records reflects a system struggling to comply with state law and reckon with—for students, parents, and teachers—the use of physical restraint.
In BPS, the larger story of physical student restraint—like the David Ellis boy’s story—remains incomplete, concealed behind a dysfunctional bureaucracy that hides trouble spots and curbs accountability, increasing risk of student mistreatment.
In BPS, physical restraint—defined as certain holds that restrict movement—is an emergency procedure of last resort permitted only to protect students or staff from imminent and serious harm.
“When this goes down,” explained Zachary Houston, BPS director of applied behavior analysis, “it’s a manifestation of the death of the bond between teacher and student.”
That bond perished 1,106 times from 2017 to 2019, according to BPS records, which include brief narrative reports, the hold employed, and a dozen other fields in an Excel sheet. Altogether, the data suggests untrained staffers have restrained students without the required threat of harm.
Still, no one, including lawyers, advocates, and BPS administrators, believe all incidents are fully and accurately reported.
“There’s no data integrity here,” McIntyre noted. “Restraints have a serious risk of harm and can cause trauma,” explained Liza Hirsch, Senior Attorney at Massachusetts Advocates for Children. Without it, Hirsch added, “We don’t know what’s really happening with kids.”
State and federal agencies, especially the US Department of Education Office for Civil Rights, require reliable data to enforce laws that govern restraint and prohibit discrimination.
In March 2021, 17 attorneys general, including Massachusetts AG Maura Healey, signed a letter urging Congress to pass the Keeping All Students Safe Act, a federal bill seeking to establish “a long-needed system of both accountability and support,” in which “children can learn without fear of abuse or discrimination.”
In the 2019-2020 school year, BPS closed to prevent COVID-19 from spreading, and the issue of student restraint went dormant. A year later, physically detaching students and teachers has extracted a profound, ongoing emotional and educational cost, especially for disadvantaged students with disabilities. Now, students are returning to classrooms.
“There are going to be emergencies,” explained Ron Hager, managing attorney for the National Disability Rights Network. “There needs to be some physical control over the environment.”
At the same time, Hager said, “We want to seriously restrict restraint,” which threatens to break the bonds with teachers that students have deeply missed and are trying to rebuild.
Flawed data and floor holds
In 2014, Massachusetts had a substandard restraint policy compared to voluntary federal guidelines, a ProPublica review of federal data found. Then, after public criticism, the state limited floor holds—lying face up (supine) and face down (prone)—that can asphyxiate a child, which is why some limit it.
Meanwhile, in California, a 13-year-old autistic student died after being restrained face down for an hour and 45 minutes, the Sacramento Bee reported. Some schools that serve disabled students have even stopped using restraints, though at the national level, in 2020, an attempt to pass a federal bill regulating restraint failed.
While Massachusetts also banned mechanical holds (handcuffs and straps), BPS data show two incidents at Excel High School of a student being handcuffed in 2017. A student “threatened to throw soda” at a school police officer and was handcuffed. In another incident, a student “pursuing another student to fight” was “non-compliant” to “verbal redirection” and was “handcuffed until her mother arrived.”
Law enforcement, including school security, are not subject to the same restraint requirements as public school officials. Still, why was it necessary to handcuff a child who threw soda, and another who chased another student, but had not hit them? The data has no explanation. BPS declined to comment.
The Massachusetts reforms also required improved data collection, a critical tool, James DiTullio, the former Mass education undersecretary, told ProPublica in 2015. “Good data leads to better outcomes,” DiTullio added. About the same time, Stan Eichner, then a Disability Law Center litigator, expressed concern that disabled students—those most often restrained—were “severely mistreated.”
In 2017, the US Department of Education began reviewing the quality of school district restraint data. And in 2018, Jim Major, head of the Massachusetts Association of 766 Approved Private Schools, whose member schools receive underserved students from public schools, noted, “We need more and better data.”
Despite reform, that concern endured. “Most special education advocates have ongoing concerns about the accuracy of restraint reporting data,” explained Rick Glassman, director of advocacy for the Disability Law Center. In April 2020, the Government Accountability Office, a congressional watchdog agency, noted large districts like BPS likely underreport data.
All school districts are required to report student restraint data to the Department of Education Office for Civil Rights. But that data is outdated—the most recent from 2017-2018—and the department inadequately tracks it, according to the GAO. The Massachusetts Department of Education has more recent data, but it contains more categories of data than BPS records. “Data helps target areas that need more support,” for teachers and students, Hager explained. “We want to know what the problem is, where it is, and the scope of it.”
While BPS’ data is flawed, it reveals some trends from the 2016-2019 school years. BPS, like other districts in the state, employs four standard holds—standing, seated, floor supine, and floor prone. The standing hold was employed most often (57%), then the seated hold (8%). About 14 BPS schools employed a prone hold (about 1%).
To employ a prone hold, certain conditions must be met, including a documented history of injury to self or others. (State law prohibits its use to protect property or discipline a student.) Still, “prone restraints present an unacceptable risk of serious harm or death,” Glassman said. Many schools logged an “other” hold (33%), classified as brief contact—like escorting a student—which may not qualify as a restraint, according to Mass guidelines.
From 2017 to 2019, 86% of restrained students had an individualized education program, a legal document that outlines services for a disability. From 2018-2019, only about 9% of incident responders were trained in de-escalation and physical restraint, while in 18% of the incidents at least one of the three responders was trained.
Andria Amador, senior director of Behavioral Health Services at BPS, who entered that data, explained it is a “crude” number. To make it, she manually cross-checked multiple lists of those trained against incomplete incident reports.
BPS, which has about 10,955 full-time staffers, has taught 1,361 employees in safety care, a train-the-trainer model that teaches prevention strategies and restraint in classes from one to three days. But restraint protocols are complicated, and not every staffer is well-trained, explained Sophia Johansson, northeast coordinator of MassFamilies, a disability advocacy group. For example, staffers may not accurately define an action as a restraint and never report it, which helps explain why advocates argue restraints are vastly underreported.
Recently, BPS claimed a technical issue prevented it from reporting restraints to the state’s Department of Elementary and Secondary Education (DESE), which had updated its online reporting system to allow real-time entry, instead of a one-time, year-end bulk data dump. Amador claimed the update initially violated the federal Family Educational Rights and Privacy Act (FERPA), preventing data entry for the 2019-2020 school year. DESE disagrees. “Our reporting system never violated FERPA,” spokesperson Jacqueline Reis wrote in an email.
Either way, BPS data exhibits a list of unknowns that fog a detailed accounting. The data is self-reported, and there is little quality control. Some Boston schools never explain why a hold was applied or who administered it. Others omit its duration. Sometimes, oddly, a hold is reported to have been administered at one o’clock in the morning. Principals are supposed to review restraint data to identify trends, according to BPS.
The data does not identify a student’s race. However, in the last six years, McIntyre estimated, her firm, Greater Boston Legal Services, had represented hundreds of students on school issues, including discipline and restraint. Three have been white. In 2015-2016, Black students were 200%, and Hispanic students 45% more likely to be restrained than white students, a recent analysis of federal restraint data found, which is consistent with past research.
“We know African Americans with disabilities are restrained the most,” Hager explained. That disparity increases the risk for emotional and physical harm, and being ostracized from the classroom.
BPS declined to say if it keeps data on race and restraints.
Nationally, in 2004, Congress passed the Individuals with Disabilities Education Act to provide children with disabilities a right to a free and appropriate public education. But, according to a GAO review of 2015-2016 national data, 70% of school districts suspiciously reported zero incidents of restraint or seclusion.
Rights rely on accurate reporting, which is in part why “there is still such concern on the overuse and restraint around the country,” Hager noted. The 9-year-old boy from David Ellis was missing from the BPS data.
How many others are there? No one knows.
“Dragged down the hall”
In Massachusetts, BPS is the largest and most diverse district, serving more disabled, economically disadvantaged, and high-needs students than other districts. Within BPS, 34 out of 117 schools and their approximately 17,000 students, mostly from long-neglected groups, scored in the “lowest performing 10 percent of schools in the state,” a 2020 review found. That review also found “systemic disarray” of BPS special education services. It did not examine restraints. Neither BPS and DESE count restraint violations.
“We acknowledge there may be violations in reporting, but have no records as such,” explained Xavier Andrews, BPS director of communications. “I’m not aware of anyone who has a tally of ‘restraint violations,’” DESE spokesperson Reis noted. Instead, it is a bureaucracy where stories are told against each other.
In December 2018, in a Boston public school, a staffer “dragged” an 8-year-old “down the hall,” violating a series of restraint policies, according to McIntyre and records from the state’s Problem Resolution System, which investigates allegations of misconduct. A therapeutic mentor saw the bruise and reported child abuse—which the school denied—to the state Department of Children and Families.
The school argued the student was not restrained, but escorted after attempting “to strike his teacher.” The school was partially wrong, DESE found. The student, crying, had spun their arms in windmill circles, “openly in the act of assaulting his teacher,” and was restrained. The school also violated policy when it failed to properly notify the parent. However, DESE concluded there was insufficient evidence to determine if the student was improperly restrained.
In 2019, Hirsch explained, DESE issued a letter of finding that BPS was not fully compliant with state regulation in reporting incident details to a parent. “Across the state, there definitely have been violations,” Hirsch explained. “Physical student restraint has been employed unlawfully as a form of control and punishment, causing great harm to students,” she added.
In the data, hundreds of bare reports describe “upset,” “unsafe,” “agitated,” “aggressive,” distressed,” and “out of control” students. Sometimes, there are anecdotes of self-harm. In 2019, at UP Academy Holland, a kindergarten student ran outside alone—after a teacher took his marker—and tried “to bang his head on the concrete.” In 2020, at Warren-Prescott, a student, after finding the school nurse door closed, kicked in its glass and hid in a vacant bathroom. The student stood on a toilet, the stall door ajar, and banged their “forehead on the cinder block wall.”
Other reports exhibit a struggle to de-escalate. “Let me go,” a student yelled, while a staff member spoke “of calming body parts,” and keeping “all bodies safe.” Sometimes, there’s mild violence, a thrown shoe or open-hand hitting. From 2017 to 2019, there were 26 reported injuries to students or teachers, all minor, such as a split lip or bruising. Rarely, there’s a weapon.
The narratives also glimpse the staff perspective. “Throwing wall paper at me in my face,” a report notes. “Assault on staffer,” or, “Student not following directions and harming staff,” others explain. These moments, noted Blake Grider, BPS assistant director for applied behavior analysis, can “happen very quickly.” The rapid choice to take away a child’s autonomy—and assume that responsibility—can turn teachers into adversaries.
After a restraint, instructors may have a debriefing about what happened and why. “It’s exceedingly rare,” Houston noted, that a teacher requires formal therapy. Still, “It’s emotionally tough to deal with,” Grider added. “It’s hard for us to see them when they’re that upset,” especially if the student feels threatened.
“It’s not our intention to physically restrain students,” explained Kristin Clark, a special education teacher at Curley K-8. “It can break or damage a relationship.” Instead, de-escalation tactics are emphasized. And, after a restraint, students—when appropriate—are debriefed in part to make them feel valued.
“We all struggle with something,” Clark said. “Tomorrow is a new day.”
In 2015, the Disability Law Center unsuccessfully lobbied DESE and the Department of Early Education and Care to ban all prone restraints. Progress was made, explained Glassman. But, the current restraint protocol is “more permissive” than in adult disability agencies. Children are “more vulnerable to injury or death”—because of their smaller bodies—“yet the legal standard is less protective.”
“I cannot explain to you why that would make sense,” Glassman added.
About 11 days after the David Ellis incident, the boy began a 45-day evaluation at COMPASS, a therapeutic day school. He’s still there, rarely restrained, and happy. But a heart condition during the COVID-19 pandemic keeps him home with faulty internet.
The incident had triggered anxiety and panic. “My baby would fight with me,” his mother recalled. “Mommy, please don’t let me get on the bus.” For COMPASS, before the pandemic, he eagerly waited on the porch in the winter. “That made me happy,” the mother said.
An accounting of restraint can seem like an adversarial process. But, “It’s not us versus them,” Houston explained. Supporting teachers, staff, and students, said Hager, is the best way to reduce restraints. Still, for the 9-year-old boy’s mother, four years later, David Ellis remains a hostile, collective “them.”
“They treated him like some kind of animal,” she recalled.
This article was produced in collaboration with the Boston Institute for Nonprofit Journalism. To support reporting like this you can contribute at givetobinj.org.
Daniel DeFraia is journalist and American Studies PhD candidate at Boston University, where he’s writing a dissertation on the historical relationship between intelligence agencies and the US news media.