Every February, the Office of the State Comptroller publishes a document that most parents will never read and every school board should. This year's edition, a seven year review of New York's school safety data, arrives with two headlines sitting side by side. The first is alarming: bullying and drug incidents have climbed past pre pandemic levels, with bullying now accounting for well over half of everything schools report. The second sounds like relief: assaults, sexual offenses, and weapons possession have collapsed to near zero across the state. Read quickly, it looks like a wash, one bad trend and one good one, cancelling each other out. Read carefully, only one of those trends is actually a trend. The other is what happens when a state quietly rewrites what counts as an incident, then measures the world against the new definition instead of the old one.

That distinction is not a technicality. It determines whether a parent looking at a school district's safety numbers is seeing a genuine improvement in student behavior or a bureaucratic sleight of hand dressed up as good news. The comptroller's office, to its credit, says so plainly. This is worth sitting with for a moment, because it changes how every other number in the report should be read.

What Changed in 2021, and Why the Comparison Broke

New York has tracked violent and disruptive incidents in schools since 2001, first through a system called VADIR and later through a second track added under the Dignity for All Students Act, which is where bullying reporting originates. In school year 2017 to 18, the state merged both systems into a single report, the School Safety and Educational Climate report, known by the acronym SSEC. The merger produced an immediate problem: reported incidents in the assault and sexual offense categories spiked so sharply that state officials suspected the new form itself, not student behavior, was driving the numbers.

The fix arrived in school year 2021 to 22, in the form of revised reporting definitions for three categories: assault, sexual offense, and weapons possession. Under the new rules, an incident only counts if three conditions are all met. The person involved must be at least ten years old. The act must rise to the level of a felony under state law. And the school must have referred the matter to law enforcement. A fight between two ninth graders that the school handles internally, with no police involvement, is no longer reportable as an assault. A sexual touching incident a principal determines does not meet the felony threshold does not appear in the data at all. None of this means the underlying behavior stopped happening. It means the state stopped counting a version of it.

The Numbers Behind the Good News

Once that filter went into effect, the collapse was immediate and dramatic. Statewide, the assault rate in secondary schools fell from 4.9 incidents per 1,000 students in 2017 to 18, to 0.5 per 1,000 in 2023 to 24. Sexual offense incidents fell from 1.6 per 1,000 to 0.1 per 1,000 over the same stretch. Weapons possession followed the same arc. The comptroller's office describes all three serious violent incidents categories as having fallen to near zero statewide, and the pattern holds across every region of the state, upstate, downstate, and New York City alike.

None of that is fabricated. The incidents that still clear the new, higher bar really have gotten rarer in the official record. But a graph that drops off a cliff exactly the year the definition changed is not describing a behavioral shift so much as a measurement one, and the report's own authors flag this explicitly rather than let the improvement stand unexamined. For a policymaker or a parent trying to judge whether a specific school or district has actually gotten safer since the pandemic, the honest answer is that this particular set of numbers cannot tell you.

Bullying: The Category That Did Not Get Redefined

Here is where the report's second headline earns its alarm. The definitions for bullying and cyberbullying have not changed since New York began collecting this data in 2017 to 18. Whatever movement shows up in these categories reflects reported behavior, not a shifting yardstick, which makes bullying the most reliable trend line in the entire document.

And the trend line is not good. Bullying is now the single most common incident category in New York schools by a wide margin, making up 61.8 percent of everything reported statewide in 2023 to 24, a total of 29,718 reported bullying incidents. Measured only among schools that reported at least one incident, the statewide rate climbed from 10.2 per 1,000 students in 2017 to 18 to 18.2 per 1,000 in 2023 to 24, a rise of roughly 78 percent.

The regional picture is where the story sharpens into something closer to a reversal. For years, bullying rates in New York City schools tracked below the rest of the state, and upstate districts consistently reported the highest rates in New York. That pattern held through the pandemic, when the city's reported bullying rate collapsed to 3.1 per 1,000 during remote learning. Then, as students returned to classrooms, New York City's numbers did something upstate's did not: they blew past every pre pandemic benchmark. The city's bullying rate rose from 10.5 per 1,000 in 2019 to 20, to 25.8 per 1,000 in 2023 to 24, an increase of roughly 146 percent. Upstate districts, meanwhile, remain higher than downstate on an absolute basis but have not yet climbed back above their own 2017 to 18 peak of 14.5 per 1,000. The comptroller's office links part of the city's surge to a 2021 revision of the Chancellor's regulations that broadened the examples and definitions used to identify bullying and harassment, particularly for online conduct, though the timing alone does not prove causation.

"It's disturbing to see the rise in bullying and drug-related incidents," Comptroller Thomas DiNapoli said, adding that the findings point to more work ahead, both inside classrooms and beyond them, to help students and teachers feel safe.

Cyberbullying's Deceptively Small Numbers

Statewide, reported cyberbullying rates look almost reassuring on their face, just 4,396 incidents in 2023 to 24, a rate of 1.8 per 1,000 students. But the comptroller's office is upfront about why that number is probably an undercount: cyberbullying happens off campus, often anonymously, and is inherently harder for school staff to detect or verify than a hallway altercation. Even taking the reported numbers at face value, the regional pattern has quietly inverted. Before the pandemic, upstate schools reported the highest cyberbullying rates in the state and New York City the lowest. By 2023 to 24, that had flipped. The city's rate climbed to 5.1 per 1,000, the highest of any region, while upstate's fell to 3.9. Two regions effectively traded places in five years, a shift easy to miss if you only glance at the statewide average.

Drugs and Alcohol: Back Above Pre Pandemic Levels

Unlike the serious violence categories, the definitions for drug and alcohol incidents have not changed since 2017 to 18, which means this trend line is as clean as the bullying data. Secondary schools reported 8,664 combined drug and alcohol incidents in 2023 to 24, with drug related incidents making up the overwhelming majority, 92.6 percent of that total. The statewide drug incident rate rose from 4.2 per 1,000 secondary students in 2017 to 18 to 6.5 per 1,000 in 2023 to 24, dipping sharply during remote learning before rebounding past its pre pandemic ceiling once students returned to classrooms. Alcohol incidents, by contrast, have stayed essentially flat, moving from 0.8 to 0.5 per 1,000 over the same period.

The regional split is the detail most likely to surprise a reader who assumes school drug problems concentrate in big cities. They do not, at least not according to this data. Upstate New York schools reported a drug incident rate of 9.8 per 1,000 secondary students in 2023 to 24, more than double the New York City rate of 4.2 per 1,000, and noticeably higher than the 6.2 per 1,000 rate in downstate districts outside the five boroughs. Whatever is driving adolescent drug use and disclosure in New York right now, it is not primarily an urban phenomenon.

The Schools That Report Nothing at All

Perhaps the most consequential number in the entire report is not a rate of anything. It is the share of schools that reported zero incidents of any kind, in any category, for an entire school year: 26 percent of all public schools statewide in 2023 to 24, including 41 percent of elementary schools and 5.9 percent of secondary schools. Among charter schools, the zero incident share was even higher, 52.6 percent overall.

The comptroller's office does not treat these as schools where nothing happened. Prior state audits of the reporting system, cited directly in this report, found a pattern of underreporting, misclassification, and incidents simply never making it into the state's records, both in the older VADIR system and in New York City's own implementation of it. A school reporting zero incidents may well maintain its own internal disciplinary records for fights or bullying that never crossed the threshold for state reporting, which means the official statewide numbers likely understate the true scope of the problem in both directions, serious incidents and lower level ones alike.

What Other Measures of the Same Schools Show

If the SSEC data leaves room for doubt, other surveys of the same student population tell a less ambiguous story. The New York City Department of Education's own annual school survey found that 51 percent of students in 2023 to 24 said students harass, bully, or intimidate one another at their school, a figure dramatically higher than the 2.6 percent of students the SSEC's incident count would imply if every reported bullying incident targeted a different child.

National survey data tells a similar story about the gap between official incident counts and lived experience. The CDC's Youth Risk Behavior Survey found that 18.9 percent of New York City high schoolers and 21.5 percent of high schoolers in the rest of the state reported being bullied on school property in the 2022 to 23 school year, a far higher share than the official incident rate would predict on its own. Nationally, the same survey found that 9 percent of high schoolers said they had been threatened or injured with a weapon at school, and 13 percent said they had skipped school at least once in the prior month specifically because they felt unsafe there.

What This Should Mean for Your District

None of this means New York's public schools are secretly dangerous, or that the state is hiding a crisis. It means the two headlines in this report are not actually comparable, and treating them as offsetting trends, one bad, one good, misreads what the data can and cannot tell a family. The drop in serious violent incidents is largely an artifact of a 2021 definitional change and should not be read as evidence that schools got meaningfully safer overnight. The rise in bullying and drug incidents, measured against an unchanged definition, is real, and New York City's recent trajectory in both categories deserves particular attention from families who assumed the city trailed the rest of the state on these measures.

For a parent trying to evaluate a specific school, the SSEC numbers are a starting point, not a verdict. A district's school safety plans, required under state law since 2000, are a better place to ask concrete questions: how bullying complaints get investigated, what counts as a reportable incident locally versus what state definitions require, and whether the district supplements its official numbers with the kind of anonymous student surveys that keep turning up experiences the official count misses entirely. The comptroller's office reaches a similar conclusion in its own recommendations, suggesting that student and family surveys, imperfect as they are, currently capture something about school climate that the state's own reporting system, by design, cannot.

Read full report.