The Road and the Woods

I graduated from Marjory Stoneman Douglas High School in 1996, among their earlier graduating classes. I didn't walk the stage, I didn't attend pep rallies, I didn't appear in the yearbooks. I was basically a renunciate, disinterested in social life. Looking back, it wasn't the healthiest but I had meaningful life outside of school, so it wasn't a big deal to me at the time. I wasn't motivated by it so didn't engage.

I entered their magnet program for art, having cleaned up my act in middle school after years of being the kid who spent too much time in the principal's office. I had an oceanography teacher who showed a personal interest in me — I wish I could remember his name. I had teachers who saw something in a kid who was a straight-A student but showed disinterest in social life. They paid attention. That attention mattered.

Twenty-two years later, seventeen human beings were killed in that building. Fourteen of them were students — children, the youngest just fourteen years old. Three were staff members who died trying to shield those children with their bodies. Fourteen families sent their kids to school that morning and never got them back. Fourteen bedrooms that still have their things in them. Fourteen seats at dinner that will never be filled again. The parents who survived — and I use that word deliberately, because something in them died that day too — have described what followed as an unending present tense: the loss does not recede. It reconstitutes itself every morning.

The impact on the Parkland community cannot be quantified. But the projected economic impact of gun violence across America can be, and has been: a peer-reviewed study published in JAMA and corroborated by the Everytown Research analysis estimates the total annual economic cost at $557 billion — 2.6% of gross domestic product. That number includes the medical costs, the mental health costs, the lost productivity, the criminal justice expenditures, and the quality-of-life losses that ripple outward from every shooting into every family and every community that is touched by it. It does not include the cost of a parent who can no longer function at work because their child is gone. It does not include the cost of a sibling who stops speaking. It does not include the cost of a community that no longer trusts the institution that was supposed to protect its children.

I did not experience the shooting. I do not claim the grief of those who did. But I spent the years between my graduation and that day becoming an intervention specialist — a behavioral researcher who worked with the most vulnerable students in the American education system, including students in crisis, students with autism, students who had been institutionalized because no one knew what else to do with them. I was trained in SAMA crisis intervention. I conducted Functional Behavior Assessments — the systematic analysis of why a person does what they do, what triggers it, what maintains it, and what could replace it.

And when I watched the coverage from Parkland, the framework I had spent my career building asked a question that nobody on television was asking:

Not: how did he get the weapon. Not: who failed to report the warning signs. Not: which law would have prevented this. Those are important questions. They are also the questions everyone was already asking. The question nobody asked — the question behavioral science is designed to answer — is the one that could change the trajectory of the next student who is walking toward the woods.

The framework nobody applied

In behavioral science, every behavior has a function. This is not a metaphor. It is a measurable, assessable, documentable fact. When a child in my classroom bit another child, I did not start with punishment. I started with assessment: What happened before the bite? What did the child gain or avoid by biting? What would serve the same function without the harm?

This is a Functional Behavior Assessment — an FBA. It is required by law in American special education. It is the most powerful diagnostic tool in applied behavioral science. And it has never been systematically applied to the problem of mass violence.

The FBA asks four questions about any behavior:

1. What is the antecedent? What conditions precede the behavior? What environmental, social, emotional, and institutional factors create the context in which this behavior becomes possible?

2. What is the function? What need does the behavior serve? In behavioral science, there are four primary functions: escape (avoiding something unbearable), attention (being seen by a system that made you invisible), control (exerting power in a context of powerlessness), and sensory/automatic (the behavior has become self-reinforcing through rehearsal).

3. What is the replacement behavior? What alternative behavior could serve the same function without the harm? This is the question that changes everything — because it shifts the intervention from stopping the bad behavior to teaching the better one.

4. What environmental changes reduce the antecedent? How do we redesign the conditions so the behavior is less likely to occur in the first place?

Let me apply each of these to what we know about mass school violence.

The antecedent analysis

The research on school shooters — particularly the work of the U.S. Secret Service National Threat Assessment Center, which has studied every targeted school attack since Columbine — reveals a consistent antecedent pattern. It is not a single trigger. It is an accumulation:

Social isolation. The majority of school shooters experienced prolonged social disconnection — not the temporary loneliness of a bad semester, but the chronic, structural isolation that rewires the stress response. The neuroscience is clear: Holt-Lunstad's meta-analysis shows that chronic isolation increases mortality risk by 26% and elevates cortisol to levels that impair the hippocampus — the brain region responsible for emotional regulation and memory. Sapolsky documented the same cascade in isolated primates. The brain, evolved for social life, interprets prolonged isolation as existential threat.

Untreated emotional dysregulation. The student who cannot name what they feel cannot regulate what they feel. This is not a personality deficit. It is a training deficit. Emotional regulation is a skill — a behavioral kata — that must be practiced under real conditions to become available under pressure. Most American schools do not teach it. They punish the absence of it.

Institutional invisibility. The student who becomes dangerous has typically been signaling distress for months or years — declining grades, social withdrawal, changes in behavior, expressions of hopelessness. These are leading indicators. The shooting is the lagging indicator. But most institutions are designed to measure lagging indicators — test scores, disciplinary incidents, attendance records — not the trajectory that precedes them.

Access to means. In behavioral research, the environment is part of the antecedent. Access to a weapon does not cause the behavior — but it compounds the lethality of the behavior when it occurs. This is the same principle that applies to bridge barriers in suicide prevention: the barrier does not address the underlying distress, but it interrupts the behavioral chain at the point where the consequence becomes irreversible. Countries and communities that have modified environmental access to weapons have seen measurable reductions in mass casualty events. This is not a political statement. It is an environmental observation consistent with the same FBA framework that governs every other recommendation in this analysis. The behavioral scientist's position is straightforward: environmental modification is one layer of intervention. It is not the only layer, and in the absence of the other layers — the ones that address the function — it is insufficient.

The evidence: from mice to meta-analysis

The isolation-aggression link is not speculative. It is one of the most replicated findings in behavioral science — across species, across methodologies, across decades.

The meta-analytic evidence. Jin et al. (2023) conducted a meta-analysis of 92 studies and found a statistically significant positive correlation between social exclusion and aggression. This was not a marginal finding — the relationship was consistent across experimental paradigms, age groups, and cultural contexts. A separate meta-analysis of 53 studies (Lin et al., 2025; N = 21,405) confirmed that social exclusion produces a significant negative correlation with prosocial behavior (r = −.10), meaning the same mechanism that increases aggression simultaneously decreases the cooperative behaviors that would serve as natural buffers against it. The excluded individual loses both the brake and the alternative route.

Leary, Kowalski, Smith, and Phillips (2003) examined case studies of school shooters directly and found that experiences of teasing, rejection, and social exclusion were present in the histories of the overwhelming majority. The Secret Service National Threat Assessment Center corroborated this in their analysis of every targeted school attack since Columbine: most attackers had experienced social rejection or isolation, and most had communicated their distress to others before the event.

The animal research — what isolation does to a brain. The human data tells us that isolation predicts aggression. The animal research tells us why — by revealing the neural mechanism that makes the connection causal, not merely correlational.

In mice, social isolation produces aggression so reliably that it is used as a standard experimental protocol — "isolation-induced aggression" is a recognized model in neuroscience. A study published in PLOS One (Ma et al., 2011) demonstrated that socially isolated male mice subjected to chronic mild stress showed significantly elevated aggression, reduced sucrose intake (a measure of anhedonia — the inability to experience pleasure), and increased anxiety-like behavior. The isolation didn't just produce hostility. It destroyed the capacity for reward from normal social activity while simultaneously increasing the drive toward aggressive contact.

In primates, the evidence is even more directly translatable to human institutions. Robert Sapolsky's thirty-year field study of wild baboons in the Serengeti demonstrated that chronic social stress — particularly the unpredictability and loss of control experienced by low-ranking individuals — produces chronically elevated baseline cortisol. This is not a metaphor for human distress. It is the identical hormone, measured in blood, following the identical neurological pathway. Low-ranking baboons show impaired hippocampal function (the brain region responsible for emotional memory and regulation), reduced capacity to terminate the stress response, and elevated rates of cardiovascular disease.

The critical finding for violence prevention: Sapolsky showed that the pathological stress response is driven not by the social rank itself but by the loss of predictability and control. During periods of social instability — when hierarchies were being disrupted and outcomes were unpredictable — even high-ranking baboons showed the cortisol profiles of subordinates. This maps directly onto the experience of the isolated student: it is not low status per se that produces the neurological cascade. It is the combination of isolation, unpredictability, and perceived loss of control — the exact antecedent pattern identified in the FBA.

The prefrontal cortex — the brain region responsible for impulse control, decision-making, and emotional regulation — is particularly vulnerable to chronic stress hormones. Marmot and Sapolsky's combined work demonstrated that social hierarchies produce measurable gradients in prefrontal cortex function, with lower-status individuals showing progressively impaired executive function. MRI studies in non-human primates (Sallet et al., 2011; Noonan et al., 2014) found that gray matter volume in the prefrontal cortex correlates with both social rank and social network size. The isolated brain is structurally different from the connected brain — and the difference is located in exactly the region that would need to function well to inhibit a violent impulse.

What this means for intervention design: The animal research makes the case that isolation-induced aggression is not a choice in the way we typically use that word. It is a neurobiological cascade: chronic isolation → elevated cortisol → impaired prefrontal function → reduced impulse control → increased aggression. The cascade can be interrupted at every stage — but only if the intervention reaches the student before the prefrontal impairment becomes structural. This is why early identification and trajectory monitoring are not optional features of a prevention program. They are the mechanism by which the intervention arrives in time.

The function analysis

This is where the conversation changes.

Most public discourse about school violence focuses on the antecedent (what went wrong before) and the consequence (what should happen after — punishment, legislation, security). Almost no public discourse addresses the function — the need the behavior is serving. This gap is not an oversight. It is a structural avoidance, because asking what need does violence serve? feels like excusing it. It is not. It is the precondition for preventing it.

Escape. For some individuals, the act is an attempt to escape unbearable internal states — pain, humiliation, despair that has become chronic and seemingly permanent. The FBA question: What other escape routes were available? Were any taught? Were any modeled? Was there a single adult in this student's institutional life who demonstrated that unbearable feelings can be survived, named, and regulated?

The replacement behavior for escape is not "stay and endure." It is the practiced capacity to sit with distress without being destroyed by it — the skill I call distress tolerance in The Engine of Trust. It must be taught through what I describe as the Environmental Grief Protocol: sitting with genuinely distressing information — real climate data, real community loss — without being paralyzed by it. It must be practiced, under real emotional conditions, repeatedly, before it is available in crisis. A worksheet about feelings is not the intervention. A practiced behavioral kata — name the feeling, locate it in the body, breathe, choose one action — rehearsed dozens of times on real frustration (the bread didn't rise, the experiment failed, the friendship broke) — is.

Attention and recognition. The manifesto pattern — the videos, the writings, the social media posts — reveals a function that is agonizing to name: the individual wants to be seen. Not admired. Seen. Recognized as existing by an institution that rendered them invisible. The FBA question: When was the last time this student was seen — genuinely, specifically, by name — by an adult in the institution? Not evaluated. Not disciplined. Seen.

The replacement behavior for attention-seeking through violence is attention received through contribution. This is exactly what prosocial recognition architecture is designed to provide: making a person's positive contributions visible to the institution, so the need to be seen is met before it metastasizes. A student whose peer recognition says "your documentation of the garden pH saved our whole experiment" has been seen. That recognition costs nothing. Its absence costs everything.

Control. The function of violence as control is the exertion of power in a context of profound powerlessness. Self-Determination Theory identifies three basic needs: autonomy, competence, and relatedness. When all three are chronically thwarted — when the student has no choices, no sense of capability, and no connection — the need for control doesn't disappear. It finds another channel.

The replacement behavior for control is genuine agency. Not the performance of choice (pick your elective) but the structural experience of making decisions that matter and seeing them produce real outcomes. A capstone project — where a student identifies a real community problem, designs a real solution, builds it, and presents results to real stakeholders — is a control replacement. The student has power. It is real. It produces visible results. The need for control is met through contribution, not destruction.

The replacement behaviors we've never systematically taught

Here is the intervention gap, stated plainly:

American schools teach content. They test recall. They punish dysregulation. They do not, as a systematic institutional practice, teach the behavioral skills that prevent the trajectory from isolation to despair to violence. Those skills exist. They have names. They have evidence bases. And they could be embedded in the curriculum without adding a single class period — because they are practiced on real emotional events that already occur in the school day.

Self-regulation katas. The Bread Didn't Rise Practice (K–2): pause, name the feeling in your body, draw it, choose one thing to do with it, try again. The Frustration Lab (3–5): the emotion becomes data — name it, separate feeling from experiment, form a new hypothesis. The Environmental Grief Protocol (6–8): sit with genuinely distressing information without being paralyzed. The Failure Debrief (9–12): structured self-analysis after setbacks. These are not add-ons. They are practiced at the moments when real emotions arise — because emotional regulation cannot be taught in theory.

Metacognition training. The capacity to observe your own thinking — to notice the rage arising, to see the thought before it becomes the action, to recognize the rut before you're in it. Hattie's meta-analysis: metacognitive instruction produces a 0.69 effect size — approximately eight additional months of learning per year. But the benefit that matters here is not academic. It is the creation of a choice point between stimulus and response. The student who can observe their own mind has options the unobserved mind does not.

Social connection architecture. Parallel play before cooperative demand. Proximity without pressure. The one-kitchen principle. Third spaces where accidental encounters occur. Every intervention in this list is an antecedent modification — redesigning the social environment so that isolation becomes structurally difficult rather than structurally inevitable.

Threat assessment that measures trajectory. The Secret Service framework already does this — it assesses the pathway to violence rather than profiling a "type." What the behavioral framework adds is the measurement of leading indicators: social withdrawal as a trajectory, not a personality trait. Emotional dysregulation increasing over time. The say-do gap between what the student presents and what they are experiencing. These are measurable. They can be tracked. And they can trigger intervention before the trajectory reaches the point of no return.

What the data says works

The interventions described above are not theoretical. They have been tested at scale, measured rigorously, and published in peer-reviewed meta-analyses. Here is what the evidence shows:

The 2023 update. Cipriano et al. (2023) updated the Durlak analysis with 424 studies from 53 countries — the most comprehensive review of SEL evidence ever conducted. The findings confirmed and extended the original results: reduced externalizing behaviors (g = 0.22, a meaningful effect), reduced emotional distress (g = 0.12), improved SEL skills (g = 0.18), and improved peer relationships (g = 0.27). The effect sizes are modest in statistical terms but massive at population scale — a 22% reduction in disruptive and aggressive behavior across hundreds of thousands of students means thousands of students who did not become the next data point.

Mindfulness-based interventions. Gillions et al. (2019) conducted a systematic review of 22 studies and found that mindfulness-based interventions effectively reduced aggression and violence in adults. The mechanism aligns precisely with the neurological cascade described above: mindfulness training strengthens prefrontal cortex function, improves the capacity to observe an impulse before acting on it, and creates the choice point between stimulus and response that the chronically stressed brain lacks.

Threat assessment — the Secret Service model. The U.S. Secret Service National Threat Assessment Center has studied every targeted school attack since Columbine and developed a behavioral threat assessment model that focuses on the pathway to violence — not the profile of a "type." Their research confirms that targeted violence is not impulsive. It follows a discernible pathway that includes grievance, ideation, planning, and preparation. At each stage, intervention is possible — but only if the institution is monitoring the trajectory, not waiting for the lagging indicator.

The institutional rut — and what we spend instead

Not to be insensitive. No number, no dollar amount can quantify the loss of even one human life. But I use the pro forma formula I developed in The Engine of Trust — a financial model I learned in commercial real estate, designed to project the future performance of an asset — because it is revealing of a truth we have long overlooked.

Against the backdrop of $557 billion in annual gun violence costs — roughly $1,700 for every American citizen, every year — the numbers on school spending are staggering. They reveal a fundamental misallocation.

American public schools now employ more security guards and school resource officers than counselors, nurses, psychologists, and social workers combined. The annual expenditure on security guards alone is $12 billion; SRO salaries and benefits add another $2.62 billion (Urban Institute, 2023 dollars). Surveillance technology — cameras, electronic locks, facial recognition — exceeds $3.1 billion per year and is growing at 8% annually (Omdia market research). Total school security spending: approximately $17.8 billion every year. Schools that experience a shooting see per-student spending increase by an average of $248 — almost entirely directed toward hardening rather than prevention.

Meanwhile, the downstream costs accumulate — and they do not reset at the end of the fiscal year. Students exposed to school shootings show a 21% increase in antidepressant use lasting two or more years after the event, a 12% rise in absence rates, a 28% increase in chronic absenteeism, and a measurable 2.4% decline in per-capita earnings in affected counties. The reactive architecture does not merely fail to prevent. It generates its own costs — costs that compound across families, communities, and economies long after the cameras stop recording.

And the reactive architecture does something worse: it actively degrades the conditions it claims to protect. Research shows that the presence of police officers in schools increases the likelihood of student arrest by a factor of 2.6. Those arrests feed the school-to-prison pipeline — student suspensions deprive students of instruction time, damage trust in the institution, increase dropout likelihood, and drive the very isolation and disengagement that the FBA identifies as the antecedent pattern for violence. The security guard does not interrupt the trajectory. The security guard accelerates it.

The reason these interventions have not been systematically deployed is not that they are unknown. It is that the institutional ruts are too deep — and they are older than most people realize.

The American public school system was not designed for human development. It was designed for compliance. In 1837, Horace Mann imported the Prussian school model — a system whose architect, the philosopher Fichte, stated its goal openly: education should shape students so thoroughly they would be incapable of thinking otherwise than as their schoolmasters wished. The Carnegie, Rockefeller, and Ford foundations shaped the expanded curriculum for factory workforce needs. Once child labor laws removed children from factories, industrialists needed parents at the machines for eight hours — and the public school system became the childcare infrastructure. The bells stayed. The rows stayed. The architecture of compliance stayed.

The result, measured across two decades of published meta-analyses, is a system operating at approximately 50% of achievable effectiveness. Chen and Yang's meta-analysis of project-based learning (66 studies) found an effect size of 0.44 standard deviations — equivalent to four to five additional months of learning per year. Applied across thirteen years of schooling, the current methodology delivers the equivalent of roughly 6.5 years of real education. The institution was already failing its students academically before it failed them emotionally — and the violence is the catastrophic tail event of that systemic failure.

This is not the first time the system has maintained a failed methodology because the institutional ruts were too deep. Emily Hanford's 2022 investigation, Sold a Story, documented how American schools taught reading using methods that cognitive science had discredited for decades — because publishers had inventory, experts had reputations, and no one measured the trajectory until millions of children could not read. The pattern described in this article is the same pattern, applied to a different domain, with higher stakes.

The rut of standardized testing rewards content recall, not emotional intelligence. The rut of disciplinary systems punishes dysregulation rather than teaching regulation. The rut of class sizes — twenty-six students to one teacher, in my daughter's classroom — makes individualized attention structurally impossible. The rut of school counselor ratios — one counselor per 372 students nationally (ASCA, 2024–25 data), but in elementary and middle schools the ratio is 571 to 694 students per counselor. Eight million American children attend a school with no counselor at all. Three million have no access to any school support staff — no psychologist, no social worker, no one whose job it is to notice the trajectory.

These are not policy debates. They are design failures. The institution was built to deliver content at scale. It was not built to develop the whole human being at scale. And when a student falls through the cracks in that design — when the institution fails to see them, fails to teach them the skills they need, fails to provide the connection their nervous system requires — the institution has created the antecedent conditions for the very behavior it will later claim it could not have predicted.

The ruts were not designed. They were carved by decades of underfunding, overcrowding, standardized testing mandates, and the accumulated weight of a system that was built for compliance in 1837 and never fundamentally redesigned. And like all ruts, they are invisible to the people inside them — until the cart breaks.

What we can build

I wrote a book called The Engine of Trust about institutional design — about building organizations, schools, and governments that encode their purpose into their structure rather than depending on the goodwill of the people inside them. The behavioral framework in that book — the Functional Behavior Assessment, the Replacement Principle, the Behavioral Kata, the Visibility Thesis — applies directly to the problem described in this article.

The solution is not a single intervention. It is an architecture:

Layer 1: Teach the replacement behaviors. Embed self-regulation, metacognition, and social-emotional skills into the curriculum — not as a separate class, but as a practiced response to real emotional events that occur in the school day. Cost: negligible. The moments already exist. The katas are what's missing.

Layer 2: Redesign the social environment. Reduce structural isolation through proximity architecture, parallel play opportunities, third-space design, and connection rituals. Ensure every student is seen — specifically, by name, for something they contributed — by at least one adult every week.

Layer 3: Measure the trajectory. Deploy leading-indicator monitoring that tracks social connection, emotional regulation, and engagement over time — not to surveil, but to intervene early. The student who is withdrawing needs a conversation in October, not a disciplinary hearing in February.

Layer 4: Modify the environment. The Department of Homeland Security's own data shows that 76% of guns used in school shootings came from a parent or close relative, and approximately half were easily accessible. This is an environmental antecedent — and it has a concrete, affordable, non-partisan solution.

A best-practice biometric gun safe — fingerprint-activated, tamper-resistant, sub-two-second access — costs between $150 and $350 for a handgun model. The technology is mature: multiple manufacturers produce UL-certified units that store 20+ fingerprints, include backup access methods, and are small enough for a nightstand or vehicle. A government subsidy program — modeled on existing safety incentive programs for smoke detectors, car seats, and home weatherization — could provide one certified biometric safe per gun-owning household at a cost of approximately $200–$300 per unit. For a district of 1,000 students, assuming 40% of households own firearms (consistent with national survey data), the cost of a universal safe subsidy would be approximately $80,000–$120,000 — a one-time investment, not an annual expense.

The legal precedent already exists. In 2024, Jennifer and James Crumbley became the first parents in American history convicted of involuntary manslaughter in connection with their child's school shooting. The conviction was based on gross negligence — purchasing a firearm for a child showing warning signs and failing to secure it. The parents were each sentenced to 10 to 15 years in prison. In Georgia, Colin Gray became the second parent charged under similar circumstances. Michigan subsequently enacted a safe storage law requiring firearms to be stored unloaded and secured with a locking device if a minor lives at or is likely to visit the property.

The framework is straightforward: provide the tool (subsidized biometric safe), establish the standard (safe storage requirement), and enforce the accountability (when a minor accesses an unsecured firearm that was involved in an act of violence, the failure to use an available, subsidized safe constitutes negligence). This is not a restriction on ownership. It is the same principle as requiring a seatbelt in a car — the tool exists, the cost is minimal, the subsidy removes the financial barrier, and the failure to use it when harm results is a foreseeable consequence of a choice not to act. Environmental modification reduces lethality. Behavioral intervention reduces likelihood. Both are needed. Neither is sufficient alone.

Layer 5: Train the adults. Teachers, counselors, administrators, and staff need the same behavioral katas the students need — the capacity to observe before reacting, to name before questioning, to regulate their own nervous systems before attempting to regulate a child's. A dysregulated adult cannot teach regulation. A teacher trained in SAMA crisis intervention, the Elicitation Standard, and the Socratic Kata has the tools. Most teachers have never been offered them.

The road

I was the kid who spent time in the principal's office. I had a roving mind and a social disinterest that could have been read, by a less attentive institution, as a warning sign. What kept me on the road was not discipline. It was the teachers who saw me — Mr. Khan with his brain teaser, the oceanography teacher whose name I've lost but whose attention I haven't, and a mother who put everything in perspective by telling me about a nun in Brooklyn who hit her across the knuckles with a ruler while she practiced piano on a piece of cardboard.

The ruts in our institutions keep us from helping the students who have left the road for the woods. The behavioral science to bring them back exists. The curriculum to teach the skills exists. The framework to measure the trajectory exists. What doesn't exist — yet — is the institutional will to implement it.

This article is not a policy proposal. It is a behavioral analysis — the same kind of analysis I conducted for every student in my classroom who was struggling, applied to the problem that no one has thought to analyze this way. The FBA doesn't take sides. It asks what the behavior is, what function it serves, what could replace it, and what environmental changes would make the replacement more likely. That is all it does. That is everything it does.

I graduated from a building that became a symbol of institutional failure. I spent the next twenty-five years learning why institutions fail and how to build ones that don't. The framework is in the book. The application to this specific, devastating problem is in the article you just read.

A note on the numbers

I hesitated to write this section. Applying a financial model to the prevention of violence against children feels wrong — and in some ways it is. No spreadsheet captures what is lost when a student is killed. No ROI calculation justifies itself against the weight of a single empty desk.

But I have spent my career watching institutions make decisions based on budgets. School boards allocate funds based on projected returns. Superintendents justify programs based on measurable outcomes. Legislators fund what they can defend with numbers. If the people who control the resources need numbers to move — then the numbers need to exist. Shining a light on the raw data is the only way to make the case for meaningful change.

So I built one. Using the same pro forma methodology I developed in The Engine of Trust — a financial projection model I learned in commercial real estate, adapted for institutional design — I modeled the five-layer prevention architecture described in this article against the current reactive security spending, projected over ten years. The model includes a Voltage Drop — a systematic discount for the gap between what the research predicts and what a real institution can capture in practice — applied to every projection.

What the national-scale projection reveals: There are 49.5 million students in American public schools. Over a decade, the current reactive architecture — security guards, cameras, lockdown systems — costs approximately $178 billion and produces zero measurable change in the conditions that create violence. The five-layer prevention architecture, at a volume-adjusted cost of $54 per student in Year 1 and $27 per student in Years 2 through 10, costs $14.7 billion over the same period and produces an estimated $99.1 billion in measurable returns from reduced disciplinary incidents, preserved academic funding, and decreased absenteeism. The net 10-year return on prevention is $84.4 billion. The ROI reaches 6.7× on the prevention spend — and $163 billion ahead of the reactive path. The annual ROI crosses breakeven in Year 1 and compounds every year thereafter as implementation fidelity improves and the behavioral skills transfer across student cohorts.

To put that in human terms: at a single 1,000-student district, the reactive path costs $4.16 million over a decade. The prevention path costs $765,000. The net advantage is $5.28 million — enough to hire thirteen additional counselors for ten years.

The most important finding is not in any column. It is the estimated 20–50 students per district over ten years who would be identified early through trajectory monitoring — students whose social withdrawal, emotional dysregulation, or loss of connection would trigger a conversation in October rather than a crisis in February. Those students do not appear in the ROI calculation. They are the reason the ROI calculation exists.

The national-scale projection — applying this framework to the full 49.5-million-student population, including the Voltage Drop methodology, per-seat cost tables, compounding analysis, and visual comparison charts — is published as a companion document: What If We Did This for Every Student in America?

An invitation

I have a potential solution. It is not the only one, and I do not pretend it is the best one. But it exists, and it is built on the evidence described in this article.

The Engine of Trust describes a complete institutional design framework — including the five-layer prevention architecture, the Behavioral Kata system (the Bread Didn't Rise Practice, the Frustration Lab, the Environmental Grief Protocol — practiced on real emotional events, not worksheets), metacognition training through the Socratic Kata methodology, and the prosocial recognition architecture that ensures no student is invisible to the institution. These are not theoretical constructs. They are designed for integration into existing school curricula without a new schedule, without a new hire, and without a new class period — because they are practiced at the moments when real emotions arise, which is the only way behavioral skills transfer from the classroom to the crisis.

But this is not a sales pitch. It is an invitation.

If you are an educator, a curriculum designer, a school board member, a legislator, a parent, or a researcher — and you can take the framework in this article and build something better — do it. Light that fire. Act on this. Design the interventions. Test them. Measure them honestly. Publish the results, including the failures. If you can find a more elegant solution than the one I have built, I will be the first to celebrate it — because the point was never the product. The point was always the student.

The behavioral science is clear. The evidence base is published. The framework exists. What has been missing is the institutional will to implement it — and the voices willing to say, publicly, that what we are currently doing is not working.

This is my voice. I graduated from a building where seventeen people were killed. I spent twenty-five years learning why institutions fail. I built a framework for how they could work. And I am telling you: the students who are walking toward the woods can be seen. The replacement behaviors can be taught. The road can be straightened.

Now it is your turn.

About the author

Jeremy E. Kunzinger is the author of The Engine of Trust: Behavioral Science, Institutional Design, and the Architecture of Human Potential (Knight of Cups, LLC, 2026) and the companion novel The General's Eye. He is a former K–12 intervention specialist, SAMA-certified behavioral trainer, autism specialist, and commercial real estate broker. He graduated from Marjory Stoneman Douglas High School in 1996.

The behavioral framework described in this article — the Functional Behavior Assessment, the Replacement Principle, the Behavioral Kata, and the institutional design methodology — is developed in full across twenty chapters of The Engine of Trust.

Chief Academic Officer, Green Life Guardian · Author of The Engine of Trust and The General's Eye