Why We Overreact to the Wrong Things on a Jobsite

Chris Fredette
Jul 20
4 min read

You have seen it. A guy stubs his toe or nicks his hand, and suddenly there is an incident report, a stand-down, a root cause meeting, and a corrective action plan. Meanwhile, three feet away, a worker is standing under an unsecured load and nobody says a word, because the load did not drop that day.

I spent years feeling like something was off about how we react to things, and I could never name it. We treat a paper cut and a near-decapitation like they sit on the same ladder. They do not. And once you understand why, a lot of the noise on a jobsite starts to make sense.

By the Numbers

A 2020 Construction Safety Research Alliance study used 17 years of data and 3.2 trillion worker-hours and found no discernible link between a company's total recordable injury rate and its fatalities.
The construction fatality rate has hovered around 10 deaths per 100,000 workers for over a decade, even as recordable injury rates fell to 20-year lows.
Construction recorded 1,075 deaths in 2023, the highest count since 2011.
Falls accounted for 39% of those deaths, and most fatal falls happened from heights between 6 and 30 feet.
In one contractor's review of 684 incidents, 88% had no realistic potential to cause a serious injury or fatality. Only 12% did.

The Number That Should Have Changed Everything

For 50 years we have measured safety with TRIR, the total recordable incident rate. Low TRIR became the trophy. Companies put it on banners and used it to win bids. The logic seemed sound: drive down injuries and you drive down deaths.

Then the data came in, and the logic fell apart. The Construction Safety Research Alliance ran the numbers across 3.2 trillion worker-hours and found no meaningful relationship between how often a company gets hurt and how often it kills someone. You can have a spotless recordable rate and still bury a worker. The two are measuring different things.

Here is why. Most recordable injuries come from low-energy hazards: a sprain, a cut, a slip on the same level. The hazards that kill people are different in kind, not just degree. A fall from height, a suspended load, an energized panel, a trench wall. Knocking down the small stuff does almost nothing to the big stuff, because the big stuff was never on the same continuum to begin with.

Why We Overreact, and Why It Matters

This is the part that finally answered my own question. When a minor injury happens, it is visible, it generates paperwork, and somebody up the chain wants to see a response. So we respond hard, because reacting to the thing in front of us feels like doing the job.

But the energy of that response is finite. Every hour spent dissecting a stubbed toe is an hour not spent on the unsecured load, the missing trench box, the guy free-climbing because the lift was tied up. We pour our attention into the incidents that could never have killed anyone and starve the exposures that could.

The contractor data makes this concrete. When one company actually sorted its incidents by potential severity, nearly nine out of ten had no path to a serious injury. Real prevention did not come from reacting harder to all of them. It came from finding the 12% that could kill someone and asking a sharper question about those.

Energy-Based Safety: The Sharper Question

The shift is simple to say and hard to unlearn. Stop sorting incidents by what happened. Start sorting hazards by how much energy is in them.

Every serious injury involves the unwanted release or transfer of energy. Gravity. Motion. Mechanical. Electrical. Pressure. Temperature. No high-energy source, no fatality. Researchers, most notably Dr. Matthew Hallowell at the University of Colorado Boulder, put a rough line on it: a high-energy hazard is one capable of causing serious harm on contact, like a fall above four feet, equipment in motion, or energized electrical. Below that line, your body usually tolerates the hit. Above it, it does not.

Then comes the question that does the real work. For every high-energy hazard, is there a direct control in place? A direct control physically blocks the energy from reaching a worker even if someone makes a mistake: a guardrail, a machine guard, lockout. An indirect control depends on a human behaving perfectly: a warning sign, a reminder, a spotter, a "be careful." If you have a high-energy hazard and only an indirect control, you are one human error away from a funeral, no matter how clean your recordable rate looks.

That is the whole game. Find the energy. Name the control. If the control is not direct, fix that before the crew touches the task.

Key Takeaways

A low recordable rate does not mean you are protecting people from the hazards that kill. The two are not correlated.
Most incidents on a jobsite have no realistic potential to cause death. Reacting equally hard to all of them wastes the attention serious hazards need.
Serious injuries come from high-energy hazards: falls, suspended loads, energized systems, pressure, motion.
The question that matters is not whether someone got hurt. It is whether there is a high-energy hazard present, and whether the control is direct or you are trusting a person not to slip.
Build your pre-task planning around finding the energy and verifying a direct control, not around counting minor injuries after the fact.

The next time a crew overreacts to something that was never going to hurt anybody, it is worth asking where that energy could have gone instead. The hazards that kill are usually quieter than the ones that generate paperwork. Learning to see them is the job.