Most RCAs find the active failure and call it done—but the latent conditions that made it possible are still there.
A welder at a fabrication shop in Louisiana died when a crane operator lifted a load over an active work area. The root cause analysis took three days, interviewed six people, and concluded: "Operator failed to follow lift plan procedures."
Case closed. Operator retrained. Safety stand-down completed. Everyone back to work.
Six months later, a rigger nearly gets crushed in the same yard when a different crane operator swings a load through a pedestrian walkway. Different operator. Same root cause listed: "Failed to follow procedures."
The Swiss Cheese Model explains exactly why this keeps happening—and why most root cause analysis never actually finds the root cause.
James Reason's Swiss Cheese Model describes how accidents happen through multiple layers of defense, each with holes. An incident occurs when the holes align—when every barrier fails at once.
The holes break into two categories:
Active failures: The last person who touched it. The operator who didn't follow the lift plan. The supervisor who didn't do a walkdown. The employee who skipped the lockout step. These are visible, immediate, and easy to blame.
Latent conditions: The pre-existing weaknesses that made the active failure possible. The lift plan that nobody can actually read in the field. The production pressure that makes cutting corners the rational choice. The fact that three supervisors have quit in six months and nobody trained their replacements. The unspoken reality that complaining about unsafe lifts gets you labeled as "not a team player."
Most root cause analysis finds the active failure and stops. It never finds the latent conditions—because finding them means admitting the system was designed to fail.
Back to that fabrication shop. The RCA found the active failure: the crane operator lifted a load over an active work area, violating the lift plan. Operator retrained. Investigation closed.
But here's what the investigation never asked:
Those are the latent conditions. The holes in the cheese that were already there, long before that operator clocked in that morning.
The reason the second incident happened six months later is because none of those holes got fixed. The investigation retrained one operator and left every latent condition intact.
The Swiss Cheese Model is not a metaphor. It's a diagnostic tool. Every layer represents a defense you built—procedures, training, supervision, engineering controls, PPE. The holes represent the gaps in each defense.
Most companies only see the holes after an incident, when they're looking backward through the stack and realize every layer failed at the same time. But those holes were always there. They just hadn't aligned yet.
The question isn't "why did this person make a mistake?" The question is: "what conditions made this mistake inevitable—and how many other holes are sitting there waiting to align?"
Latent conditions hide in plain sight. They sound like:
The companies that actually find latent conditions don't do it through incident investigations. They do it through:
Near-miss investigations that treat close calls like actual incidents. The crane operator who almost swung a load over a work area is giving you the same data as the one who actually did. Most companies ignore it because nobody got hurt.
Anonymous safety perception surveys. Employees know where the latent conditions are. They won't tell you in a toolbox talk with their supervisor standing there. They will tell you anonymously.
Management walkthroughs that ask "why" five times. Observation: rigger isn't wearing a hardhat. Why? "I left it in the truck." Why? "It doesn't fit over my hood." Why? "We don't have the right size." Why? "Procurement says we're over budget." That's a latent condition.
Auditing your corrective action backlog. How many corrective actions are overdue by 30+ days? Every one of those is a known hole in the cheese that hasn't been fixed. Track the age of your backlog—it's a leading indicator of how many latent conditions you're ignoring.
Here's the question that separates a real root cause analysis from theater: "If we fired the person responsible, would this happen again?"
If the answer is yes, you didn't find the root cause. You found the active failure.
In the fabrication shop case: fire the crane operator who violated the lift plan, hire a new one, and six months later a different operator does the same thing. Because the latent conditions—production pressure, bad site layout, unspoken expectations, lack of stop-work authority—are still there.
Root cause analysis that stops at human error is just paperwork. It's a record that you investigated, not evidence that you fixed anything.
The Swiss Cheese Model, developed by James Reason, explains that accidents occur when holes (failures) in multiple layers of defense align. Each layer—procedures, training, supervision, engineering controls—has weaknesses. An incident happens when all the weaknesses line up at once, allowing a hazard to pass through every barrier.
An active failure is the immediate, visible error: the operator who didn't follow the procedure, the worker who removed a guard. A latent condition is the pre-existing weakness that made the active failure possible: the unreadable procedure, the production pressure that rewards speed over safety, the broken equipment that was never fixed. Active failures are easy to find. Latent conditions require asking why the active failure was able to happen in the first place.
Because finding the active failure is fast, visible, and doesn't implicate leadership decisions. Finding latent conditions means admitting the system was designed to fail—that site layout, staffing, budgets, production schedules, or management priorities created the conditions for the incident. Most RCAs stop at "human error" because going deeper is politically uncomfortable.
Latent conditions show up in near-miss reports, anonymous safety surveys, corrective action backlogs, and employee interviews. If the same near-miss keeps happening, if corrective actions sit overdue for months, if employees say "we've always done it that way" or "there's no time to do it right"—those are latent conditions. Track your near-miss trends, audit your corrective action aging, and ask "why" five times during management safety walks.
Absolutely. A recordkeeping error on your OSHA 300 log doesn't happen because one person fat-fingered a date. It happens because multiple defenses failed: the initial incident report was incomplete, the supervisor didn't review it, the safety manager didn't catch it during quarterly audits, and leadership never allocated time for a proper recordkeeping review process. Same model. Same alignment of holes.
Root cause analysis, corrective action tracking, OSHA 300 log accuracy, incident investigation documentation—it's necessary, but it's not where you add value. Your job is to find the latent conditions before they align. Our job is to handle the paperwork so you have time to do that.
EHS Inc. manages your OSHA recordkeeping, compliance documentation, training records, and ISNetworld updates so you can spend your day doing actual safety work instead of filling out forms. If the admin burden is keeping you from the work that actually prevents incidents, let's talk.
Aaron West
Founder, EHS, Inc. — 18+ years in EHS compliance and contractor safety
Aaron West has spent over 18 years helping contractors and businesses navigate OSHA compliance, ISNetworld® certification, and workplace safety management. He founded EHS, Inc. to make enterprise-level EHS accessible to companies of all sizes — serving contractors and businesses nationwide — without long-term contracts or enterprise overhead.
Our team handles the complexity so you can focus on running your business. No long-term contracts, no learning curve.
Talk to EHSFramework to achieve zero incidents
Stop hitting paywalls
54 topics in English & Spanish