Why Blaming the Person Almost Never Fixes the Problem

The impulse to look for a person before looking for a structure is not a leadership failing. It is a well-documented cognitive default — and acting on it without checking first produces a specific kind of waste that organizations pay for, on a delay, in the next hire.

Something goes wrong inside an organization, and the first question almost everyone asks is the same question: who's responsible.

It feels like the right question. It's usually the wrong one.

Not because people aren't sometimes responsible. They are, plainly, often. But because the instinct to look for a person before looking for a structure is not a judgment call. It's a cognitive default, and it has been measured, named, and studied for fifty years.

The bias has a name, and it isn't subtle

In 1977, the psychologist Lee Ross gave a name to something every manager has done without noticing: the tendency to explain someone else's failure by pointing at their character, and to explain your own failure by pointing at your circumstances. Ross called it the fundamental attribution error, and the name has held up because the finding has held up. Watch someone else miss a deadline, and the easy story is that they're careless or unmotivated. Miss your own deadline, and the easy story is that the brief changed twice, the dependency was late, the calendar conspired against you. Same failure. Different story, depending entirely on whose failure it is.

Inside a company, this bias doesn't sit still — it climbs the org chart and gets sharper as it climbs. A landmark 1980 study of how supervisors actually diagnose underperformance (Mitchell and Wood) found the same pattern running underneath formal performance reviews: managers consistently explain a subordinate's poor output by pointing at the subordinate, low effort, low ability, rather than at the task design, the resourcing, or the conflicting priorities that made the outcome close to inevitable, and they respond more punitively the more times the pattern repeats. The diagnosis comes first. The intervention follows from the diagnosis. So the fix is almost always aimed at the person, because the person is what got blamed.

It would be reassuring to think experience cures this. It doesn't. A 2013 study of admissions officers and hiring managers, people whose job is specifically to assess performance fairly, found that even they take a strong result as proof of ability without discounting for how much easier the situation made that result (Swift, Moore, Sharek, and Gino). The candidate who had the better year because they had the better territory gets credit for the year, not the territory. Expertise doesn't filter this out. It just makes the wrong conclusion look more credible.

Two fields that have nothing to do with each other found the same thing

If this were only a psychology finding, it would be interesting but easy to wave off as soft science. It isn't only a psychology finding.

Aviation, nuclear power, and hospital safety all confronted the same problem decades ago, from a completely different angle: a plane crashes, a patient is harmed, an accident happens, and someone wants to know who made the mistake. For most of the twentieth century, the answer was almost always "the person closest to the failure" — the pilot, the surgeon, the operator. Herbert William Heinrich's 1931 study of industrial accidents, one of the most cited studies in the history of safety science, concluded that eighty-eight percent were caused by unsafe acts on the part of workers. The fix, naturally, was to retrain the workers. (Later researchers challenged both Heinrich's methodology and the specific ratio — but the fact that the finding shaped industrial safety practice for the next several decades regardless is not in dispute, and it's the more relevant fact for this piece's purposes.)

The fix didn't work, because it was aimed at the wrong layer. The framework that eventually replaced it — now standard across high-reliability industries — draws a hard line between two different things: the visible action that happened right before the failure, and the conditions, built years earlier by people who were nowhere near the incident, that made that action close to unavoidable. The visible action is real. It is also, almost always, the last domino — not the first one. The conditions are invisible, structural, and usually the actual cause. Punishing the domino that fell doesn't move the ones standing behind it.

This is not a metaphor borrowed for effect. It's the literal architecture safety science uses to investigate failure in industries where getting the diagnosis wrong costs lives, not quarters. The fact that an organizational diagnosis and an air-crash investigation reach for the same structure is the strongest evidence available that this isn't a quirk of management thinking. It's a pattern that shows up wherever something complex fails and someone has to figure out why.

The fix that doesn't fix anything

Peter Senge gave this failure mode its own shape in systems thinking, separate from psychology and separate from safety science: he called it shifting the burden. When a problem shows up, there's almost always a fast, visible response available, and a slower, harder, structural one. The fast response feels like progress. It relieves the immediate pain. And it quietly trains the organization to keep reaching for the fast response instead of the slow one, because the fast one always seems to work, for a while.

The trap closes slowly. Each round of the fast fix makes the organization a little less capable of doing the harder work, because the harder work never gets practiced. The underlying condition keeps generating the same symptom, on a loop, while everyone involved becomes increasingly confident that they're handling it.

This is the same finding research on training and behavior change keeps producing, with real numbers attached. A meta-analysis covering decades of training-transfer studies found that on average, only ten to fifteen percent of what gets trained ever shows up as sustained behavior change on the job (Blume, Ford, Baldwin, and Huang, 2010). The behavior didn't fail because people forgot it or didn't believe in it. It failed because nothing in the surrounding environment was rebuilt to support it, and an unsupported behavior is exactly as durable as it sounds. Change the structure instead of just the behavior, and the pattern in the literature looks different: interventions that pair new behavior with new environmental support hold up. Without that pairing, decay is close to universal.

What this actually means for the question everyone asks first

None of this says individual accountability doesn't matter. People can be careless. People can be the wrong fit for a role. Sometimes the person actually is the problem, and saying otherwise would be its own kind of dishonesty.

What the research says is narrower and more useful: before settling on a person as the answer, it's worth asking a second question that almost never gets asked with the same energy as the first one. Not just who did this, but what arrangement of resources, incentives, and constraints made this close to the most likely outcome for almost anyone standing in that spot. If the answer to the second question is "nothing in particular — this really was about them," that's a real finding, and it should lead somewhere. If the answer is "the structure pretty much guaranteed this," replacing the person solves nothing. The next person inherits the same structure, and the same outcome, on a delay.

The bias toward the first question over the second isn't a character flaw in any particular leader. It's the default setting of how people process other people's failures, confirmed independently by psychologists studying judgment, safety engineers studying catastrophe, and systems theorists studying why fixes don't hold. The default is strong. It's also wrong often enough, and expensive enough when it's wrong, that it's worth checking before acting on it — every time, not just the times it feels warranted.