I’ve always found the compliance machinery inside large financial institutions fascinating, mostly because of how elaborate it is. Every year, these firms spend tens of millions of dollars on training. Employees sit through ethics modules, data‑security refreshers, risk‑mitigation videos, and the inevitable quizzes at the end. They click through attestations confirming they understand the rules. The underlying assumption is straightforward: if you teach people the right behaviors, the culture will follow.

But if you look at how these same institutions actually operate, the incentives tell a different story. The behaviors that get rewarded — aggressive lending, creative interpretations of policy, shortcuts that boost quarterly numbers — rarely match the behaviors emphasized in the training. In many cases, the people who advance the fastest are the ones who treat the rules as flexible guidelines rather than constraints.

When something goes wrong, leadership points to the training logs. “We told them what to do,” the argument goes. “They chose not to.” It’s a neat narrative, but it rests on a flawed assumption: that what an institution says it values is a reliable indicator of what it actually values.

In practice, people don’t follow instructions. They follow incentives.

If this dynamic feels familiar, it’s because it shows up more often than it seems — and once you see it, you start noticing it everywhere. Subscribe for new Trust Signal essays twice a week.

And when incentives conflict with the rules, the incentives win.

This is what I think of as Safety Theater — the appearance of control without the underlying structure to enforce it. It’s much easier to produce training modules than to redesign the incentive architecture that drives day‑to‑day behavior. Training is visible. Incentive reform is expensive, politically difficult, and often slows down the very performance metrics leadership is rewarded for.

Once you see the structure, the outcomes stop being surprising. If following a compliance procedure slows you down, and bypassing it leads to better evaluations, higher bonuses, and faster promotions, the rational choice is obvious. Over time, employees learn to read the organization not through its stated values but through its reward system.

In effect, they become behavioral sensors. They watch who gets promoted. They watch who gets protected. They watch who gets quietly sidelined. If the person who bends the rules to hit a target is celebrated, while the person who follows the rules precisely is treated as slow or difficult, the message is clear. The training may say one thing, but the system is communicating something else entirely.

Trust, in this context, isn’t about what people say. It’s about what they rely on when it matters. When employees choose between compliance and performance, their choice reveals what they believe the institution truly values. Those choices, repeated across thousands of decisions, become the real operating logic of the organization.

Why does this persist? Because changing the incentive structure requires confronting trade‑offs most institutions would rather avoid: slower growth, lower short‑term profitability, and more complex management. Training, by contrast, is a perfect substitute for structural change. It signals seriousness to regulators and the public without altering the underlying system.

The result is a widening gap between narrative and reality. Institutions present themselves as risk‑conscious and ethically grounded, while the behaviors they reward tell a different story. When scandals occur, they’re treated as isolated failures rather than predictable outcomes of the system itself.

If you want to understand what an institution actually values, don’t read the policies. Look at the reward structure. Who advances? Who gets protected? Who gets punished? Those are the real indicators of trust allocation inside the system.

There’s a simple diagnostic I keep coming back to: find the person who was penalized for achieving strong results while violating internal guidelines. If that example is hard to find, the guidelines aren’t functioning as constraints. They’re functioning as narrative.

And that’s the point. The system isn’t failing when a scandal happens. It’s doing exactly what it was designed to do. The incentives are the truth. Everything else is theater.

What makes this difficult to see is that nothing here feels like a decision. No one explicitly chooses these outcomes. The shift happens quietly, through what gets rewarded and what gets ignored.

And once that pattern is in place, it doesn’t stay confined to a single context. It shows up across different systems—wherever behavior is shaped under pressure and constraint.

If this way of seeing things is useful, you’ll likely want the next few essays.

I publish Trust Signal twice or thrice a week. Subscribe below!