Escalation as Adaptive Behavior, Not Pure Bias

We have misdiagnosed escalation of commitment as a cognitive failure when it is often a rational response to uncertainty.

The standard narrative is familiar: a decision-maker invests in a failing course of action, then doubles down rather than cut losses. We call this the sunk cost fallacy. Kahneman and Tversky documented it. Business schools teach it as a cautionary tale. The implication is clear—escalation happens because we are irrational, emotionally attached to our choices, unable to think clearly about the future.

But this framing obscures something important. Escalation sometimes works. And when it works, it works because the decision-maker has detected a signal that justifies continued investment.

Consider what actually happens when someone escalates. They are not ignoring new information. They are interpreting it. A project is behind schedule. The team could abandon it, or they could interpret the delay as a temporary friction point—a problem that additional resources might solve. Both interpretations are defensible. The choice between them depends on what you believe about the underlying situation: Is this a fundamentally flawed strategy, or a sound strategy encountering expected obstacles?

This is where the behavioral literature becomes misleading. It treats escalation as though the decision-maker is trapped in a psychological loop, unable to update their beliefs. But many escalators are updating constantly. They are simply updating in a direction that favors continuation. Why? Because they have information—or believe they do—that suggests the project remains viable.

The problem is not that escalators ignore sunk costs. Most sophisticated decision-makers understand sunk cost logic perfectly well. The problem is that they face genuine uncertainty about whether the project will succeed if they persist. In that environment, the question becomes: what does continued investment signal about your confidence in the underlying strategy?

Here is where priming effects matter in ways the traditional bias literature misses. If you frame a decision as a choice between "cutting losses" and "seeing it through," you prime different mental models. The first invokes narratives of prudent withdrawal. The second invokes narratives of resilience and breakthrough. Neither frame is objectively correct. Both are interpretations of an ambiguous situation.

A manager who has been subtly primed toward persistence—through language, through organizational culture, through the way peers discuss similar situations—will interpret the same evidence differently than one primed toward caution. This is not irrational. It is adaptive. It is how we navigate situations where the data underdetermines the decision.

The real insight from Kahneman's work is not that escalation is a bug in human reasoning. It is that our reasoning is sensitive to framing, to the narratives we inherit, to the subtle cues embedded in how choices are presented. We do not fail to update our beliefs. We update them in ways that reflect our prior assumptions about what signals mean.

This has a practical implication that the bias literature often misses: you cannot eliminate escalation by simply educating people about sunk costs. They already know about sunk costs. What you can do is change the signals they receive about what persistence means. If organizational culture frames continued investment as stubborn attachment to a failed strategy, escalation becomes less likely. If it frames it as disciplined follow-through on a sound strategy, escalation becomes more likely.

The choice between these frames is not a choice about rationality. It is a choice about what kind of organization you want to build. Do you want one that cuts quickly and moves on? Or one that commits deeply and works through obstacles? Both have costs. Both have benefits. The escalation you observe is not a cognitive failure. It is the behavioral expression of the cultural signals you have sent.