Sunk-Cost Fallacy: The Founder’s Most Expensive Emotion

The sunk-cost fallacy is the tendency to continue an endeavor because of previously invested resources — time, money, or effort — rather than on the merits of what lies ahead. ¹ For founders, the stakes are categorically higher than for most decision-makers: the investment is not just financial, it is biographical. The practical corrective is to evaluate every continuation decision as though the prior investment belonged to a stranger — and to build the structural conditions that make that reframe possible.

Why this is the wrong frame: it is not about waste, it is about identity

The standard framing of the sunk-cost fallacy treats it as a bookkeeping error — a failure to recognize that spent money cannot be unspent. That framing is accurate but insufficient. Normative economic theory is clear: costs incurred in the past are irrelevant to future marginal payoffs. ² The decision that matters is always the next one. Yet the research literature, accumulated over four decades, shows that human beings — including highly intelligent, financially sophisticated ones — systematically violate this principle.

The foundational study was published in 1985 by Hal Arkes and Catherine Blumer in Organizational Behavior and Human Decision Processes. ¹ Their design was elegant: theater subscribers who had paid full price for season tickets attended significantly more performances than those who had received discounts — including performances during bad weather and on inconvenient evenings. ³ The ticket price was already spent. It could not be recovered. Yet it continued to govern behavior months later.

What Arkes and Blumer documented was not irrationality in the colloquial sense. It was a systematic, predictable deviation from economic rationality, driven by a specific psychological motive: the desire not to appear wasteful. ⁴ That motive is social and moral before it is financial. And for founders, it is amplified by a further layer that the theater study could not capture: the original decision was not just made by the subject — it was made publicly, loudly, and at the cost of other opportunities foregone.

The three engines of the trap

1. Loss aversion and the asymmetry of pain

The deepest root of sunk-cost reasoning is loss aversion, the finding by Daniel Kahneman and Amos Tversky that the psychological pain of a loss is approximately twice as intense as the pleasure of an equivalent gain. ⁵ Their 1979 paper, “Prospect Theory: An Analysis of Decision under Risk,” published in Econometrica, established this asymmetry as a robust feature of human judgment — consistent across cultures, income levels, and decision types. ⁶ Kahneman was awarded the Nobel Memorial Prize in Economic Sciences in 2002 partly for this work. ⁶

For a founder, crystallizing a loss — shutting down a product line, firing a team, writing off eighteen months of engineering — registers as a loss of roughly twice the psychological magnitude of the equivalent gain from redirecting those resources. This is not weakness. It is neurology. The asymmetry means that the rational case for stopping must be approximately twice as compelling as the rational case for continuing before the emotional scales balance. Most founders never build that margin into their decision criteria.

2. Personal responsibility and the self-justification loop

Barry M. Staw’s 1981 paper in the Academy of Management Review identified a second, compounding mechanism: escalation of commitment. ⁷ Staw showed that when managers received negative feedback on an initial investment decision, they allocated significantly more additional funds to that investment if they — rather than someone else — had made the original call. ⁸ Personal accountability for the original decision creates an ego investment that compounds the financial sunk cost, making objective reassessment structurally difficult. ⁹

The mechanism is self-justification: continuing the project is the only way to prove, retrospectively, that the original decision was sound. ¹⁰ Founders are the worst possible candidates for objective continuation decisions, because they are always the person who made the original call. The research on escalation of commitment has documented this phenomenon across new product development, information systems, strategic management, and entrepreneurship. ¹¹

3. Identity fusion: when the company becomes the self

The third engine is the most insidious and the least discussed in standard treatments of the fallacy. Sunk costs are particularly powerful when the investment is tied to identity or public commitment. ¹² A founder who has spent five years building a company is not merely weighing financial returns — they are weighing what stopping says about them, about their judgment, about the years of their life devoted to the venture. Admitting the project should end feels like admitting those years were wasted, which triggers not just financial loss aversion but ego threat. ¹²

This is precisely what happened at Intel in the early 1980s. As Japanese manufacturers turned DRAM chips into a commodity product, Intel’s senior management team — including Andy Grove — resisted the exit from memory chips not primarily because the economics were unclear, but because their personal identity had become inseparable from the memory business. ¹³ They continued adding funds to build infrastructure and provide additional training, looking for a solution that would justify their old investments and past choices. ¹⁴ The sunk cost bias delayed the transition to microprocessors — a transition that, once made, produced a 4,500% increase in Intel’s market capitalization during Grove’s tenure as CEO. ¹⁵

Grove’s eventual solution was a reframe that has since become one of the most cited examples of deliberate debiasing in business history. He asked Gordon Moore: “If we got kicked out and the board brought in a new CEO, what do you think he would do?” Moore answered: “He would get us out of memory.” Grove replied: “Why shouldn’t you and I walk out the door, come back in, and do it ourselves?” ¹⁶ By mentally separating themselves from the accumulated investments and reframing as fresh decision-makers, Grove and Moore could evaluate the situation on its forward-looking merits.

The Concorde problem: when sunk costs become policy

The behavioral economics literature has a name for the institutional version of this failure: the Concorde fallacy. The British and French governments continued to fund the development of the Concorde supersonic airliner for years after it became clear the project would never be commercially viable. ¹⁷ Billions had already been spent, and neither government wanted to be the one to write off the investment. ¹⁷ The Concorde entered service in 1976 and operated at a loss for its entire commercial life before being retired in 2003. ¹⁸

The lesson for founders is not that governments are uniquely irrational. It is that sunk-cost reasoning scales with organizational complexity and public commitment. The more stakeholders who know about the investment, the more reputational capital is tied to its continuation, and the harder the exit becomes. A startup that has raised a public seed round, hired a team, and announced a product roadmap has, in structural terms, replicated the Concorde’s bilateral treaty: the irrationality has been written into the social contract.

A 2022 study published in the Journal of Economic Behavior & Organization found that escalation of commitment occurs when decision-makers continue investing in failing projects despite clear signals that termination would be more rational — and that this tendency has been documented across new product development, pricing, entrepreneurship, and strategic management. ¹¹ The phenomenon is not sector-specific or culture-specific. It is a feature of human decision architecture.

A counterintuitive finding: intelligence does not protect you

One of the most practically important findings in the recent literature is that cognitive ability does not reduce susceptibility to the sunk-cost bias. A laboratory experiment published in Journal of Economic Behavior & Organization found strong evidence of the fallacy specifically among high-cognitive-ability subjects. ¹⁹ Higher intelligence, in other words, does not inoculate against the bias — it may even provide more sophisticated rationalizations for continuing a failing course of action.

This matters for founders because the startup ecosystem systematically selects for high-cognitive-ability individuals who are also unusually good at constructing narratives. The same facility with language and pattern recognition that helps a founder raise capital and recruit talent also helps them build a compelling internal case for why the failing product just needs one more quarter, one more feature, one more hire. The bias does not announce itself. It arrives dressed as strategic patience.

Three interventions that actually work

The outsider reframe

Grove’s question to Moore is not merely an anecdote. It is a formalized debiasing technique: mentally substituting a fresh decision-maker for yourself, then asking what that person would do. The research basis is the finding that sunk-cost effects are substantially weaker when the person making the current decision is not the person who made the original one. ⁸ The reframe does not require a new CEO. It requires a deliberate cognitive act of dissociation from the original decision. Practically, this means asking: “If I were evaluating this opportunity cold, with no prior investment, would I commit the next dollar, the next month, the next hire?”

The pre-mortem

Psychologist Gary Klein formalized the pre-mortem technique in a 2007 Harvard Business Review article. ²⁰ The exercise asks a team to assume, as a stipulated premise, that the project has already definitively failed — then work backward to identify every reason why. ²¹ Research on prospective hindsight, the cognitive mechanism the pre-mortem exploits, found that imagining an event has already occurred increases the ability to correctly identify failure reasons by approximately 30% compared to standard forward-looking risk analysis. ²² The technique received broader attention when Kahneman described it in Thinking, Fast and Slow as one of the most practically useful decision tools he had encountered in decades of studying judgment. ²³

For founders, the pre-mortem serves a second function beyond risk identification: it creates a documented record of the failure conditions that were anticipated before the decision was made. When those conditions subsequently materialize, the exit decision is no longer a capitulation — it is the execution of a plan that was written in advance. The founder who pre-commits to exit criteria is not giving up early; they are making the exit decision at the moment of maximum clarity, before the sunk costs accumulate.

Mindfulness and present-moment focus

A 2014 study by Hafenbrack, Kinias, and Barsade, published in Psychological Science, investigated the debiasing effect of mindfulness meditation on the sunk-cost bias. Across four studies — one correlational and three experimental — the results found that increased mindfulness reduces the tendency to allow unrecoverable prior costs to influence current decisions. ²⁴ The mechanism is temporal: mindfulness draws attention away from the past investment and the future anxiety about waste, anchoring judgment in the present-moment evaluation of forward-looking value. ²⁵

This is not a recommendation to meditate before every board meeting. It is a structural insight: decision processes that force present-moment evaluation — zero-based budgeting, rolling 90-day resource allocation, quarterly kill criteria — replicate the cognitive effect of mindfulness at the organizational level. They remove the accumulated weight of prior investment from the decision frame before the meeting begins.

What this means

The founders who build durable companies are not the ones who never make expensive mistakes. They are the ones who recognize, faster than their peers, when a decision is being governed by the past rather than the future — and who have built the structural conditions to act on that recognition before the cost of continuation exceeds the cost of the original error. That is not a personality trait. It is a decision architecture. It can be designed, installed, and practiced.

At FounderWise, the work of Business Growth Accelerator (a FounderWise brand) is built on exactly this premise: that the quality of a founder’s decisions, not the quality of their original idea, is the primary determinant of long-run outcomes. The sunk-cost fallacy is the single most expensive place where decision quality degrades — and the one most amenable to deliberate correction.