Sunk-Cost Fallacy

A project is late, over budget and missing the original strategic need it was meant to address. Still, the leadership team keeps funding it because "we've already spent too much to stop now." That statement feels practical, but it describes a decision error. The money already spent cannot be recovered by continuing. What continuation actually does is place new resources at risk in the hope of redeeming old ones. The sunk-cost fallacy is the tendency to let irrecoverable past investment govern a forward-looking choice.

What sunk cost means

A sunk cost is a cost already incurred that cannot be changed by the current decision. Money spent on a failed build, months invested in a stalled initiative, reputational capital used to back a strategy, or effort already devoted to a hiring process are all examples. Once the cost is sunk, the relevant question is not how to get it back through persistence. It is what choice now creates the best future outcome.

The fallacy arises when leaders treat sunk cost as if it were recoverable through continued commitment. They reason that abandoning the initiative would "waste" what has already been spent, so they keep going even when the expected future case is weak. In reality, stopping may prevent further waste, while continuing may compound it.

This is why the sunk-cost fallacy is not about accounting. It is about decision framing. Past spending belongs in postmortems and learning reviews. It does not belong in the direct logic of whether to commit the next unit of resource.

Why the bias is so persistent

The bias persists because stopping is psychologically and politically hard. Ending an initiative can feel like admitting error, exposing a sponsor, disappointing a team, or making earlier sacrifice look pointless. Continuing preserves hope. It also preserves the story that success is still just over the next hill.

Organizations add further pressure. Projects accumulate champions, budgets, dependencies and symbolic value. The more visible the initiative, the harder it becomes for leaders to separate the future business case from the identity attached to having chosen it. Teams may also confuse effort with entitlement, believing that because they worked hard, the initiative deserves more time.

These pressures are understandable, but they distort capital allocation. Future money, attention and talent become instruments for protecting the emotional meaning of past decisions rather than for pursuing the best available option from today onward.

Past investment is not irrelevant

A common misunderstanding is that sunk costs should be ignored in every sense. That is too simple. Past investment matters for learning. It can reveal what the organization has discovered, what capabilities it has built, what assets are reusable and what assumptions proved false. It also matters for accountability:

leaders should understand why the investment was made and whether governance failed

What sunk costs should not do is serve as the main reason to continue. The relevant continuation question is forward-looking: given what is now known, what are the expected future costs, benefits, risks and alternatives? If the project were not already underway, would the organization choose to start or continue it from this point?

This distinction preserves learning without allowing history to dominate resource allocation. A project can produce valuable knowledge and still be the wrong place to invest next.

How escalation happens

Escalation of commitment often begins with a plausible rationale. The project is behind, but one more quarter may reveal the payoff. The market turned temporarily, but recovery may come. The team has already built most of the core system, so stopping now seems premature. Each step feels incremental. The cumulative effect is that the organization keeps funding an initiative long after the expected value has turned negative.

Escalation is especially likely when milestones are vague, success is easy to redefine and no one wants to own the shutdown decision. A project then survives by narrative flexibility. Setbacks become proof that more support is needed. Delays become signs that the effort is too important to abandon. Each added commitment is justified partly by the one before it.

This is why sunk-cost fallacy often overlaps with governance failure. Without explicit decision rules and independent challenge, teams can continue not because the case is strong but because stopping has become socially harder than continuing.

Where organizations fall into the trap

Capital projects are a classic example. Construction overruns, software implementations, acquisitions and turnaround programs can all develop a momentum that persists after the original economics deteriorate. Product portfolios show the same pattern when declining offerings keep receiving support because they once mattered strategically or because teams have invested years building them.

Hiring decisions can also suffer from sunk-cost thinking. A long recruitment process may make leaders reluctant to reject a final candidate who no longer looks right. Litigation, partnerships and market entries all create similar pressure:

once the organization has publicly committed, retreat feels more expensive than reevaluation

The trap is strongest where status, identity and visibility are intertwined with the investment. The more a decision has become part of who the organization believes itself to be, the easier it is to keep paying for the story.

Design reviews around the next decision

The best antidote is to redesign governance so that continuation is treated as a fresh choice rather than the default. Stage gates, pre-defined milestones and independent reviews help because they shift the question from "how much have we already spent?" to "what is the expected value of the next commitment?" That reframing is central.

A good review asks a small set of hard questions.

1. What would be the expected future value if the initiative started from this point today?
2. What additional resources are required and what are the opportunity costs of using them here?
3. Which earlier assumptions have failed and which are still necessary for success?
4. What assets or learning can be salvaged if the initiative is stopped or redesigned?
5. What objective exit criteria would trigger pause, pivot, or termination?

These questions allow leaders to preserve what is reusable without pretending that more spending redeems prior spending.

Make stopping legitimate

Organizations often celebrate persistence so strongly that stopping looks like weakness. That norm is useful when the main challenge is short-term discomfort. It is damaging when the real challenge is objective reassessment. Mature cultures distinguish disciplined persistence from costly denial.

This requires visible examples. When leaders shut down a failing initiative and explain the reasoning in forward-looking terms, they teach the organization that stopping can be a mark of judgment rather than failure. When they salvage useful assets and reassign talent effectively, they reduce the fear that ending a project erases all prior contribution.

Leaders should also avoid punishing the messenger who surfaces deterioration. If people learn that honest downside information threatens careers, sunk-cost escalation becomes easier because the casefor continuation never gets tested cleanly.

Past investment should not determine whether a failing initiative continues. The right decision rule is simple but demanding:

evaluate the next commitment on its future merits, use prior investment for learning rather than justification and make it institutionally acceptable to stop when the expected future case no longer holds