CompanyGraph
Why a past investment can distort a present capital-allocation decision, and how a company can make stopping rationally possible.
The Next Dollar Has No Memory
A company may have spent years and millions on a plant, software platform, acquisition, research program, or store network. That spending cannot be recovered by deciding what to do today. The relevant question is narrower: given the asset as it now exists, what will the next dollar of labour, cash, or management attention produce compared with the alternatives?
The sunk-cost fallacy appears when the answer to that forward-looking question is displaced by the emotional or organizational pain of recognizing the earlier loss. A write-off makes the loss visible. Continuing can preserve a story in which the investment eventually pays off, even if the additional commitment has negative expected value. That is why the bias is a capital-allocation problem as well as a psychological one.
What the Research Actually Shows
In controlled experiments, Arkes and Blumer found that people were more likely to continue with a failing course of action when they had already paid for it. Their result establishes a tendency to let prior expenditure influence a current choice; it does not show that every continuing project is irrational or that managers ignore all new information.
Staw’s organizational research describes escalation of commitment as persistence in a failing course when decision-makers have responsibility for the original choice. Career reputation, the desire to appear consistent, and the threat of a public reversal can make continuation safer for the individual than recommending abandonment. Those mechanisms are hypotheses about why organizations continue; they are not proof that a particular company is acting from bias.
How Escalation Becomes a Corporate Process
Escalation rarely arrives as one dramatic decision. A project misses a milestone, the forecast is revised, and a small rescue budget is approved. The revised plan creates a new milestone and a new justification for the following budget. Each review then compares abandonment with a larger visible write-off, while the losses from another year of delay remain distributed across payroll, maintenance, interest, and foregone opportunities.
Project champions matter because the person who proposed a strategy may also be the person who controls the information presented to the board. This does not require dishonesty. Optimistic assumptions can survive because the champion knows the project’s history, has invested reputation in it, and sees more clearly what would be lost by stopping than what another team could do with the resources.
External commitments add a second layer. Customer promises, supplier contracts, debt covenants, public targets, and partnership announcements can make a change of plan costly. The commitment may be economically sensible to end, but termination fees, employee transfers, stranded equipment, or regulatory approvals can make the cash and time required for exit difficult to obtain.
A Famous Case Is an Illustration, Not a Rule
The Concorde programme is often presented as a sunk-cost example because the British and French governments continued development despite escalating costs and a limited commercial market. The history is more complicated: national industrial policy, prestige, technological learning, and strategic relationships were also part of the decision. The case therefore illustrates how political and organizational commitments can keep a project alive; it does not prove that sunk-cost bias alone caused continuation. Britannica’s historical overview records both the aircraft’s technical achievement and its commercial limits.
Stopping Is Not Free
Rational abandonment can require cash before it releases cash. A company may need to settle leases, write down inventory, compensate staff, remediate a site, unwind supplier agreements, or fund customer migration. Those costs are real and may justify a controlled wind-down rather than an immediate closure. They do not, however, make the original expenditure recoverable or justify funding operations whose remaining returns are worse than the available alternatives.
Timing and authority determine whether a stop decision is feasible. A manager may see the deteriorating economics but lack authority to cancel a programme. A board may approve a closure but not have the liquidity to pay termination costs. A business may have cash on the balance sheet while its debt agreements, customer commitments, or regulated approvals restrict how that cash can be used. Capital-allocation discipline therefore requires an explicit exit budget, independent review of the remaining case, and a decision-maker who does not benefit personally from continuation.
What Records Can and Cannot Establish
Budgets show what management planned to spend. Forecasts show assumptions about future revenue, cost, and timing. A board paper can show that a project was approved or that a target was missed. None of those records, by itself, establishes whether continuation is now economically rational. That requires comparing the remaining cash flows and risks with realistic alternatives, including the value of redeploying people, equipment, capacity, and attention.
A write-down can make a loss visible without proving that the underlying project was a mistake at inception. Conversely, a project can remain on the balance sheet while its useful options have narrowed. The investor should therefore read changes in estimates, repeated milestone resets, impairment timing, and capital expenditure together with operating evidence rather than treating any single accounting line as a verdict.
Questions for Investors
- What changed since the original approval? Separate new evidence about demand, cost, technology, regulation, and execution from the amount already spent.
- Who benefits from continuation? Look for governance arrangements that give project champions control over forecasts without an independent review.
- What would stopping require? Identify termination payments, customer migration, remediation, debt restrictions, and the cash and authority needed to exit.
- What alternative use is being displaced? Continuing is a choice to withhold resources from another project, acquisition, maintenance programme, or balance-sheet use.
A disciplined company does not treat every cancellation as success or every continuation as bias. It makes the remaining economics visible, separates accountability for the original decision from the next decision, and funds an orderly exit when that is the higher-value use of scarce resources.