Institutions continue pursuing failed strategies because they cannot separate the cost already spent from the value of continuing.
Sunk Costs in Institutional Contexts
The sunk cost fallacy — the tendency to continue an action because of the resources already invested rather than the expected future return from continuing — is well-documented in individual decision-making but more consequential in institutional contexts, where it operates through mechanisms that are structurally more powerful than any individual's psychological bias. Individual sunk cost errors can be corrected when the individual recognises the bias. Institutional sunk cost errors are reinforced by accountability structures, reputation management incentives, and the political dynamics that make acknowledging failed investments costly for the actors who made them.
The institutional actor who authorised a significant investment has a specific interest in the investment being perceived as successful — their professional reputation is tied to the investment's outcome in ways that make objective assessment of failure costly. This interest produces the specific institutional sunk cost dynamic: the escalation of commitment to a failing strategy, not because the expected future return justifies escalation, but because acknowledging failure is personally costly for the actors whose decisions created the investment.
The Escalation Cycle
Escalation of commitment follows a predictable cycle. An investment is made based on an expected return. The return fails to materialise, but the failure is attributed to insufficient investment rather than to the strategy's fundamental flaws. Additional investment is authorised to address the shortfall. The additional investment fails to produce the expected return for the same structural reasons that produced the original failure. The failure is again attributed to insufficient investment. The cycle continues until either the sunk cost becomes large enough to force an external review that overrides the internal escalation logic, or the failure becomes too public to be attributed to insufficient investment any longer.
Institutional Mechanisms for Correction
Correcting institutional sunk cost errors requires governance mechanisms that separate the assessment of whether to continue from the institutional interests of the actors who made the original investment decision. Independent review processes, portfolio-level performance assessment that does not attribute individual investment outcomes to specific decision-makers, and explicit pre-commitment to decision criteria that will trigger strategy reassessment — each of these reduces the grip of sunk cost logic on institutional decision-making by creating accountability for forward-looking decisions rather than for the justification of past ones.
The sunk cost is gone whether the institution continues or stops. The question is what the future will bring, not what the past cost. Institutions that cannot make this separation fund failures with good money in perpetuity, and call it commitment.
Discussion