Persisting with a failing line of reasoning, collection effort or product because of what has already been invested in it — the sunk-cost fallacy applied to analysis. The investment may be time, money, an established source, or professional reputation.
In intelligence analysis
- A collection programme continues because of its cost rather than its yield; the effort to have built it is offered as a reason to keep it.
- An analytic line is defended past the evidence because abandoning it would mean conceding the earlier judgement was wrong — which is why the decision to withdraw a judgement should be made structurally, not personally.
- It interacts with Belief Perseverance: the longer a judgement has stood and the more products have carried it, the harder it is to retire.
- In cyber contexts it appears as continued investment in a detection capability or attribution method that has stopped producing, because the sunk cost is visible and the opportunity cost is not.
Countermeasure
- Judge the programme or judgement on forward-looking value only: if it did not exist today, would we build it?
- Set explicit review points with a pre-agreed condition for withdrawal, decided before the investment deepens.
Related
- Cognitive Bias — the taxonomy this bias sits within
- Loss Aversion — the same asymmetry seen from the loss side
- Belief Perseverance — the analytic face of the same attachment
- The Intelligence Cycle — the evaluation and review phase is where withdrawal is meant to happen