Losses loom larger than equivalent gains, skewing risk judgement. From prospect theory: the same outcome is weighed differently depending on whether it is presented as giving something up or as failing to gain it, and the avoidance of loss dominates the pursuit of equivalent benefit.
In intelligence analysis
- Risk judgements are asymmetric: the risk of underestimating a threat attracts more institutional attention than the risk of overestimating it, so assessments drift upwards under scrutiny.
- Warnings are over-produced relative to their base rate, because a missed warning is a visible failure and an unnecessary one is not — the asymmetry is structural, not personal (see Base-Rate Neglect).
- It shapes collection trade-offs: closing a visibility gap feels like a loss, whereas the foregone opportunity of not collecting elsewhere feels like nothing at all.
- The same asymmetry governs disclosure decisions — the cost of over-sharing classified material is immediate and attributable; the cost of under-sharing is diffuse and usually unattributable.
Countermeasure
- State both error types explicitly in the product: the cost of being wrong in each direction, not only the risk of the threat materialising.
- Price the recommendation against a baseline: what would be lost by doing nothing, and what by acting?
Related
- Cognitive Bias — the taxonomy this bias sits within
- Escalation of Commitment — the same attachment, applied to prior investment
- Status Quo Bias — the mirror preference for the existing state
- Estimative Language — forces both error types into the assessment