Treating the absence of evidence as evidence of absence — reading "we found nothing" as proof that nothing exists. In intelligence the error runs in both directions, and the second direction is the more dangerous: failure to find a capability is read as concealment rather than as weak collection.
In intelligence analysis
- The fallacy is a property of the collection, not of the target. No finding may mean the capability is absent, the source is silent, the sensor did not cover it, or the question was never tasked.
- A denied or unanswered request from an adversary is interpreted through the prevailing hypothesis: non-cooperation becomes concealment, whereas the same behaviour under a different hypothesis reads as a state with nothing to declare.
- Its mirror image — absence of evidence as positive proof of concealment — is the more common failure in practice, because it manufactures confidence instead of draining it.
- Confidence statements must distinguish "we assessed X" from "we did not look", or the two become indistinguishable in the product.
Countermeasure
- State every judgement of non-existence together with the collection effort that supports it; a negative finding without a collection record is an assertion.
- Ask the reverse question explicitly: if the capability existed, what would we expect to see, and did we look there?
Related
- Cognitive Bias — the taxonomy this bias sits within
- Confirmation Bias — supplies the hypothesis that gives absence its meaning
- Structured Analytic Techniques — assumption checks surface unexamined negative findings