Cognitive Biases · CB-39
People will pay disproportionately more to eliminate a small risk entirely than to achieve an equal or larger reduction in a bigger risk — completely eliminating something feels categorically better than merely reducing it, even when the numbers say otherwise.
A preference for completely eliminating a small risk over achieving a proportionally larger reduction in a bigger risk, even when the larger reduction would save more expected lives, money, or harm overall — driven by the psychologically satisfying, categorical certainty of 'zero' versus the merely incremental feeling of 'less.'
Documented in behavioral economics research including studies by Kahneman and Tversky's broader prospect theory framework, and analyzed extensively in risk-perception research examining public preferences for environmental and safety regulation spending.
The Mechanism
Eliminating a small risk beats a bigger reduction of a larger risk, in preference — not in lives saved
Survey respondents asked to choose between public safety spending options consistently preferred fully eliminating a small, rare risk over a much larger reduction in a bigger, more common risk — even when the larger reduction option would save substantially more expected lives, the psychological appeal of reaching exactly zero dominated the actual numbers.
01 · 'ZERO' HAS A CATEGORICAL, QUALITATIVE APPEAL THAT 'LESS' DOESN'T MATCH
Complete elimination removes the need to think about the risk at all, unlike any partial reduction
Reducing a risk from 10% to 3% still leaves something to worry about and account for; eliminating a risk from 0.1% to 0% removes the category of concern entirely — this qualitative shift from 'some risk' to 'no risk' carries psychological weight well beyond what the numerical difference in expected harm would justify.
02 · IT SYSTEMATICALLY DISTORTS PUBLIC POLICY AND REGULATORY SPENDING PRIORITIES
Resources get allocated toward achieving zero on rare risks rather than toward the largest expected-harm reduction
Regulatory economics research has repeatedly found that public and political preference for eliminating a specific, salient risk entirely can direct spending toward interventions with a far higher cost per expected life saved than alternative interventions that would reduce a larger, more diffuse risk by a smaller relative amount — a real, costly consequence of the bias at a policy scale.
03 · IT INTERACTS WITH HOW VIVIDLY AND SPECIFICALLY THE RISK IS FRAMED
A named, specific, identifiable risk gets disproportionate attention relative to a diffuse, statistical one
Zero-risk bias tends to be strongest for identifiable, specific risks (a particular chemical, a named hazard) rather than diffuse statistical risks of similar or greater expected harm — closely related to the broader tendency for vivid, specific threats to receive disproportionate attention and resources compared to abstract, statistically larger ones.
Where It Fails / Inversion
Where it fails / inversion
Achieving true zero risk sometimes genuinely does carry disproportionate value beyond the pure expected-harm calculation — for catastrophic, irreversible risks (a small chance of an unrecoverable outcome), completely eliminating the possibility can be rationally worth more than the raw expected-value math suggests, distinct from the bias's more common, costly application to ordinary, recoverable risks.
How To Use It
Worked example · allocating a safety or risk-mitigation budget rationally
Before committing disproportionate resources to fully eliminating one specific, salient risk, calculate the expected harm reduction per dollar spent across all the available risk-mitigation options, including options that would only partially reduce a larger risk — and be explicit about whether the case for prioritizing complete elimination rests on the actual math, or on the psychologically appealing feeling of reaching exactly zero.
How to use it
Before prioritizing a spending or effort decision aimed at completely eliminating a small risk, calculate the expected-harm reduction per unit of resource spent, and compare it honestly against alternative options that would only partially reduce a larger risk — resist letting the psychological appeal of 'zero' override a clear-eyed comparison of the actual numbers.
See Also