Your critical thinking gym
Home
The Commandments
Join the Club
Your critical thinking gym
● Cognitive bias
← The term was coined by economist Jonathan Baron, psychologist Rajeev Gowda, and economist Howard Kunreuther in their 1993 study of hazardous-waste cleanup attitudes for the journal Risk Analysis. An earlier, purely economic demonstration of the same pattern appeared in W. Kip Viscusi, Wesley Magat, and Joel Huber's 1987 consumer health-risk study.
Preferring to bring one risk all the way down to zero, even when a different option would remove more total risk overall, just not all of it, from any one place.
No widely used alternate name. It's often mentioned alongside the certainty effect, a related but broader idea from prospect theory, not an interchangeable synonym (Section 09).
A preference for reducing a risk all the way to zero, over an alternative that reduces the total amount of risk by more but leaves some of it, in some place, still standing.
Plain version
Say you can spend a fixed budget one of two ways: fully wipe out a small danger, or take a much bigger bite out of a larger one without finishing the job. Most people gravitate toward "fully wiped out," even when the math says the second option saves more lives, prevents more harm, or removes more danger overall. Zero isn't just a smaller number than "a little risk left." It feels like a different category of thing entirely: done, versus not done.
Aliases
None in common use. Some writers use "zero-risk bias" and "the certainty effect" as if they're the same thing, they aren't (Section 09 has the actual distinction).
Related, but not the same thing
The certainty effect, from Kahneman and Tversky's prospect theory, is the broader finding that people weight a move from "probable" to "certain" far more heavily than an equal-sized move somewhere in the middle of the probability scale. Zero-risk bias is the specific real-world version of that pattern that shows up when the "certain" option means eliminating one risk entirely, at the cost of doing less overall. Every zero-risk bias case involves the certainty effect; not every certainty-effect case involves picking a worse overall outcome to get there.
Scope
One of the harder biases on this site to catch yourself doing, because the pull toward "zero" doesn't feel like a shortcut misfiring, it feels like caring about safety. There's no error signal, because from the inside, choosing complete elimination looks exactly like the responsible choice.
Why your brain does this
Staying alert to a threat costs energy and attention that can't be spent on anything else. For most of human history, threats were mostly local and resolvable: a predator is nearby, or it isn't; a fire is burning, or it's out. Once a threat like that was actually gone, it made sense to stop monitoring it completely and move on. A mind that treats "fully resolved" as a distinct, valuable state, worth a real premium over "mostly resolved", is a mind that correctly stops spending resources on yesterday's danger. Zero wasn't an arbitrary number to prefer; it was the signal that vigilance could finally switch off.
Where it misfires
Modern risks are rarely local or binary. Chemical exposure, disease probability, financial exposure, and environmental contamination are usually questions of degree, not presence-or-absence, and they usually can't be driven to a literal, verifiable zero at all. Applying an ancient "fully resolved, stop monitoring" instinct to a world of probabilities and trace amounts means chasing the psychological state of closure even when it costs more total safety than a real, if incomplete, reduction elsewhere would have delivered.
The live academic debate, briefly
The Kahneman and Tversky heuristics-and-biases tradition treats this as a clear deviation from rational, expected-value reasoning: the math says pick the option that removes more risk, and this pattern reliably doesn't. Gigerenzer's ecological-rationality tradition would frame the underlying instinct, an option that ends monitoring entirely is worth a premium, as a reasonable rule of thumb for a world of resolvable, local dangers, one that simply wasn't built for probabilistic, never-quite-zero modern risks. This page doesn't take a side; both traditions agree on the mechanism, they differ on whether to call the shortcut itself flawed or just mismatched to its environment.
How it activates, step by step
1. You're choosing between risk-reduction options that differ in whether they drive some risk all the way to zero or merely shrink it.
2. The "zero" option registers as a different kind of outcome, not just a smaller number, but a state of being finished with the problem.
3. That felt difference (worry present versus worry gone) gets weighted far more heavily than the actual size of the risk reduction.
4. You choose, or support, the zero-risk option, even when a rival option would remove more total danger, because it promises closure, not just improvement.
Zero-risk bias isn't only something people fall into on their own. It's a promise communicators can make deliberately, because "completely safe" is a far more persuasive sentence than "safer than before."
A message first raises the specter of a danger, then offers a solution framed as total elimination rather than reduction, "completely protected," "zero exposure," "eliminates the threat entirely." The zero-risk instinct does the persuading from there: once "fully safe" is on the table, "safer" stops sounding good enough by comparison.
Sustained, exaggerated emphasis on a specific danger, even a rare one, makes the audience specifically crave a zero-risk solution to that one thing, regardless of whether resources spent eliminating it entirely would prevent more total harm somewhere else.
Urgency compresses the time available to compare options properly. Under time pressure, the option that promises a clean, complete resolution is easier to grab onto than one that requires weighing partial, harder-to-picture tradeoffs, so "act now to eliminate it completely" out-competes a more genuinely effective but partial fix.
Outside politics
In February 2010, the UK's Royal Society of Chemistry, working with the Cosmetic, Toiletry and Perfumery Association, publicly offered a £1 million prize for the first genuinely "chemical-free" consumer product, a bounty designed to be impossible to collect, since everything, including water and air, is made of chemicals. The RSC's own research had found that 52% of women and 37% of men surveyed said they actively sought out "chemical-free" products. The prize was a direct response to marketing language that sells the emotional promise of zero, not an accurate description of anything.
What increases it
Strong: proportional framing. People respond far more to the percentage of a specific risk removed (100% of this one) than to the absolute amount of harm avoided across all options on the table. A choice framed as "eliminate this risk entirely" pulls harder than the identical choice framed as "prevent five more cases."
Strong: dread-heavy domains. Risks involving contamination, cancer, poisoning, or involuntary exposure, the kinds of danger psychologist Paul Slovic's risk-perception research has long shown people find viscerally "dreadful" regardless of actual probability, are where the pull toward total elimination is strongest.
Moderate: a small, well-defined target. It's much easier to picture "zero" for one specific, named site, product, or substance than for a diffuse, larger-scale risk, so zero-risk bias shows up more reliably when there's a clean, single thing to point at and say "gone."
What does NOT predict it
Being numerically sophisticated doesn't reliably protect against it. Baron, Gowda and Kunreuther's original 1993 survey deliberately included legislators, judges, business executives, environmentalists, and economists, professionals fluent in exactly this kind of cost-benefit reasoning, and the bias still showed up in a substantial share of respondents.
State-dependent factors
Heightened fear or dread plausibly strengthens the pull toward zero, since the whole appeal of the "zero" option is that it ends the specific worry driving the fear in the first place, though no study isolating fear, stress, or time pressure specifically for this bias (as opposed to risk perception generally) turned up in this research pass. Treat this as a reasonable inference from the dread-risk literature, not a directly tested finding.
The pattern across all four: a research survey of policy professionals, a decades-old federal law, a marketing bounty designed to be unwinnable, and an everyday purchase all land on the same trade, giving up a larger total reduction in risk for the specific, singular relief of a risk being all the way gone.
First documented
The term "zero-risk bias" comes from Jonathan Baron, Rajeev Gowda, and Howard Kunreuther's 1993 paper "Attitudes Toward Managing Hazardous Waste: What Should Be Cleaned Up and Who Should Pay for It?" in Risk Analysis, 13(2), 183-192, built around the two-landfill cleanup scenario described in Section 05.
The earlier empirical groundwork
W. Kip Viscusi, Wesley Magat, and Joel Huber's 1987 paper "An Investigation of the Rationality of Consumer Valuation of Multiple Health Risks," in the RAND Journal of Economics, 18(4), 465-479, demonstrated the underlying pattern a few years earlier without using the same name: in surveys about household products like insecticides, people were willing to pay disproportionately more to eliminate a small risk entirely than to achieve an equal or larger reduction that still left some risk behind.
Theoretical grounding
Both studies connect back to Daniel Kahneman and Amos Tversky's prospect theory (1979), specifically the certainty effect: the finding that people weight a shift from probable to certain far more heavily than an equally sized shift somewhere in the middle of the probability range. Zero-risk bias is what that general pattern looks like when applied to real safety and cleanup decisions.
Cross-tradition note
See Section 02 for Gerd Gigerenzer's ecological-rationality framing, the relevant non-English research tradition for this page's adaptive-origin section.
Score: 1 out of 5 (see the breakdown in the card after Section 01).
The in-the-moment question
Am I choosing this option because it removes the most total risk, or because it's the one that lets me stop thinking about this specific danger? If I'd take a different answer once the word "zero" is off the table, that's the bias talking.
Why this is so hard to catch
The pull toward zero doesn't feel like a mistake, it feels like caring properly about safety. There's no internal alarm to notice, because from the inside, wanting a danger completely gone looks exactly like the responsible, careful choice, not like a shortcut overriding the actual numbers.
Debiasing research note
The evidence here leans structural rather than psychological. Baron, Gowda and Kunreuther's own professional-respondent sample suggests that simply being expert or numerate doesn't switch the bias off, which is why their proposed fix targets the decision process itself (forcing a shared unit of comparison) rather than individual willpower or awareness. No dedicated study testing a training intervention that reliably reduces zero-risk bias at the individual level turned up in this research pass; treat the individual-tier advice above as a reasonable habit, not a proven fix on its own.
Commonly co-occurs with
Loss aversion supplies part of the emotional charge, an existing risk framed as something you're currently exposed to feels like a standing loss, and fully removing it feels like the only way to truly stop the bleeding. The framing effect determines how easily the bias gets triggered in the first place: describe an option as "eliminates the risk" and it pulls harder than the mathematically identical option described as "reduces the risk by 90%." Affect heuristic also plays a role: the stronger the gut-level dread a risk produces, the more the "gone completely" option out-competes a larger but partial reduction.
Often mistaken for
Probability neglect is ignoring how likely or unlikely an outcome is altogether once it's vivid or emotionally charged enough, focusing entirely on the worst case regardless of its odds. Zero-risk bias is narrower and more specific: it's about which of several risk-reduction options you pick, favoring the one that reaches exactly zero even when a rival option removes more total risk. Probability neglect can make a risk feel worth worrying about in the first place; zero-risk bias is what happens next, once you're deciding how to respond to it.
Compounding effect
Zero-risk bias tends to feed negativity bias in resource-allocation settings: once one danger has been fully eliminated, remaining risks that were never brought to zero can start to feel disproportionately alarming by comparison, even if they were always smaller than the one that just got the "complete" treatment.
Primary sources
Accessible reading