Extension neglect
Judging a set by a typical member, not by how many members it has.
What it means
Extension neglect is the tendency to judge a set by a representative case while all but ignoring its size — the number of items, people, or units at stake. Asked a hard question about an aggregate, the mind substitutes an easier one about a single prototype and reads value off that image, so willingness to pay or felt concern stays nearly flat even as the underlying quantity climbs a hundredfold. The neglect is a default of intuition rather than an incapacity: point attention directly at the count and sensitivity largely returns. Extension neglect is the umbrella mechanism behind scope insensitivity and duration neglect (both well replicated); it has also been invoked to explain the disproportionate pull of a single identifiable victim, though large pre-registered replications have since failed to find that effect reliably and flagged publication bias in the original literature.
Valuation by prototype
The mechanism Kahneman named is attribute substitution: asked a hard question about a whole set, the mind quietly answers an easier one about a single representative case. To judge how bad an oil spill is, it conjures one oiled bird and reads the horror off that image; the number of birds is a separate fact that must be retrieved and multiplied, work the intuitive system does not do on its own. Because a prototype has no size — one image stands in for two thousand cases as readily as two hundred thousand — the resulting feeling is flat across magnitudes. Extension enters a judgment only when attention is deliberately pointed at it, which is why the neglect is a default rather than an incapacity.
What the evidence shows
The signature result is a willingness-to-pay curve that barely moves with scale. In the migrating-birds survey, stated payments to prevent 2,000, 20,000, and 200,000 deaths were about $80, $78, and $88 — statistically flat across a hundredfold range. Duration neglect shows the same flatness in time: in the cold-pressor study a majority chose to repeat a longer, objectively worse ordeal because its final seconds were milder, and in the colonoscopy trial patients' remembered pain tracked the average of the peak and final moments while total length barely registered. Across these paradigms the felt or remembered value correlates strongly with a salient moment or image and close to zero with the aggregate — the empirical fingerprint of extension neglect.
Limits, replication, and debiasing
Not every member of the family is equally solid. Scope insensitivity in stated valuation and duration neglect in remembered pain replicate well, but the most emotionally vivid case — the single identifiable victim — has fared badly under scrutiny. A large pre-registered replication of Small, Loewenstein and Slovic found no reliable identifiable-victim effect and evidence of publication bias in the earlier literature. The neglect also lifts when extension is made unavoidable. Presenting magnitudes as a graph rather than a bare number, or evaluating small and large cases side by side rather than one at a time, restores much of the missing sensitivity, because joint and visual formats force the aggregate into view. Extension neglect is therefore largely a property of how a question is framed, not a fixed limit of judgment.
Where it shows up
The bias bites wherever a value ought to scale with a count but is elicited one case at a time. Contingent-valuation surveys used to price environmental damage produce numbers that ignore how much is at stake, which is why regulators treat single-scenario willingness-to-pay with caution. Charitable appeals lean the other way on purpose, replacing a statistic with one named face because the aggregate cannot generate feeling on its own. Experience and service design exploits the time version: a flight, a hospital stay or a support call is remembered by its worst moment and its ending, so a graceful finish can outweigh hours of mild friction. Public-health and disaster triage face the hardest form, where resources should follow the numbers but attention follows the prototype.
Examples
People report similar willingness to pay to save 2,000, 20,000, or 200,000 migrating birds, because each scenario evokes the same prototype image of a single drowning bird.
A named child needing surgery draws donations within days, while an appeal for a famine affecting millions raises far less. The larger number never reaches the feeling that drives giving.
A colonoscopy that lasts longer but ends gently is remembered as less unpleasant than a shorter one stopped at peak pain. The extra minutes barely register in the verdict.
Told a data breach exposed either ten thousand or ten million records, people voice much the same alarm; the felt violation attaches to 'my data leaked,' not to the size of the count.
A three-day getaway ending with a perfect sunset dinner is recalled more warmly than a two-week trip that petered out in airport delays; remembered joy tracks the peak and the finish, not the days.
First described in Daniel Kahneman & colleagues; Kahneman (2000).
Key references
- Maier, M., Wong, Y. C., & Feldman, G. (2023). Revisiting and rethinking the identifiable victim effect: Replication and extension of Small, Loewenstein, and Slovic (2007). Collabra: Psychology, 9(1), 90203. doi.org/10.1525/collabra.90203
- Small, D. A., Loewenstein, G., & Slovic, P. (2007). Sympathy and callousness: The impact of deliberative thought on donations to identifiable and statistical victims. Organizational Behavior and Human Decision Processes, 102(2), 143-153. doi.org/10.1016/j.obhdp.2006.01.005
- Kahneman, D., Ritov, I., & Schkade, D. (1999). Economic preferences or attitude expressions? An analysis of dollar responses to public issues. Journal of Risk and Uncertainty, 19(1-3), 203-235. doi.org/10.1023/A:1007835629236
- Redelmeier, D. A., & Kahneman, D. (1996). Patients' memories of painful medical treatments: Real-time and retrospective evaluations of two minimally invasive procedures. Pain, 66(1), 3-8. doi.org/10.1016/0304-3959(96)02994-6
- Kahneman, D., Fredrickson, B. L., Schreiber, C. A., & Redelmeier, D. A. (1993). When more pain is preferred to less: Adding a better end. Psychological Science, 4(6), 401-405. doi.org/10.1111/j.1467-9280.1993.tb00589.x