The Framing Effect
Logically identical choices made differently depending on wording: gains invite caution, losses invite risk.
The framing effect is the finding that logically equivalent descriptions of the same outcome produce different choices. Tversky and Kahneman's 1981 "Asian disease" experiment is the canonical demonstration: told that 600 people will die in an epidemic, most participants chose the certain program ("200 people will be saved") over a gamble with the same expected value. When the identical programs were described in terms of deaths — "400 people will die" — most participants flipped to the risky option. Nothing about the outcomes changed; only the reference point moved from lives saved to lives lost.
The effect follows directly from loss aversion: outcomes framed as gains invite risk-averse choices, while outcomes framed as losses invite risk seeking, because the sure option in a loss frame feels like a loss — and losses are weighted roughly twice as heavily as gains. The same asymmetry appears in laboratory gambles, in insurance and warranty purchases, and in clinical decisions described as survival versus mortality rates, where both patients and physicians respond to the wording.
Framing works at the level of the whole question, too. Johnson and Goldstein (2003) compared European countries' organ-donation consent rates: opt-in countries averaged single digits to low tens of percent, opt-out countries above 90 — with the default, not attitudes, doing most of the work. Message framing has similar leverage in public health: gain-framed messages ("brushing prevents cavities") tend to work better for prevention behaviors, loss-framed ones ("not screening risks missing cancer") for detection behaviors, though the boundary is fuzzy. Even food advertising exploits the effect — "90% fat-free" outsells "10% fat" for the same product.
The effect replicates reliably in the classic paradigms, but its size is context-dependent: it weakens with domain expertise, repeated choices, and when options are compared side by side rather than sequentially. Framing is not a trick layered onto rational judgment — it is a consequence of how reference points are set, which is why it pairs with confirmation bias as a reminder that the presentation of evidence is part of the evidence.
Tags
behavioral economics decision making prospect theory psychology