Trap/Cognitive Bias/No. 0403

Framing Effect

The framing effect is a cognitive bias in which different descriptions of equivalent options change people’s judgments or choices. Studied by Amos Tversky and Daniel Kahneman, it can make the same outcome seem like a gain or a loss, with effects varying by context.

a trap: easy to walk into

Evidence: Well established
Read: 6 min
Links: 10 connections
Useful when: Deciding under uncertainty · Evaluating a claim · Persuading and explaining · Reading data and statistics · Risk and safety

01You've seen this when…

in life
At the grocery store, meat labeled 75% lean seems healthier than meat labeled 25% fat. You put the lean package in your cart without comparing anything else.
at work
A presentation says the service retains 80% of customers each year. The team feels comfortable. Another slide describes the same result as losing one customer in five, and suddenly everyone wants a recovery plan.
out in the world
You support a flood barrier when the notice emphasizes the 90 homes it protects. Hearing that 10 of the same 100 homes remain exposed makes the proposal seem inadequate.

02The idea

The numbers haven’t changed. Your reaction has.

A framing effect occurs when different descriptions of equivalent or closely comparable options change people’s judgments or choices. A treatment can offer a chance of survival or a chance of death. A project can preserve jobs or eliminate them. Each description brings a different part of the situation into view.

The clearest cases hold the underlying outcomes constant. If changing only the description changes your preference, the frame is doing some of the deciding.

This is not the same as loss aversion. Loss aversion means losses tend to hurt more than equivalent gains please. Framing concerns what happens when the description changes, sometimes making the same outcome register as a gain in one version and a loss in another.

Nor does positive wording always win. A loss frame can make people more willing to gamble, support intervention, or take a warning seriously. The direction depends on the decision.

03Why it happens

Different kinds of framing need different explanations. Several processes can contribute:

The description establishes a starting point. Saving 200 people emphasizes improvement from a disastrous baseline. Losing 400 emphasizes a shortfall from keeping everyone alive. This is reference dependence: an outcome’s meaning depends partly on what you compare it with.
Gains and losses can change the appeal of risk. In prospect theory, additional gains or losses have progressively less subjective impact. This diminishing sensitivity can make a guaranteed gain attractive, but a chance to escape a guaranteed loss worth gambling for. It helps explain risky-choice framing, not every framing effect.
A highlighted feature colors the whole judgment. Lean meat brings a desirable attribute to mind; fat brings an undesirable one. The description can change the overall impression even when no gamble is involved.
Wording carries information about the speaker. Someone emphasizing failures may be signaling concern. Someone emphasizing successes may be recommending the option. Listeners can respond to that implied message as well as to the literal numbers.

These processes can overlap. There isn’t one universal rule that predicts every response to every frame.

04A worked example

The best-known demonstration comes from Tversky and Kahneman’s 1981 study. Participants imagine an outbreak expected to kill 600 people and choose between two programs. This is a hypothetical decision, not a report of an actual outbreak.

What it looks like In the gain version, Program A saves 200 people for certain. Program B offers a one-third chance of saving all 600 and a two-thirds chance of saving nobody. Seventy-two percent choose Program A. They appear to prefer a guaranteed rescue over a gamble.

What’s actually going on Another group sees the same programs described through deaths. Under Program A, 400 people die for certain. Program B offers a one-third chance that nobody dies and a two-thirds chance that all 600 die. Only 22% choose Program A. Now the gamble is more popular. Saving 200 out of 600 and losing 400 out of 600 describe the same result; the risky program’s outcomes are also unchanged. The description changes the group’s preference.

What would have helped A comparison that gives both survivor and death counts for every outcome. That makes the real choice visible: a guaranteed partial rescue versus a gamble on rescuing everyone. Neither preference is automatically wrong. Letting the wording switch you between them is the problem.

05How to spot it

One side of the outcome disappears. A pitch gives successes without failures, jobs preserved without jobs lost, or savings without remaining costs.
Your preference changes after a rewrite. You choose differently even though the probabilities, amounts, and consequences stay fixed.
People argue over the label rather than the outcome. Discussion centers on whether something counts as a rescue, a cut, or a loss.
Different audiences receive different descriptions. Supporters hear what the proposal preserves; opponents hear what it sacrifices.

06What to do instead

Write the complementary description. Translate survival into mortality, and describe retention and jobs saved in terms of churn and jobs lost. Keep the denominator and time period identical.
Put both descriptions together. Present both outcomes in one statement: 90 survive and 10 die out of every 100. This makes selective emphasis in competing slogans easier to notice.
Separate outcomes from attitudes toward risk. List what can happen and its probability first. Then decide whether you prefer certainty or a gamble. Matching average outcomes can leave meaningful differences in how desirable the options are.
State your decision criterion before hearing the pitch. You might set a minimum acceptable result or choose between prioritizing expected benefit and guarding against the worst possible outcome. A stable criterion gives the wording less room to steer you.
Check what changed. If the second description changes your choice, identify the new fact or priority behind the change. If you can’t, pause before committing.

A standard comparison table is a simple cognitive forcing strategy: it requires a check instead of relying on your first impression.

07When it isn’t just a change of words

Two descriptions may look equivalent while leaving out different information. A 90% survival rate for healthy patients and a 10% death rate for all patients measure outcomes in different populations. The comparison can also change when the time period shifts or an outcome’s definition differs. Check for these differences before diagnosing a bias.

Frames can also communicate legitimate priorities. Emphasizing the homes left unprotected may draw attention to unequal exposure, helping people assess a flood barrier’s merits even as it influences support. A reader who learns something about the speaker’s concerns may reasonably update their judgment.

Researchers distinguish risky-choice framing from two other forms: attribute framing and goal framing. The patterns vary across these types. The familiar gain-versus-loss reversal is well documented, but its size and direction depend on wording and context. It’s a reason to inspect a decision. Its variability limits its use in predicting everyone else’s behavior.

08Roots

In 1981, Amos Tversky and Daniel Kahneman put a public-health emergency on paper. Six hundred lives were at stake in the scenario. The intervention they tested was a change of words. Their Science paper challenged a basic expectation of rational choice: your preference should stay the same when the same options are described differently.

They had developed prospect theory two years earlier to explain how people judge gains, losses, and risk. Framing showed why the description mattered before the choice was even made: it helped establish what counted as a gain or a loss. Later research separated risky-choice effects from other forms, including the positive or negative presentation of a product’s attributes. The idea traveled into medicine, marketing, policy, and survey design.

An adaptive hypothesis offers another angle: protecting a gain may favor caution, while escaping a serious shortfall may justify risk. That could make flexible risk-taking useful in some environments. But this is a speculative account, not an established explanation for why equivalent wording changes a modern decision.

09How solid is this?

Framing effects have substantial experimental and meta-analytic support. They are not uniform: risky-choice, attribute, and goal framing differ, and wording and decision context affect the size and sometimes the direction of the effect.

10Connections

Often confused with Loss Aversion
Countered byCognitive Forcing Strategy
Part of Choice Architecture, Reference Dependence, Prospect Theory, Preference Construction
See also Anchoring Bias, Diminishing Sensitivity, Risk Aversion, Salience Bias

11Origin and sources

Amos Tversky and Daniel Kahneman’s 1981 paper, The framing of decisions and the psychology of choice, established framing effects as a central topic in decision research.

Suggest an edit· Updated 2026-10-02