Framing effect

Area: Cognitive biases and heuristics · Evidence: strong (How we rate the evidence). Gain-versus-loss framing replicates in large multi-lab studies and meta-analyses; political emphasis framing is robust in experiments but weaker when audiences hear competing frames, and it fades over time.

In brief

A framing effect occurs when the way a choice or issue is presented changes people's judgments, even though the underlying facts are the same or nearly so. Researchers distinguish two kinds. Equivalence framing describes the same outcome in logically identical terms ("200 saved" versus "400 die"). Emphasis framing highlights some considerations over others ("free speech" versus "public safety"). All political communication frames its subject; the question is whether the frame informs the audience or hides something from it.

What the research shows

The classic study of equivalence framing is Tversky and Kahneman's "Asian disease" problem [1]. Participants imagined preparing for an outbreak expected to kill 600 people. One group chose between a program that would save 200 people for certain and one with a one-third chance of saving all 600 and a two-thirds chance of saving no one; 72 percent chose the certain option. A second group saw the same programs described in terms of deaths: 400 people will die for certain, or a one-third chance that nobody dies and a two-thirds chance that all 600 die. Now 78 percent chose the gamble. The outcomes were identical. Only the description changed, from lives saved to lives lost, and preferences reversed. People tend to avoid risk when options are framed as gains and accept risk when they are framed as losses.

Choices in the "Asian disease" problem: the same two programs chosen differently when described as lives saved (72 percent took the certain option) or as deaths (78 percent took the gamble). Drawn by FRAME from Tversky and Kahneman (1981) [1]. CC BY 4.0.

Irwin Levin, Sandra Schneider and Gary Gaeth showed that "framing" covers several distinct effects [2]. Besides risky-choice framing of the Asian disease kind, there is attribute framing, in which a single quality is described positively or negatively (meat labeled "75 percent lean" is rated better than meat labeled "25 percent fat"), and goal framing, in which the same action is urged by stressing what is gained by doing it or lost by not doing it. The effects differ in size and in why they occur.

Political scientists mostly study emphasis framing. Dennis Chong and James Druckman define it as a speaker's choice of which considerations to put forward, so that the audience weighs those considerations more heavily when forming an opinion [3]. In a well-known experiment, Thomas Nelson, Rosalee Clawson and Zoe Oxley showed participants one of two local news stories about a Ku Klux Klan rally [4]. The facts were the same, but one story framed the event as a free-speech issue and the other as a threat to public order. Viewers of the free-speech version expressed more tolerance for the rally. Kenneth Rasinski found a similar effect from single words in survey questions: Americans were far more willing to increase spending on "assistance to the poor" than on "welfare" [5].

Has it held up?

Equivalence framing is well established. The gain-versus-loss version of the Asian disease problem was among the effects replicated in the first Many Labs project, across dozens of samples [6]. Anton Kühberger's meta-analysis of risky-choice framing studies found a reliable effect of small to moderate size, which varied considerably with how the problem was set up [7].

Emphasis framing is also robust in experiments, with important limits. Chong and Druckman found that when people hear competing frames, as they do in real campaigns, the effects are altered and often reduced, though not eliminated, and that the strength of a frame mattered more than how often it was repeated [8]. Druckman also found that frames work mainly when they come from a source the audience considers credible [9]. Many framing experiments measure opinion minutes after exposure, and later research has found that the effects often weaken within days [10]. For these reasons the rating reflects a strong mechanism whose real-world size is usually smaller than a single laboratory study suggests.

How it shows up in political messaging

• "Daisy," 1964. Lyndon Johnson's campaign did not argue policy detail in its best-known advertisement. It framed the election as a choice about nuclear war, with an image of a child and a countdown, in response to Barry Goldwater's statements on nuclear weapons [11].

Frame from "Peace, Little Girl" ("Daisy"), Lyndon B. Johnson presidential campaign, broadcast 7 September 1964. From the FRAME gallery: Peace, Little Girl ("Daisy")

• The "welfare queen," 1976. Ronald Reagan's campaign speeches described a Chicago woman who had defrauded public-assistance programs using many names and addresses [12]. The story was based on a real but exceptional case whose details the retelling inflated: in 1976 Linda Taylor was charged with defrauding about $8,000 using four aliases, while Reagan's speeches cited 80 names and $150,000 a year [13]. It framed the welfare debate around fraud rather than need, the same shift of emphasis that Rasinski's survey wording produced.
• "Living wage," 2012 onward. Striking fast-food workers and their unions described a $15 minimum as a "living wage" [14], framing the question as what a full-time worker needs to live on rather than as a labor cost.
• Vote Leave, 2016. The Leave campaign in the UK referendum presented a gross EU contribution figure alongside the suggestion that the money could be spent at home, most famously on the National Health Service. The UK Statistics Authority called the figure misleading [15], but the frame turned an accounting question into a choice about hospitals.
• Public health, 2014. The U.S. Food and Drug Administration's "The Real Cost" campaign framed smoking for teenagers as a matter of appearance, showing a young actor's skin peeling away The Real Cost — "Your Skin". It emphasized the consequence most salient to its audience rather than the most common one.

Frame from "Your Skin," The Real Cost campaign, U.S. Food and Drug Administration, 4 February 2014. From the FRAME gallery: The Real Cost — "Your Skin"

Framing through word choice shades into Euphemism / doublespeak, where a harm is renamed so the audience never evaluates it, and Semantic hijacking, where a loaded word is stretched to cover a new set of cases.

How to spot it

• The same fact could be stated as a gain or a loss, a percentage or a count, and only one version appears.
• A single consideration (safety, freedom, fairness, cost) is presented as the whole issue.
• A word carries a verdict ("handout," "investment," "regime," "government") that a neutral synonym would not.
• A striking individual case is used to set the frame for a policy affecting millions.
• A useful check is to restate the claim in its mirror-image frame and see whether it still seems as persuasive.

Persuasion or manipulation?

Framing cannot be avoided: any message chooses what to emphasize, and choosing the considerations one sincerely thinks matter most is ordinary, honest persuasion. The FRAME Writer's toolkit relies on this page for exactly that kind of honest message framing. Framing tips into manipulation when the frame hides information the audience would need to judge fairly, such as the denominator, the tradeoff or the other side's strongest consideration; when an equivalence frame is chosen specifically to trigger risk-seeking or risk-avoidance; or when a loaded label replaces a description so that evaluation never happens. Honest framing leaves the audience able to see what was left out.

Archive techniques that lean on this

• Euphemism / doublespeak — Euphemism / doublespeak: a frame at the level of a single word, chosen so the audience never meets the concrete act.
• Semantic hijacking — Semantic hijacking: captures a word's existing frame and stretches it to cover the speaker's preferred cases.

Related topics

• Anchoring — Anchoring
• Availability heuristic — Availability heuristic
• Priming — Priming
• Negativity bias — Negativity bias
• Moral foundations — Moral foundations

Sources

1. Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology of choice. Science, 211(4481), 453–458. https://doi.org/10.1126/science.7455683
2. Levin, I. P., Schneider, S. L., & Gaeth, G. J. (1998). All frames are not created equal: A typology and critical analysis of framing effects. Organizational Behavior and Human Decision Processes, 76(2), 149–188.
3. Chong, D., & Druckman, J. N. (2007). Framing theory. Annual Review of Political Science, 10, 103–126. https://doi.org/10.1146/annurev.polisci.10.072805.103054
4. Nelson, T. E., Clawson, R. A., & Oxley, Z. M. (1997). Media framing of a civil liberties conflict and its effect on tolerance. American Political Science Review, 91(3), 567–583.
5. Rasinski, K. A. (1989). The effect of question wording on public support for government spending. Public Opinion Quarterly, 53(3), 388–394.
6. Klein, R. A., Ratliff, K. A., Vianello, M., Adams, R. B., Jr., Bahník, Š., Bernstein, M. J., et al. (2014). Investigating variation in replicability: A "Many Labs" replication project. Social Psychology, 45(3), 142–152. https://doi.org/10.1027/1864-9335/a000178
7. Kühberger, A. (1998). The influence of framing on risky decisions: A meta-analysis. Organizational Behavior and Human Decision Processes, 75(1), 23–55.
8. Chong, D., & Druckman, J. N. (2007). Framing public opinion in competitive democracies. American Political Science Review, 101(4), 637–655.
9. Druckman, J. N. (2001). On the limits of framing effects: Who can frame? Journal of Politics, 63(4), 1041–1066.
10. Chong, D., & Druckman, J. N. (2010). Dynamic public opinion: Communication effects over time. American Political Science Review, 104(4), 663–680.
11. Museum of the Moving Image. "Peace, Little Girl (Daisy)," 1964. The Living Room Candidate. https://www.livingroomcandidate.org/commercials/1964/peace-little-girl-daisy
12. NPR. (2013, December 20). The truth behind the lies of the original "welfare queen." NPR Code Switch. https://www.npr.org/sections/codeswitch/2013/12/20/255819681/the-truth-behind-the-lies-of-the-original-welfare-queen
13. Levin, J. (2013, December 19). The welfare queen. Slate. https://www.slate.com/articles/news_and_politics/history/2013/12/linda_taylor_welfare_queen_ronald_reagan_made_her_a_notorious_american_villain.html
14. Democracy Now! (2013, August 2). "We are slowly dying": Fast-food workers launch strike for living wage and right to unionize. https://www.democracynow.org/2013/8/2/we_are_slowly_dying_fast_food
15. Office for National Statistics. Leave campaign claims during Brexit debate (freedom of information release summarizing the UK Statistics Authority's 2016 statements). https://www.ons.gov.uk/aboutus/transparencyandgovernance/freedomofinformationfoi/leavecampaignclaimsduringbrexitdebate

Images

Image	Source	Licence
Bar chart: "Asian disease" gain and loss frames (Tversky and Kahneman 1981)	Drawn by FRAME from source 1 (make_cognitive_bias_diagrams.py)	CC BY 4.0, FRAME
"Daisy" advertisement frame (1964)	FRAME gallery, Peace, Little Girl ("Daisy")	Public domain (published without a copyright notice, per Wikimedia Commons)
"Your Skin," The Real Cost (2014)	FRAME gallery, The Real Cost — "Your Skin"	Public domain (U.S. government work)