The Framing Effect

The framing effect is the cognitive phenomenon by which the same information, presented in different ways, produces different decisions. The classic demonstration shows people preferring “90% lean” ground beef to “10% fat” ground beef, even though these describe the same product.

The effect is small in any single instance and substantial in aggregate. Most of the decisions we make are influenced by framing, whether or not we notice the framing at work.

The Asian disease problem

Tversky and Kahneman’s 1981 study is the canonical demonstration. Subjects were told a hypothetical disease threatened to kill 600 people. They were offered two policy options.

Frame 1: Program A saves 200 lives with certainty. Program B has a one-third chance of saving 600 lives and a two-thirds chance of saving none. Most subjects chose A.

Frame 2: Program A kills 400 people with certainty. Program B has a one-third chance of zero deaths and a two-thirds chance of 600 deaths. Most subjects chose B.

These are the same programs. A saves 200 of 600 (equivalent to losing 400 of 600). B has a one-third chance of saving all or losing all. The expected values are identical. The framing — saved versus killed — shifted preferences. Subjects were risk-averse in the gain frame and risk-seeking in the loss frame.

Where it shows up

Framing effects shape political polling, where the wording of questions strongly influences responses. Asking whether the government should “assist the poor” produces different responses than asking whether the government should “provide welfare,” even though many of the underlying programs are the same.

Framing shapes medical decisions, where the description of risks (90% survival versus 10% mortality) influences treatment choices. The same statistic, framed differently, produces different elected treatments.

Framing shapes consumer behavior. A discount “applied at checkout” feels different from a price “before tax.” The total paid is the same. The experience differs.

Why it persists

Framing effects persist because we don’t process information in a frame-independent way. Cognitive psychologists describe two processing systems: a fast intuitive system and a slower analytical system. Most decisions rely heavily on the fast system, which is highly sensitive to surface features of the information including how it’s framed.

Even when we know about framing effects, the fast system still operates. Reading about the Asian disease problem does not eliminate susceptibility to it. The slow system can sometimes intervene, but most decisions are too quick or too low-stakes to engage it.

How to mitigate

For important decisions, the most effective practice is to deliberately reframe the question. If you’re presented with a “90% lean” option, also describe it to yourself as “10% fat.” If you’re considering a treatment with “90% survival rate,” also describe it as “10% mortality rate.”

The reframing exercise often reveals that the original framing was doing work the underlying information didn’t warrant. A preference that survives reframing is more likely to track substance. A preference that flips with framing was probably tracking presentation.

This is a small habit with large aggregate effects. Across hundreds of small decisions, deliberate reframing produces measurably better outcomes — not because it eliminates the bias entirely, but because it catches the cases where the bias was doing the most work.