Six Hundred Will Die: The Asian Disease Problem and the Physics of Framing

This is the same experiment as chapter 16, and it appears twice in this guide on purpose. There it was a demonstration that preferences are constructed rather than retrieved. Here it is the founding document of linguistic manipulation, because it establishes something stronger than the claim that words matter.

It establishes that a description can be complete, accurate, and non-deceptive, and still determine the answer.

Recall the structure. Six hundred people expected to die. Two programs. Under the first description, Program A saves 200 for certain and Program B has a one-third chance of saving all 600. Seventy-two percent take the certain option. Under the second, Program C means 400 will die and Program D has a one-third chance that nobody dies. Seventy-eight percent take the gamble.

There is no falsehood anywhere in that. Two hundred saved out of six hundred is four hundred dead. Both statements are true, both are complete, and neither omits anything the other includes.

Prospect theory explains the reversal, as chapter 15 describes: the value function is concave for gains and convex for losses, and steeper on the loss side, so the same option is evaluated by a risk-averse system in one frame and a risk-seeking system in the other. What determines which system evaluates it is the reference point, and the reference point is set by the wording.

For the purposes of this part of the guide, the consequence is the important thing.

Every defense most people have against being persuaded is a defense against falsehood. We check facts. We ask for sources. We look for contradictions and omissions. We ask whether the speaker has an interest. All of these are aimed at detecting statements that are not true, and all of them pass a well-constructed frame without noticing it, because there is nothing in it to catch.

This is why frame control is worth more than winning an argument, and why professionals fight over it. Once the terms are set, the argument inside them can proceed honestly and rigorously and still arrive where the frame pointed. The debate about the estate tax and the debate about the death tax are conducted by the same people with the same facts and reach different places. The debate about undocumented workers and the debate about illegal aliens likewise. Aid to families and welfare spending. A public option and a government takeover.

Tversky and Kahneman also showed the effect is not a laboratory artifact confined to undergraduates. A Cochrane review of framing in health communication finds attribute and goal framing effects in real clinical messaging, and surgeons themselves have been shown to make different treatment recommendations when identical outcome data is presented as survival rather than mortality.

The countermeasure is the one in chapter 16 and it is worth repeating because it is the only reliable one available, and it is mechanical rather than a matter of vigilance.

Rewrite the choice in the opposite frame before deciding. Gains as losses, counts as percentages, relative as absolute. Then see whether your answer changes. If it does, you have not yet formed a preference about the world; you have formed one about a sentence, and you now know exactly how much of your position the sentence was carrying.