The Wrath of Kahneman

Cass Sun­stein, David Schkade, and Daniel Kah­ne­man, in a 1999 pa­per named Do Peo­ple Want Op­ti­mal Deter­rence, write:

Pre­vi­ous re­search sug­gests that peo­ple’s judg­ments about puni­tive dam­age awards are a re­flec­tion of out­rage at the defen­dant’s ac­tions rather than of de­ter­rence. This is not to say that peo­ple do not care about de­ter­rence; of course they do. Our hy­poth­e­sis here is that they do not at­tempt to pro­mote op­ti­mal de­ter­rence; for this rea­son they do not make the kinds of dis­tinc­tions that are ob­vi­ous, even sec­ond­na­ture, for those who study de­ter­rence ques­tions. Above all, they may not be­lieve that in or­der to en­sure op­ti­mal de­ter­rence, the amount that a given defen­dant is re­quired to pay should be in­creased or de­creased de­pend­ing on the prob­a­bil­ity of de­tec­tion, a cen­tral claim in the eco­nomic anal­y­sis of law.

If we’re af­ter op­ti­mal de­ter­rence, we should pun­ish po­ten­tially harm­ful ac­tions more if they’re hard to de­tect, or else the ex­pected di­su­til­ity of the pun­ish­ment is too small. But ap­par­ently this does not ac­cord with peo­ple’s sense of jus­tice.

Does this mean we should change our sense of jus­tice? And should we ap­ply op­ti­mal de­ter­rence the­ory to in­for­mal so­cial re­wards and pun­ish­ments, such as by get­ting an­grier at an­ti­so­cial be­hav­iors that we learned of by (what the wrong­doer thought was) a freak co­in­ci­dence?