A Criticism of Bernheim & Sprenger's (2020) Tests of Rank Dependence1

Bernheim and Sprenger (2020, Econometrica ; SB) claimed to experimentally falsify rank dependence in prospect theory. This comment criticizes SB’s results and novelty claims. Their experiments only captured well-known heuristics and not genuine preferences. Many more falsifications of rank dependence have been acknowledged before, where SB’s equalizing reductions also have been used before. SB thought to identify probability weighting and utility where they are unidentifiable, which invalidates all SB’s related claims. SB used an incorrect formula of original prospect theory. SB’s suggested alternative of rank-independent probability weighting with dependence on number of outcomes (“complexity aversion”) has long been discarded