Biased recommendations from biased and unbiased experts

Abstract

When can an expert be trusted to provide useful advice? We experimentally test a simplified recommendation game where an expert recommends one of two actions to a decision maker who may take either action or instead pursue an outside option. Consistent with predictions from the cheap talk literature, we find that decision makers partially discount recommendations for the action a biased expert favors, but that recommendations can still be persuasive in that they reduce the chance of the outside option. If the decision maker is uncertain whether the expert is biased toward an action, biased experts lie even more, while unbiased experts follow a political correctness strategy of recommending the opposite action so as to be more persuasive by appearing unbiased. Even if experts are known to be unbiased, experts pander by recommending the action that the decision maker already favors, and decision makers discount the recommendation. The results highlight that transparency of expert incentives can improve communication, but need not ensure unbiased advice.