Good News Is Not a Sufficient Condition for Motivated Reasoning

Abstract People often receive news that makes them feel better or worse about the world around them. This paper studies how news valence affects belief updating. Using large-scale experiments, I test whether people engage in motivated reasoning to overly trust “good” or “bad” news with minimal functional and ego relevance. Results show a precisely-estimated null effect for motivated reasoning in either direction, and modest effects can be ruled out. Complementary survey evidence finds that participants expect good news to increase happiness but not to systematically bias inference. Results suggest that belief-based utility is insufficient for distorting inference to favor preferred beliefs.