Preference evolution, attention, and happiness

We present the first model of preference evolution in an environment where the fame of an agent affects selection. Specifically, agents who are famous are more likely to be selected for comparison by other agents. Agents compare happiness, and then switch preferences if the other agent is happier. Without the attention economy, only 'happy' preferences survive - happy preferences being those which, when followed, give agents maximal subjective well-being. In an attention economy, however, unhappy preferences can persist if there is incomplete information. With incomplete information, agents may make errors when assessing the happiness of others. Furthermore, famous agents may be systematically less happy than others; empirical happiness research suggests that happiness comes from factors like family and religion, not fame. With these two possibilities in play, ordinary agents may be matched frequently with famous people who seem happy but are not. In these matches, ordinary agents will adopt preferences that actually make them less happy, and this allows unhappy preferences to persist in equilibrium. Our model contributes a theoretical explanation for the empirical finding that people who pay more attention to media generally score lower on scales of subjective well-being.