The origins of hits: Cumulative advantage vs. multiplicative returns in cultural markets

Abstract

The popularity of cultural objects is often distributed as many unpopular “flops” alongside a few “hits.” Hits can be several orders of magnitude more popular than typical objects but are difficult to predict ex-ante. Most explanations focus on cumulative advantage (CA): rich-get-richer processes wherein the success of cultural objects breeds future success, creating high inequality in popularity and decoupling popularity from the properties of objects. I present an additional formal model, multiplicative returns (MR), which assumes that cultural objects are judged as interdependent ensembles of their properties. Like CA, the MR model produces “hit” objects whose popularity is difficult to predict before the fact. I show that the MR model generates popularity distributions consistent with that of US baby girls’ names. I conclude that highly unequal popularity distributions of cultural objects are over-determined: they can arise from MR, CA, or both.