Under what conditions do democracies emerge and consolidate? Recent theories suggest that inequality is among the leading determinants of both democratization and consolidation. By contrast, this article argues that inequality harms consolidation but has no net effect on democratization. The author shows that the existing theories that link inequality to democratization suffer from serious limitations: (1) they are useful only for understanding transitions from below and thus do not apply to many other transitions (that is, those from above); (2) even for democratization from below, their predictions are unlikely to hold, since inequality actually has two opposite effects; and (3) they ignore collective action problems, which reduces their explanatory power. However, these objections do not affect the relationship between inequality and consolidation. In particular, while inequality has two opposite effects on the probability of transition to democracy, it unambiguously increases the probability of transition away from democracy. This article conducts the most comprehensive empirical test to date of the relationship between inequality and democracy. It finds no support for the main democratization theories. Contrary to what they predict, estimation suggests neither a monotonic negative nor an inverted U-shaped relationship. Yet inequality increases the probability of backsliding from democracy to dictatorship.