The oft-observed persistence of economic inequality in society after democratization has spurred growing discontent with democracy in many parts of the world and pose an important puzzle to social scientists. This study seeks to explain why democracy performs poorly in some instances but successfully in others. To do so, I examine democracy's heterogeneous effect on how income is initially distributed in the market before taxes and transfers. Thus far, the existing literature on the democracy-inequality nexus has disproportionately concentrated on the redistribution of income through taxes and transfers. Accordingly, researchers have generally equated lack of improvements in equality with the failure to redistribute progressively and attempted to explain such failure by pointing to elites' capture of democracy. However, the almost exclusive focus on redistribution obscures the fact that market income distribution is logically and temporally prior to redistribution and, as such, shapes the policy leeway available for taxes and transfers. In highlighting the significance of democracy's effect on market income distribution, this study seeks to improve our understanding of the challenges and opportunities in redressing inequality in democracies.Democracy, I argue, can reduce inequality of market income but only under the appropriate conditions. By enhancing the protection of civil liberties, democracy strengthens the bargaining power of workers and increases the number of workers covered by a collective wage bargain. Such a change holds the potential to lessen inequality of market income, but democracy's effect on inequality is conditioned by trade openness and level of wage bargaining and becomes muted under high trade levels and decentralized wage bargaining. In examining how democracy interacts with trade openness and the level of wage bargaining to influence market income distribution, this study moves beyond the still-unresolved debate on whether democracy affects inequality, suggesting that a more fruitful research agenda involves asking under what conditions democracy can and cannot bring down inequality.To test my arguments, I integrate quantitative and qualitative methods. I employ data on labor's share of national income and personal income inequality before taxes and transfers to run statistical analyses that model the conditional effects of democracy. Furthermore, an in-depth study of South Korea, based on original primary sources, offers the process-tracing evidence that since the 1980s, political contention over market income distribution among labor unions, businesses, and the government has been influenced by the interactions between democratization and trade openness and the level of wage bargaining as theorized.