Under what conditions is land redistributed to correct for rural land inequality? Many scholars have sought to explain the occurrence of land transfers by thinking about the demand for land by the landless, the historically dispossessed or political elites. I use this dissertation to argue that to understand contemporary land redistribution, which is often voluntary and compensated, we must also consider the supply of land by landholders. Characteristics of agricultural institutions determine whether a commodity sector provides land for land reform. This argument has implications for our understanding of how to resolve rural inequality and historical injustices under a market-based framework. I rely on four subnational case studies of agricultural commodities in South Africa: sugar, fruit, livestock and grain. After apartheid, the South African state promised to redistribute 30% of all agricultural land from white farmers to dispossessed or landless Black farmers. While the program has floundered and failed to meet even half this target, some sectors--especially the sugar sector--have redistributed land at relatively high rates. Under market-based land reform, the characteristics of agricultural institutions affect landholders' supply of land. I develop a typology based on two factors. First, some sectors have actors who depend on the regulation of their sector and thus landholders have an incentive to redistribute land because the state has leverage over them. However, other sectors have deregulated and therefore the state has minimal leverage to encourage land transfers. Second, some sectors centralize power in processors, which give actors collective action capacity to act on such incentives. With decentralized power among producers, sectors will supply land in a more patchwork fashion.