I study how landownership affects labor allocation, income distribution, and poverty in less developed countries. I focus on three prototypes of ownership classes: landlords, smallholders, and landless people. Agents are identical except for their ownership of assets. On the basis of optimizing behavior, they divide into urban workers in the modern sector, urban workers in the informal sector, agricultural laborers, subsistence farmers, and landlords. The impact of land reform on production and poverty depends on the amount of fertile land per capita. A more egalitarian distribution of landownership reduces poverty where land is scarce but not where land is abundant.

We investigate the effects of wage compression through centralized collective bargaining when growth depends on the continual reallocation of labor from older, less productive plants to new, more productive plants. We first study the compression of wage differentials that derive from decentralized bargaining in heterogeneous plants. We then consider wage compression when wage difterentials arise from competition among employers over workers of differing quality. We show that wage compression through centralized bargaining can result in higher profits and greater entry of new plants than either decentralized bargaining or a competitive labor market.

To explain the rise in Afghan opium production, we explore how rising conflicts change the incentives of farmers. Conflicts make illegal opportunities more profitable as they increase the perceived lawlessness and destroy infrastructure crucial to alternative crops. Exploiting a unique data set, we show that Western hostile casualties, our proxy for conflict, have a strong impact on subsequent local opium production. Using the period after the planting season as a placebo test, we show that conflict has a strong effect before but no effect after planting, indicating causality.