This dissertation uses multilevel modeling to examine the effects of corporate and community characteristics on rates of sulfur dioxide emitted by facilities in the electrical power industry. The conceptual framework draws from ecostructural theory to emphasize the social-structural causes of pollution. It also draws from organizational resource dependence theory and the shareholder conception of value. This framework suggests the contemporary transformation in corporate form and the changes in the basic relationship between the corporation and its shareholders have created dependencies, opportunities, and incentives that affect pollution. At the local community level, the conceptual framework also draws from theoretical insights of environmental justice scholars and other scholars in the environmental sociology and social-movement literature. The power plants examined in this dissertation are owned by the largest corporations in the electrical power industry and are located in many different communities across the United States. The multilevel models include three corporate characteristics and four local community characteristics as independent variables. They also include several facility and local community characteristics as control variables. In accordance with ecostructural theory, the findings demonstrate that the total number of subsidiaries in the corporate structure and the dividend payments to shareholders have significant positive effects on the power plant emissions rates. The analysis of community demography shows that relationships involving the power plant emissions rates and percent African Americans, percent families in poverty, and median home values are contingent on the geographic unit of analysis. Hence, the demographic analysis does not consistently support any theory of environmental inequality. On the other hand, all models show that the prevalence of non-profit organizations in the county has a significant negative effect on the power plant emissions rates. This follows in accordance with both ecostructural theory and the path of least resistance theory that underpins the sociopolitical model of environmental inequality. Lastly, all models show that facility control variables involving size, age, and fuel mix have significant effects on the emissions rates. In sum, this dissertation brings together and simultaneously tests theoretical insights from several lines of research to demonstrate that different levels of social structure explain environmental pollution.