Following Freudenburg’s framework of the “double diversion,” this dissertation aims to understand environmental inequality as the product of two interrelated processes: (1) inequality in the generation of environmental harm, or “disproportionality,” and (2) inequality in the ability to shape discussions about environmental harm through discourse, or “privileged accounts.” I employ a mixed-methods approach in order to assess both disproportionality and discursive power in the debate over coal-fired power in the United States. First, I analyze emissions data at the facility and parent company levels to assess whether a minority of producers is disproportionately responsible for the majority of CO2 generated in the sector. Results indicate that inequality in the generation of emissions is more extreme at the parent company level than at the facility level, with only three companies responsible for the worst 25% of emissions in 2015. Second, I analyze qualitative data from in-depth interviews (n=209) with policy elites at the federal level and in the state of Ohio to identify the dominant narratives and discourse coalitions that shaped the debate over coal-fired power surrounding the 2016 election. I identify the “legitimating discourses” used in support of coal-fired power, then compare these “privileged accounts” to anti-coal counterframes. Discourse analysis findings illustrate how pro-coal interests shifted their discursive strategies to adapt to changing policy contexts, as well as the shortcomings of the anti-coal narratives that sought to shift the discourse toward environmental interests. Finally, to understand the connections between patterns of disproportionality, I explore how the “extreme emitters” identified in quantitative analysis appear within interview data. Together, these analyses illustrate the influence of privileged accounts over the debate, definition, and response to persistent environmental problems.