In the past several decades, wildfires in the western U.S. have become more severe, frequent, and damaging. Federal and state governments bear substantial responsibility for managing these incidents. Yet we know little about how government environmental managers make decisions, whether in this context, or in the many other contexts in which government administrators play an important role. In this dissertation, I use the example of federal wildfire management to study decision-making among government environmental managers. In the first essay, I estimate avoided losses to structures due to wildfire suppression. Though preventing losses to structures is a primary goal of wildfire suppression, avoided losses to structures do not justify costs of suppression for many wildfires, especially those that begin in remote areas. In the second and third essays, my collaborators and I explore consequences of behavioral biases among communities affected by wildfire management. In the second essay, we show that, due to pressure individuals place on government administrators, behavioral biases can affect the decision-making of public land management agencies. In the empirical context of this study, government decisions over where to locate wildfire risk reduction projects, this can result in inefficient policy outcomes. The third and final essay uses behavioral bias-induced shocks to community demands for wildfire risk reduction projects to study differences in responsiveness among government administrators to demographically-varying communities. We find that government administrators are more responsive to communities in which a greater percentage of residents are white, educated, or young.