Wildfire protection strategies (WPS) afford mitigation planners a diverse set of tools to reduce risk to households and communities before, during, and after a wildfire event. However, ex-post effectiveness of WPS are poorly understood. Since the 2003 Healthy Forests Restoration Act, Community Wildfire Protection Plans (CWPP) have become a standard planning and mitigation tool to prevent loss from wildfire events in the United States. Although efforts have been made to document CWPPs and analyze some key aspects of these documents, there remains limited insight into which how CWPPs are developed, which WPS these plans contain, and how effective they have been in preventing wildfire loss. There is also no generalizable, national effort to assess this, which leaves communities without evidence-based research to refine and improve existing and novel plans. This paper proposes an innovative approach to improving our understanding of WPS effectiveness by studying CWPPs at a national level over the full history of these plans. First we build upon the national-level CWPP database to achieve greater spatiotemporal coverage. Then we derive WPS variables from CWPPs using recent advances in natural language processing. Next we build a unique dataset to understand the development and effectiveness of WPS in CWPPS effectiveness by utilizing open source data on wildfire hazard, event characteristics, and environmental context. We then design a series of statistical models to assess CWPPs under various spatiotemporal and treatment effect specifications. Lastly we analyze how trends in CWPP policy and implementation responds to wildfire events and interact with the effectiveness inferred from our statistical models.
