To support improved wildfire incident decision-making, in 2017 the US Forest Service (Forest Service) implemented risk-informed tools and processes, together known as Risk Management Assistance (RMA). The Forest Service is developing tools such as RMA to improve wildfire decision-making and implements these tools in complex organizational environments. We assessed the perceived value of RMA and factors that affected its use to inform the literature on decision support for fire management. We sought to answer two questions: (1) What was the perceived value of RMA for line officers who received it?; and (2) What factors affected how RMA was received and used during wildland fire events? We conducted a qualitative study involving semi-structured interviews with decision-makers to understand the contextualized and interrelated factors that affect wildfire decision-making and the uptake of a decision-support intervention such as RMA. We used a thematic coding process to analyze our data according to our questions. RMA increased line officers’ ability to communicate the rationale underlying their decisions more clearly and transparently to their colleagues and partners. Our interviewees generally said that RMA data analytics were valuable but did not lead to changes in their decisions. Line officer personality, pre-season exposure to RMA, local political dynamics and conditions, and decision biases affected the use of RMA. Our findings reveal the complexities of embracing risk management, not only in the context of US federal fire management, but also in other similar emergency management contexts. Attention will need to be paid to existing decision biases, integration of risk management approaches in the interagency context, and the importance of knowledge brokers to connect across internal organizational groups. Our findings contribute to the literature on managing change in public organizations, specifically in emergency decision-making contexts such as fire management.
A pupillary index of susceptibility to decision biases
