New Approaches to Rating Public Fire Protection

New Approaches to Rating Public Fire Protection

Phil LeGrone | Vice President of Data & Actuarial Solutions


Fire is a large, consistent source of covered loss that must be effectively evaluated to ensure long-term profitability for insurers. Insured fire losses total more than $22 billion every yeari, so it’s critical insurers have the data to understand how best to insure for loss as well as for developing thorough risk-analysis tools for effective underwriting decision making.

Using exciting new data-driven risk modeling and a unique approach to data measurement and analysis, AAIS is making a difference for our members.

Using Data to Quantify Total Fire Risk
AAIS has taken on the challenge of developing a new methodology to quantify variations in claim severity related to public fire protection response. The initial work is focused on the efficacy of the data itself.

Currently, AAIS is exploring many datasets, concentrating on information from 10 states between incident years 2012 to 2016 to better understand the variation in public fire protection that impacts claims severity. The team has reviewed more than a dozen distinct data sources thus far, including the National Fire Incident Reporting System (NFIRS) database managed by the United States Fire Administration, data from the US Census Bureau, the United States Geological Survey public data sharing resource, AAIS member data, and more.

A key dataset comes from NFIRS. The raw NFIRS data in the initial ten-state area consists of more than 150 data elements and approximately three million reported fire incidents. AAIS is studying the data, including fire station response times, number of apparatus and number of responding personnel, to build its new methodology. 

Predictive Data and CommunityRisk Reduction
AAIS has found that just looking at ‘classic’ data measurements, like distance to a fire department, isn’t enough to truly understand the full spectrum of fire loss drivers in the context of public fire protection, and it isn’t enough to subsequently develop a robust predictive methodology around these drivers. The team is looking at everything from the location of the fire station and fire hydrants to fire station staffing levels, building stock characteristics, weather at the time of the incident, and geographic area protected. They’re also studying communities, including population density by zip code, average building value, and relative population statistics.

Past data analysis shows the importance of community risk reduction (CRR) or working with communities to improve their understanding of what’s causing fires and how to prevent them. More information about this exciting program will be available soon.

Today, AAIS is hard at work on data research and development. The goal is to begin implementing the new public fire protection model into select model insurance programs and states by January 2020.

This innovative methodology will apply to both commercial and residential risks – a departure from past industry efforts, where only residential fires were studied. AAIS believes that by studying data across industry lines, members will have an end-to-end, holistic solution for their customers.

iPer the Insurance Information Institute (


View PDF Version