Fighting Fire with Data: How Public Fire Protection Can Save Lives

Jun 13, 2020 / by Phil LeGrone

With more than $22 billion in losses recorded annually, fire insurance is burning a hole in the insurance industry’s proverbial pocket. Having used the same approach to assessing public fire protection (PFP) risk for the past 30 years, insurers have failed to evaluate the impact it has on overall loss cost.

 

With both long-term profitability of insurers and the safety of insureds at risk, AAIS has developed a new data-driven methodology that examines the relationship between claim severity and PFP by quantifying variations of severity.

A New, More Predictive Methodology
Knowing that a massive amount of data was needed to create a compelling model, AAIS leveraged data from The National Fire Incident Reporting System (NFIRS), the U.S. Census Bureau, and Member fire claims, among many other resources.

Data was analyzed and grouped into three predictive areas of focus: fire station information, physical and geographic features, and community characteristics, all of which impact claim severity.

From these groups, the AAIS Data & Actuarial Solutions team identified predictive variables such as station staffing levels, building height, and incident state. These predictive variables are more specific than the standard ‘distance to a fire station’, accounting for all important aspects of predictive areas.

In the end, AAIS’s completed model has shown the chosen variables to be successfully predictive throughout the United States.

AAIS Public Fire Protection Model
AAIS’s Public Fire Protection Model utilizes the identified predictive variables to examine fire risk by zip code. Every variable was assigned a point value representing their relative importance. The sum of the points determines a zip code’s score…the higher the score, the higher the fire risk… the lower the score, the lower the fire risk.

The model codes are color coded, making it easy to visually identify fire risk by location on a map. A shaded state map shows that risk depends on more than the distance to a fire station.

While AAIS currently utilizes its model predominantly among eastern states, it is applicable nationwide. Our Fire Protection Model predicts fire risk unlike any model before. The model is designed to be adjusted as new risks arise, decreasing the need to create new models as new perils are determined, saving insurers time, money, and rework as they stay current with industry trends about fire protection.

Florida Homeowners By-Peril
AAIS had developed Homeowner By-Peril programs for most U.S. states, providing policy language that addresses standard U.S. home perils such as tornados and theft. In 2019, AAIS received approval for its Florida Homeowners By-Peril Program (FL-HOBP), our first HOBP program tailored to perils and market conditions unique to one state.

Why Florida? First, it’s one of the most litigious states in the country. Insurers experience constant regulation changes at the hands of the state government. Florida homeowners also experience perils unique to their environment, like sinkholes, in addition to standard perils like hurricanes, tornadoes and floods. Because of a complex and varied peril map, carriers lacked access to approved programs. Simultaneously, they were dealing with the effects of the constantly changing market conditions dictated by State-owned Citizen’s Insurance, the Florida Insurance Guarantee Association, and reinsurers.

AAIS’s FL-HOBP provides carriers with many tools to meet the unique market conditions in Florida. Tools include, but are not limited to, forms, algorithms, coverages, and endorsements unique to Florida.

FL-HOBP, Meet AAIS’s PFP Model
AAIS’s Fire Protection Model will bring vast improvements in fire-risk prediction. Our PFP methodology is utilized for testing and research purposes, and uses a generalized linear model (GLM), which is standard in insurance rating plans. AAIS designed the Public Fire Protection Model to work in conjunction with our Homeowners by Peril Rating Plan. Once the model is approved, AAIS intends to introduce it via the FL-HOBP Program.

Sidebar: UK Fire Services; A Case to Study
Recently, UK Fire Protection expert John Bonney joined AAIS VP Phil LeGrone to discuss his approach to risk reduction and its impact on fire risk in the United Kingdom.

While serving as the Chief of Hampshire Fire and Rescue Service, Bonney, now a leading figure in the UK Fire and Rescue Service, helped to formulate a new approach to risk reduction. He stresses the idea that community risk reduction is as significant as organizational change, and the need to build a model that addresses both principles.

The UK Fire Service collected data by visiting communities and talking with homeowners. While there, they educated homeowners on easy fire prevention tools, like the importance of smoke detectors, safe appliance use, and more. They were also able to identify hi-risk groups, such as older adults, adjusting their attention accordingly.

Using this data, Bonney and his company, Alchemy Management Solutions, developed an acuity model, which begins by asking ‘what we know.’ Next it asks, ‘what do others know.’ Data is then analyzed to identify correlations between lifestyle and incidents, as well as health and fire risks. The next step is to ‘plan and deploy,’ which involves developing a response, identifying preventative measures, and implementing protection. The final step is to evaluate and learn.

For more insight from John Bonney and details about the U.K.s Fire Protection Model, watch his discussion with AAIS’s Phil LeGrone: ‘A Changing Model for Fire Services.’

 

Tags: Community, Technology, Issues & Trends, Data & Technology, Homeowners, Modeling/Predictive Analytics, Fire, Data/Tech, Underwriting, Modeling/Actuarial, National Science Foundation, Homeowner By-Peril, NFIRS

Phil LeGrone

Written by Phil LeGrone

VP of Data & Actuarial Solutions - Phil has more than 25 years of experience in various risk management activities from both the insurer and risk modeling perspective. His career began in the Loss Control Division at Chubb surveying commercial and industrial risks primarily from a fire protection perspective. Phil has worked for primary insurers and specialty insurance carriers. His expertise includes data, data analytics, catastrophe modeling, software product management, and geospatial data and analysis.

COVID-19 Resource site ad

Lists by Topic

see all

Posts by Tag

See all