If you live within a mile of a church, you’re far less likely to have a car accident than drivers who live more than a mile from a church. But if you live within a mile of a restaurant, you face a significantly greater risk of an accident than most other drivers.
Those are among the key findings of a study released Quality Planning Corp., a predictive analytics company based in San Francisco. QPC helps insurance companies price insurance more accurately and fairly.
QPC examined the relationship between where a vehicle owner lives and the likelihood he will be involved in an auto accident and concluded that the riskiest place to live is within one mile of a restaurant. In fact, if the owner of an automobile lives within one mile of an eating establishment he is 30 percent more likely to crash his car than if he lived more than one mile from the restaurant.
The study examined more than 15 million policyholders and two million claims, mapping the proximity of vehicle-owners’ addresses to various types of businesses, including amusement centers, bars, churches, dentists’ and doctors’ offices, parking lots, banks, car dealers, car washes, child day-care centers, gas stations, medical buildings, movie theaters, optometrists’ offices, schools and shopping centers.
The study found that the riskiest places to live near are restaurants, grocery stores, schools and banks. At the other end of the scale, individuals that live within one mile of an airport, park, forest or racetrack are much less likely to suffer vehicle damage.
When it comes to car crashes, churches are the least risky neighbor of all. People who live within one mile of a church are 10 percent less likely to have an accident resulting in a property damage claim than if they lived one more than one mile from the church.
Commenting on the statistics, Dr. Daniel Finnegan, founder and CEO of QPC, noted: “It’s well known that auto insurers use a policyholder’s ZIP code to calculate the risk he or she represents. New technology enables us to be even more accurate in determining the level of risk associated with a policy by identifying the specific risk factors associated with that policyholder’s home address.
“In our research to develop a new predictive loss model for auto insurers, we have identified more than 500 variables that are highly correlated to auto accidents, many of which are specific to a policyholder’s home address. Among the more interesting variables we found are hail storms, crime rate, topography, traffic patterns, occupation, street width and chiropractors per capita.”
Insurance companies have historically based policyholders’ rates on their ZIP code or where their vehicle is kept. While ZIP codes may be convenient and necessary for speedy mail delivery, they are not a particularly good predictor of property/casualty insurance losses. The ability to assess risk at the street-address level is a major breakthrough in private passenger auto underwriting and will eventually lead to more accurate rating and could reduce premiums for some drivers.
“It’s important to remember that these observations are indicative of the area and we would naturally expect higher accident rates in higher traffic areas,” added Bob U’Ren, QPC vice president of marketing. “Traffic patterns and density are often key considerations when selecting sites for restaurants and grocery stores.
“There are also comparatively fewer homes and apartments, and generally lower vehicle use, close to parks and forests. But who would have thought it is more dangerous to live by an elementary school than a liquor store? Or a bank versus a hotel?”
QPC periodically releases snapshots and analyses of auto insurance data to raise awareness of the factors that determine what consumers pay for auto insurance.
Previous reports have examined the fraudulent use of Social Security numbers when applying for insurance, the abuse of the ‘farm discount,’ older drivers and auto accidents versus violations, the relationship between occupations and auto accidents, teenage drunk driving, and the discrepancy between reported and actual mileage.
Rating integrity, competitive advantage
QPC assists auto insurers in their efforts to minimize rating error. QPC takes an auto insurance company’s book of policyholders and processes it through a battery of more than 150 proprietary tests, cross-reference checking and pattern-matching algorithms to identify errors and discrepancies that might suggest fraud and misrepresentation on the part of consumers.
QPC also provides insurers with additional services such as policyholder phone interviews to discover missing drivers, verify garaging addresses, determine annual mileage and other key rating information. Over time, insurance companies with accurate rating information are better able to compete and are more financially stable.