Watch your calendar.

Within a year, Lamont Boyd “fully expects” that his firm--Fair Isaac, Minneapolis, Minn.--will be asked to develop a personal lines pricing or tiering model that utilizes an applicant’s occupation and education level as rating factors.

Fair Isaac, formerly based in San Rafael, Calif., has long been the nation’s premier developer of credit scores and credit-based insurance scores. Boyd is Fair Isaac’s insurance market manager.

According to Boyd, the catalyst for such a request would be the publicity generated when the Consumer Federation of America (CFA) recently protested GEICO’s use of occupation and education level in a New Jersey personal auto manual.

Whatever one thinks of the fairness of using those factors in underwriting or rating, Boyd says other carriers are likely to believe GEICO has demonstrated that education and occupation are correlated to insurance loss.

“I absolutely think that, by the way GEICO uses occupation, it has found it to be predictive,” Boyd says. “The [New Jersey] insurance department apparently saw no reason why it wasn’t.”

If that were not enough, Allstate released a study in May supporting its initiative to use education and occupation as factors in awarding auto insurance discounts.

Beyond credit

The development of predictive models using education and occupation, or any other individual characteristics, will present challenges to homeowners carriers beyond those confronted when credit-based insurance scoring was introduced.

In particular, credit-based scoring was built upon massive, standardized databases of credit information on most Americans. Every carrier had access, indirectly, to the same data.

There is nothing comparable to credit bureaus with national information on individuals’ occupations and education levels. Companies would have to gather such information on their own.

Today, while it is common for homeowners applications to ask for an applicant’s occupation, there is no standardized approach for doing so, says Claudine Modlin, a senior consultant for Watson Wyatt Worldwide, an international actuarial consulting firm.

“Some companies leave it to the agents’ discretion whether to record it, some have consumers choose from a list, some have agents or applicants enter it free form,” she says.

Of the insurers Fair Isaac has dealt with so far, only one could provide collected data on occupation, says Boyd, and that information simply categorized applicants and insureds as “professional” or “non-professional.”

Data on education level is, perhaps, easier to standardize (high school, some college, college graduate, etc.), but has not been commonly captured on homeowners applications.

Resistance

Even if carriers make the necessary investment in gathering and analyzing personal data, there’s no guarantee that regulators will allow them to use it.

The CFA is challenging the use of occupation and education level in insurance underwriting and rating “because of their correlation to income and race,” says Robert Hunter, CFA insurance director.

To the extent that the CFA and like-minded groups are successful, insurers will have to factor regulatory resistance into their cost-benefit analysis for gathering and analyzing personal information.

If consumer groups argue that an individual’s education and occupation shouldn’t affect the underwriting of auto insurance, where drivers are the principal risk, they are even less likely to accept that those factors should be considered in homeowners insurance, where property perils cause the most losses.

Historically, homeowners premium was primarily based on rating factors associated with risk of loss to building property. That risk was principally a function of construction, fire protection, and territorial hazards, with a small share devoted to liability coverage.

Today, however, “the trend in the industry is to gather more information on the inhabitants of houses,” says Modlin. “A lot of that [information gathering] has to do with the degree of care likely to be shown by an insured.”

“It’s not unreasonable to speculate that there could be a correlation between education [level] and a propensity to maintain a property,” says Chester Szczepanski, vice president and actuary for the Donegal Insurance Group, Marietta, Pa.

Szczepanski, who was chief actuary for the Pennsylvania Insurance Department, believes companies may be given more latitude to use occupation and education in homeowners than in personal auto, because the former is not required by law, as is auto insurance, and fewer people need homeowners than auto insurance.
Other factors

Jeff Kucera, a consulting actuary with the international firm EMB America, believes that, in the near future, utilization of other data captured on homeowners applications will be a priority over utilization of occupation and education level.

“I think there are enough other things companies need to look at first from a rating perspective,” he says. “Some of the factors currently being used, although not extensively, include the age of the named insured, the number of occupants, original home buyer, first-time homebuyer, marital status, and gender.”

All or most of that information is already collected by carriers and readily available for testing.

“There’s no hard and fast formula for determining how valuable a variable may prove to be,” Kucera adds. “Occupation by itself may not be overly predictive. Education may not be very predictive.

“But with three, four, or five variables providing slight improvements, it may become more relevant.”

Proper analysis needs to extend to specific causes of loss, Kucera says. “Some factors will prove to be more predictive for fire or liability losses than for weather-related losses.”

Whether education and occupation figure prominently or not, Kucera believes homeowners carriers will develop predictive models that identify “risk-averse” people who take steps to avoid or minimize loss.

By so doing, they will establish a reliable and justifiable connection between personal characteristics and property losses.