Freedom of Choice.

Individual life insurance applicants are more likely to purchase a policy if they can choose from a variety of underwriting requirements, weighing the burden of completing requirements against the appeal of a lower premium.

Consumers shopping for individual life insurance can choose among a broad range of products, carriers and distribution systems. Once they select an insurer, sales channel and product, however, they are typically offered only one combination of underwriting requirements and premium. If that combination is unacceptable, they often must select another product, insurer or sales channel. This process is inconvenient for buyers and wastes resources used to promote a particular company or present a product.

A buyer may find an offered underwriting/premium combination unacceptable for a variety of reasons. More common reasons include these:

* The premium is perceived to be too high.

* The underwriting is considered inconvenient or too extensive.

* The buyer is averse to medical examinations or disclosure of medical data.

* The buyer prefers a more rapid application and approval process than the one offered.

To avoid these objections, an insurer can offer a range of underwriting/premium combinations and allow the applicant to balance price, intrusion, privacy and convenience issues. The emphasis in this strategy, however, is on designing an attractive process for applicants, not accommodating conventional pricing and underwriting processes. Implementing this multiple underwriting/premium combination structure will require new and creative approaches.

Applicant Considerations

Price: The cost of underwriting tends to be inversely related to the cost of death claims, and products with little or no underwriting generally experience higher death-claim rates than products with extensive underwriting. But the relationship is not proportionate, so the cost of additional underwriting is generally more than offset by the resultant reduction in death claims. Therefore, gross premiums or cost of insurance charges are relatively high for minimally underwritten products and relatively low for extensively underwritten products.

Intrusiveness: The intrusiveness of underwriting also tends to be inversely related to experienced death claims. For example, products issued based on application responses alone tend to experience higher death-claim rates than products issued based on medical underwriting. More extensive underwriting tends to be more intrusive in terms of the time and tasks required of the applicant and the amount and type of information gathered about the applicant. Gross premiums are typically high for products based on minimally intrusive underwriting processes and low for intrusive processes.

Privacy: Individual privacy is an increasingly important issue for life insurance applicants. In designing underwriting/premium combinations, one or more combinations should allow an applicant to choose minimum disclosure of personal or sensitive information. This design allows applicants to weigh disclosure of additional information against potential premium savings and select the most comfortable balance.

Convenience: Like privacy, convenience has a value that varies by individual. One or more underwriting/premium combinations should allow applicants to weigh additional time against potential premium savings.

Underwriting/Premium Structure

The basis of the applicant-choice approach is that some number of discrete underwriting/premium combinations are offered. While it is conceptually possible to design a fully flexible system, such that appropriate premiums are calculated given whatever underwriting information an individual offers, it is not currently feasible to do so for several reasons. First, there is insufficient experience data available to put a mortality value on each piece of information an applicant might offer. Second, while it is computationally possible to go well beyond current pricing processes, it is not yet feasible to price "on the fly." Finally, the application experience could be overcomplicated by such an approach.

Offering a range of discrete underwriting/premium classes provides a meaningful choice to applicants within a structure based on relatively well-understood mortality relationships and striking a reasonable balance between choice and underwriting complexity.

Designing meaningfully different pairings of underwriting requirements and coverage premium is the key to this structure. While the tendency would be to fall back on existing combinations of application questions and tests, a potentially more productive approach is to consider the acceptance or aversion of applicants to various questions and tests and design combinations accordingly. Assumptions for the mortality effects of anti-selection would be applied to each of the combinations.

In addition to the actual data provided by the requirements, underwriters are concerned with the source of the data and the reliability (verifiability) of the data. Basic combinations of requirements could be composed of information obtained without face-to-face contact with the applicant (such as by direct mail or Internet) and not verified by another source (for example, MIB Group Inc.'s fraud-protection service or an attending physician's report). More advanced combinations would include some face-to-face contact (with a sales adviser or paramedical examiner) and outside verifications of several types. Optimal definitions of classes and associated underwriting requirements will vary by distribution channel. Limited benefits and restrictive maximum policy sizes may be necessary for minimally underwritten classes.

The relationship between and presentation of classes are critically important to pricing and the applicant experience. The simplest presentation would be a display of all classes at once, allowing the applicant to choose the most appealing. But this approach has drawbacks. First, premiums for a combination are potential premiums. If during underwriting the applicant is found to be substandard for the selected combination, appropriately rated premiums or cost of insurance rates could be larger than those for less select classes. Second, simultaneous presentation of rates and requirements for several classes, plus smoking and other distinctions, could easily overload an applicant with information.

Perhaps a more successful approach is an interactive presentation starting with the least underwritten class and progressing through the classes until the applicant decides the additional information required to move to the next class is not worth the associated premium reduction. This approach alone does not solve the substandard problem mentioned above, but it makes possible a premium-ratcheting arrangement that does address the problem. The ratcheting works by guaranteeing that any adverse underwriting finding from information made available in an attempt to move up in class will not produce a premium or cost of insurance rate higher than that for the class already attained. Attempting to qualify for a lower premium is, therefore, financially risk-free and encouraged.

Unique Characteristics

Pricing assumptions for the class structure described above have elements common to conventional pricing, plus there are some unique elements. For example:

* Each class, except for the most select, includes individuals that belong to the class per its underwriting requirements, plus individuals who would qualify for a more select combination but chose not to provide additional information.

* Each class, except for the most select, also includes individuals who theoretically belong in a less select class but were guaranteed not to fall out of the class by the ratchet rule.

* The most select class should contain risks and have premiums similar to those in the most select class in a normal underwriting structure.

* Preferred and ultrapreferred rates can be offered to qualifying individuals in the more select classes.

* The cost of the underwriting component of the gross premium will be very low for the basic classes.

Adjusting mortality assumptions for individuals electing not to move up and for substandard applicants protected by the ratchet rule are important actuarial challenges for this class structure. As with normal structures, lapse rates can be expected to vary by class. An interesting option would be to allow insureds to move to more select classes after issue. Such an option would be complicated by issues concerning who pays for the option and how commissions should be adjusted.

The underwriting/premium class structure may be suitable for a variety of distribution channels. Two promising applications are the Internet and work-site sales, where the applicant can be interactively led through the available classes. Insurance sold through banks or other financial institutions is another potential application, customized so that use of client data from the financial institution is a substitute for more intrusive information gathering.

For life insurers attempting to design more applicant-friendly underwriting processes, the applicant-directed approach is a reasonable alternative to the standard underwriting process. Implementation requires unique approaches to pricing and administration. But benefits should include a better return-on-marketing effort and greater applicant satisfaction.

David Cook is a consulting actuary in the Omaha, Neb., office of Milliman & Robertson Inc. Deb Schmidt is a consultant in the Hartford office of Milliman & Robertson Inc.

Examples of Possible Life Insurance Underwriting Classes

Though only an example, this structure illustrates several important design considerations. Four classes may be appropriate for a general-purpose product. A work-site of mass-market product might have fewer classes, while some markets would tolerate more. The number of classes tends to be limited by two factors: the availability of studies assigning a value to nontraditional underwriting requirement models and the complexity of presentation to potential applicants.

Class    Underwriting                             Issue Speed
Base     Simple application questions depending   Immediate accept or
         on the market, such as actively at work, decline
         not diagnosed terminal, etc.
Simple   Complete nonmedical application, MIB     Immediate accept or
         Group fraudprotection services, motor    decline
         vehicles report
Normal   Simple plus paramedical or medical exam  Normal turnaround
         with fluids testing
Advanced Extensive underwriting with few or no    Normal turnaround
         limits
Class    Policy Sizes
Base     Small through core
         expected range
Simple   Small through core
         expected range
Normal   Core expected range
         and larger
Advanced Core expected range
         through maximum