Academic journal article Risk Management and Insurance Review

An Analysis of the Quality of Internet Life Insurance Advice

Academic journal article Risk Management and Insurance Review

An Analysis of the Quality of Internet Life Insurance Advice

Article excerpt

ABSTRACT

This article analyzes the quality of the life insurance recommendations generated by Internet Web sites. The quality of a Web site's recommendations is measured by comparing how closely the recommendations match two benchmarks chosen by the authors. The authors submitted four relatively simple scenarios to 48 Web sites offering advice on how much insurance is appropriate in given circumstances. The article presents several statistical measures of the wide variation in advice offered by these Web sites, and it suggests several explanations for this variation. Statistically, the Web site recommendations differed from the authors' chosen benchmarks, but did not differ by the type of organization maintaining the Web site.

INTRODUCTION

This article analyzes the quality of the life insurance recommendations generated by Internet Web sites. The quality of a Web site's recommendations is measured by comparing how closely the recommendations match two benchmarks chosen by the authors. Individuals wanting to generate a purchase recommendation at a Web site are asked to input personal family and financial information. The site then relies on a mathematical model to transform the applicant's information into a recommendation for the amount of life insurance to purchase. This article focuses on the quality of this output.

This article is important because commentators have historically assumed that the general public was neither motivated to make the life insurance purchase nor capable of understanding such a complex purchase. From these two beliefs came the notion that the life insurance sale required the services of a well-trained, commissioned sales force. The Internet sale of life insurance substitutes Web advice (and motivation) for the advice provided by state-regulated insurance agents. The presumed calamitous effect on society of giving bad life insurance advice has provided a basis for agent licensing and regulation. Therefore, it is reasonable to ask, and important to understand, whether Web site-provided advice is generally good or bad.

Throughout the article, the authors refer to "good," "bad," "accurate," or "inaccurate" advice. These definitions are relative, as a following section of the article describes. For the purpose of this article, good advice regarding the amount of insurance to purchase is defined as advice that produces neither over- nor underinsurance; bad advice is advice that does produce over- or underinsurance. Over- or underinsurance is measured relative to a logical solution to the problem or circumstances presented. In this article, the logical solutions are measured relative to two standards. The first is the output of the Economic Security Planner (ESPlanner) software program, and the second is the authors' own life insurance model solution.

This article follows a significant contribution by Jagadeesh Gokhale, Laurence J. Kotlikoff, and Mark J. Warshawsky (GKW, 1999) titled "Comparing the Economic and Conventional Approaches to Financial Planning." GKW compared the output of two mathematical software programs that were designed to calculate an appropriate amount of life insurance and savings in given circumstances. "This paper illustrates the different saving and insurance recommendations provided by economic financial planning software ..." The insurance recommendations analyzed by GKW were generated by computer software distributed for use on personal computers, not results generated by Internet Web sites, which is the case in the present study. GKW completed their analysis by taking one basic scenario and making it more complex over a series of 20 iterations as well as by submitting four complete case studies for the software packages to analyze. This article follows the latter pattern of submitting complete (albeit fairly simple) case studies for the Web site models to analyze. Finally, this article uses the same computer program that GKW used as one benchmark in this analysis. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.