Vol. 3 No. 3 | Winter 2002

IN THIS ISSUE

Life Insurance in the Information Age

Can the Market Add and Subtract?

Entrepreneurs: Will They Stay or Will They Go?

Tax Benefits in Acquisitions of Privately Held Corporations

___________

This Issue In Spanish

This Issue In Chinese

Archives

About

Contact Us

 

 


Life Insurance in the Information Age

Internet Comparison Sites Create Major Marketplace Change


Research by Austan Goolsbee

The dramatic decline in the price of term life insurance in the late 1990s was considered an unexplainable phenomenon-until now.


To many in the life insurance industry, the decline in the cost of term life insurance in the late 1990s was puzzling at best. In 1993, the average annual premium paid per $1,000 for a renewable one-year term life insurance policy was $3.20. By 1997, the average had fallen more than 20 percent to $2.50, resulting in huge savings for consumers and lost profits for insurance companies.

Within the same period, several Web sites were launched that provided fast and easy price comparisons for life insurance policies across different companies. Could a relationship between these two phenomena be proven and explained?

Austan D. Goolsbee, professor of economics at the University of Chicago Graduate School of Business, and Jeffrey R. Brown of Harvard University's Kennedy School of Government address this question in their paper, "Does the Internet Make Markets More Competitive? Evidence from the Life Insurance Industry."

In the traditional economic view, the Internet reduces the time and effort necessary to search for the lowest-price product and therefore should lower overall prices, creating a more competitive marketplace. Goolsbee's findings support this view and offer the first empirical evidence of the Internet's impact on offline price competition.

Based on his results, the growth in Internet usage can explain about three quarters of the total declines in term life prices. The rise in Internet use from 1995 to 1997 reduced term life prices by an estimated 8 to 15 percent, implying that consumers saved between $115 million and $215 million per year.

"In the long run, the greater market competition created by the comparison sites will definitely be in the consumers' best interests and will undermine the market power of the leading players in the industry," says Goolsbee.

Choosing a Plan

In 1998, more than 52 million life insurance policies were purchased in the United States, with a face value of nearly $2.2 trillion. There were 358 million total insurance policies, with a face value of $14.47 trillion. As one of the most widely held financial products in the U.S., even a seemingly small percentage decline in policy prices indicates substantial savings for consumers and enormous implications for the industry.

The various forms of life insurance protect one's dependents against major financial losses in the event of the insured's death. For individual life insurance policies, there are two basic types: term and whole. Term life policies provide coverage for specific periods of time, such as one year or five years or until the policyholder reaches a certain age, while whole life policies are not term dependent.

In economic terms, search cost refers to the costs associated with going out and finding information. The time and effort expended by driving around to find the cheapest gallon of gasoline is one example.

Prior to the launch of the comparison sites, the majority of consumers contacted individual insurance agents to gather information on policies and prices from different agencies. The time needed to contact multiple agents and answer health surveys to get quotes meant that relatively few people were informed regarding the best deal on a particular policy. Searching for policy prices online therefore marked a significant reduction in search costs.

In 1996, many insurance-oriented Web sites were launched, offering consumers immediate access to online quotes for insurance products. In nearly all cases, the sites did not sell insurance policies online, but rather served as a means for easy comparison and referral. Consumers simply had to answer a site's medical questionnaire and enter a desired amount of coverage. In a matter of seconds, the site's search engine produced a list of companies that offered matching policies and their price quotes.

These comparison sites focus almost exclusively on term life insurance. By 1999, more than 5 million households had researched life insurance online, a testament to the convenience offered by comparison shopping on the Internet.

Determining the Cause of Price Decline

To observe how certain types of policyholders' insurance prices behaved over time, Goolsbee collected data on the prices of insurance policies and demographic information on policy owners from LIMRA International Buyer's Studies, 1992 to 1997. These studies are annual surveys on the purchase of individual life insurance contracts in the U.S., based on a sample of 30,000 policies issued by 46 companies per year.

Next, he used buyer characteristics to match LIMRA data to information from the Technographics 1999 survey on the growth of Internet usage and online insurance research by Forrester, a leading market research company. His calculations were designed to determine whether those groups whose Internet usage grew the most also experienced a more dramatic decline in insurance prices, and whether those developments occurred at the same time.

Using statistical models to calculate policy prices based on the characteristics of policies and individuals, Goolsbee estimated the likelihood that the individual used the Internet during the given time period, including to check online insurance sites. Characteristics from the Forrester survey, such as age, state, occupation, and income were used to calculate the probability of Internet usage for each individual in each year. This estimate was necessary because the LIMRA study did not specifically ask individuals if they had used online insurance sites.

In addition to providing information on computer ownership, Internet use, and online buying behavior, the Forrester survey was especially useful because it asked individuals whether they used the Internet to research products online, including life insurance.
To see how Internet use affected policy prices for term life insurance, Goolsbee studied changes in the levels and growth patterns of online usage between different groups of individuals and then compared price changes among groups whose Internet usage grew at different rates.

"You can think of the share of people using the Internet to compare prices as an indicator of how many informed people there are," says Goolsbee.

One analysis compared data on individuals from California, Virginia, and Washington, where more than 40 percent of the population was online in 1997, with individuals from Alabama, Louisiana, Kentucky, and Arkansas, where Internet use was about 25 percent in 1997. The results show that prices for identical term life policies in states with high Internet use fell significantly faster at the end of the sample period, with 1997 prices 32 percent below 1992 levels, than they did in states with low Internet use, where 1997 prices were about 13 percent below 1992 levels.

Similar results were obtained by comparing policy prices for people in high-skill occupations with high levels of Internet use with people in low-skill occupations. In addition, more substantial price declines could be seen for people under age 30, whose Internet use reached 46 percent in 1997, than for people over age 45, where only 34 percent of the group was online.

Using Forrester data, Goolsbee calculated the likelihood of Internet usage and related it to the price of insurance policies. The results show that prices for individual term life policies for people in a given group fell more during periods in which the group quickly adopted the Internet. Increasing the share of a demographic group that uses the Internet by 10 percentage points lowers prices for the group by 1.5 to 4.5 percent, depending on the combinations of individual characteristics.

The results from the statistical models strongly suggest a direct connection between Internet usage and the total decline in prices of term life insurance policies over this time period.

To rule out alternative explanations for the dramatic price declines, Goolsbee considered changes in mortality and life expectancy and their impact on insurance prices. The degree to which mortality declined from 1992 to 1997 was minimal, and therefore unlikely to cause such sharp drops in price.

Further, increases in life expectancy should influence both whole and term life prices. Since the comparison sites do not cover whole life policies, rising Internet use should not affect prices of these policies. Goolsbee's data supports this explanation for why whole life prices were unaffected during this period.

Goolsbee also considered the possibility that rising Internet use for a group was associated with an unobserved factor that led to price declines. To test this assumption, he estimated the effect of Internet usage on term life insurance prices from 1992 to 1995, the period before online insurance sites existed. Results from this analysis show that rising Internet usage had no significant effect on prices before online insurance sites existed.

Benefits for Consumers

In addition to determining the extent to which the Internet caused term life prices to fall, Goolsbee's evidence also supports the hypothesis that when no one has access to full information about prices, giving that information to a small number of people will increase the range of prices because only a few people will get the best deal. However, in this study, once the proportion of Internet users in the group exceeds about 5 percent, the range of prices then shrinks, reflecting greater market competition.

Goolsbee's research contradicts earlier assumptions that the Internet would be better for sellers than buyers.

"Although you see less of this nowadays, it's worth thinking through claims of future profits from a given Internet business model, because those benefits may simply be passed to the consumer, rather than to the company."


Austan D. Goolsbee is professor of economics at the University of Chicago Graduate School of Business.

>> Request a copy of the research paper featured in this article

 

   
___________________________________________________

Capital Ideas Home
 |  More Faculty Research  |  GSB Home