Jeffrey R. Brown and Austan Goolsbee assess the impact internet use has had on term life insurance prices in the US, in the winning paper for the e-commerce section of the Lumina Awards.

This paper provides empirical evidence on how the rise of internet comparison shopping sites has affected the prices of term life insurance in the US during the 1990s. Using data on individual life insurance policies and internet use, the results indicate that a 10% increase in the share of individuals in a group using the internet reduces average insurance prices for the group by as much as 5%. The results suggest that growth of the internet reduced term life prices by 8% to 15% from 1995 to 1997, and increased consumer surplus by $115m-$215m per year and perhaps more. The results also show that the initial introduction of the internet search sites is initially associated with an increase in price dispersion within demographic groups, but as the share of people using the technology rises further, dispersion falls.

The last five years have witnessed an explosion in the growth of internet marketplaces as alternatives or supplements to traditional retail markets. For many products, including insurance, consumers can now go online and comparison shop between hundreds of vendors with much less effort than in the physical world. The traditional economic view suggests that the internet should reduce search costs for consumers and thereby reduce prices and make markets more competitive.

Despite this presumption of increased competition, empirical work on the internet has not been supportive of the theory. Previous research has generally found large dispersion of prices online, and prices either modestly lower or actually higher compared to their offline counterparts. Little is known, however, about the impact of the internet on offline prices. In this paper, we will present the first empirical evidence on the impact of internet competition on prices and dispersion offline. By combining internet and life insurance industry data sets over time, we are able to document how important the internet - and the reduction in search costs that it creates - can be for market competition.

We examine the US market for term life insurance for several reasons. First, in the mid-1990s, a group of internet price comparison sites began that dramatically lowered the cost of comparing the prices (i.e. premiums) of term life policies across companies. Second, life insurance is one of the most widely held financial products in the US, with premiums amounting to several percent of GDP annually. If the internet reduces prices in this market, the potential welfare implications are enormous. Third, there has been a large price decline in the cost of term life insurance in the 1990s that has taken place concurrent with the rise of the internet. We will examine the ways in which the two are related.

We match individual policy-level data to data on the growth of internet usage and online insurance research by various owner characteristics. We fit regressions for the price of life insurance on policy and individual characteristics, and then include a measure of how likely the individual is to have used the internet or to have researched insurance online. The results indicate that once the online insurance sites began, the faster a group adopted the internet, the faster prices of term life insurance fell for that group. They are robust in that rising internet use did not have any effect on prices during the period before the insurance websites existed, nor did it affect the prices of types of life insurance that were not covered by the websites (i.e. whole life policies). Neither can the results be explained by changes in mortality across groups. Interestingly, the data also show that the internet-induced reduction in search costs actually increased price dispersion upon introduction. As internet use became more widespread, price dispersion fell. These results suggest that the internet has had a substantial effect on the pricing, profitability and competitiveness of the insurance industry.

Life and the net
The US market for life insurance is the largest private individual insurance market in the world. In 1999, nearly 59m life insurance policies were purchased, with a face value of over $2.5 trillion dollars, bringing the total number of policies in force to 366m, with a total face value of $15.5 trillion. Individual life policies accounted for 20% of the policies and 56% of the face value of coverage sold. Over half of this individual coverage was term insurance.

By 1996, there were a number of insurance-oriented websites that provided US consumers with access to online quotes for term life insurance. These sites function by the customer providing answers to medical and demographic questions, and entering the amount of coverage they seek. The sites report numerous companies that would offer such a policy and give a price quote from each. In most cases, these sites are strictly a comparison device, as the individual does not buy the product online directly from them, and the connection to the offline seller remains. With the creation of these sites, the costs of comparing prices became extremely low. Users can obtain dozens of quotes in a matter of seconds that would previously have taken much searching. By 1999, more than 5 million households had researched life insurance online.1

Search costs and pricing
Our approach is to think of the internet as reducing search costs and analyse its impact empirically. Of the numerous models of the effect of search costs on prices, the most relevant for our empirical work is that of Stahl (1989). This model begins with most customers paying a search cost for each price quote they receive, while a small fraction have no search cost. Customers search sequentially and the equilibrium prices involve the stores choosing prices from a distribution. The positive search cost customers have an endogenously determined reservation price and stop searching whenever they find a price below that. The low search cost customers get price quotes from all the firms and buy from the lowest priced one.

While in reality internet comparison sites do not reduce search costs to zero, they do dramatically reduce the costs of price searching. We will, therefore, interpret the share using the internet to research insurance as a proxy for the share of the population that is fully informed, and examine what happens to prices as this fraction changes.

There are three basic results stated in the model that have direct predictions for our empirical work (and summarise key findings of the search literature). First, when some consumers have zero search costs and others do not, we should expect to see price dispersion in equilibrium. Second, as the share of consumers with no search costs increases, average prices should fall. Third, as the fraction of informed consumers rises from zero to one, price dispersion may at first increase, and then decrease. This is because when all consumers face very high search costs, all firms can charge a single (high) price. When all consumers face zero search costs, all firms must charge a single (perfectly competitive, lower) price. Thus, increasing the share with no search costs will increase price dispersion at first as some consumers seek out lower prices, and others do not. Eventually, however, when there are enough informed consumers in the market, price dispersion begins to decline.

Data on prices and usage
To test for the impact of the internet on prices and price dispersion, we combine two data sets. First, we use annual surveys of life insurance purchases conducted by LIMRA International from 1992 through 1997. Second, to create a measure of the probability of internet usage for each individual in each year based on the person's observable characteristics, we use the Technographics 1999 survey by Forrester, a leading market research company on the information economy. This provides us with approximately 11,000 person-year observations. Summary statistics are presented in Table 1.

According to this data, over the last half of the 1990s consumers witnessed a large decline in the price of term life insurance. Without taking account of any controls, the average annual premium paid per $1,000 of renewable one-year term coverage was $3.20 in 1993 and by 1997 had fallen more than 20% to $2.50. Over this same period, the overall share of people with online access rose from 2.6% in 1992, to 38.8% in 1997. Of key importance for our regressions is the considerable variation in both the levels and growth patterns of online usage between groups. Not all groups grew at the same rate over time.

Our regressions will attempt to explain the premium paid for term policies. The dependent variable is the log of the annual premium per $1,000 of face value of insurance.

The main variables of interest are variables that measure the probability that an individual uses the internet, as well as the probability that an individual shops for insurance online.2

Basic results
A. Overview of price trends
The results from this regression are listed in column 1 of table 2. The explanatory power of the regression is high3 and the coefficients on the explanatory variables are fully in line with expectations. Policies for men cost about 20% more than identical policies for females, for smokers 45% more than for non-smokers. When interest rates rise (lowering the inverse interest rate term), this reduces prices. Most importantly, the results show a dramatic decline in prices of term life insurance, especially toward the end of the sample. Relative to real prices in 1992, prices for identical policies were about 1% lower in 1994 but almost 19% lower in 1996 and 27% lower in 1997.

Prices seemed to fall most at the time that the internet insurance comparison sites came online. Whole life prices make an interesting comparison since the insurance sites did not cover such policies, and column 2 reports a specification using whole instead of term policies. At the start of the sample the whole and the term price changes were very similar. In 1996 and 1997, however, prices dropped dramatically for term policies while whole life policies remained constant or even rose slightly.

B. Results
In table 3, we add the probability of internet usage to the price regressions. We compute the internet usage for individuals in each year based on age-state, age-occupation, age-income and occupation-state groups, as listed at the top of the column. In every case, the coefficients are negative and significant, suggesting that prices for identical term life policies for people in a given group fell more during those periods in which the group had faster adoption of the internet.

Note that because there are age, occupation, state and year indicators in the regression, these results cannot be explained by level differences in price or life expectancy across groups or time periods. People aged between 25 and 30 may have lower prices than people aged 45-50 because of health differences, lifestyle choices and many other reasons that may be correlated with internet usage. But this will not appear as a positive coefficient on internet usage in our regression because these effects are absorbed in the age variables.

The coefficients indicate that increasing the share of a demographic group that uses the internet by 10% lowers prices for that group by about 1.5% to 4.5% depending on the specification.4 The internet usage variable seems to explain a large fraction of the total decline in prices over this period. In the baseline results without internet use (table 2), prices fell 27% over this period. Once we control for internet usage, the total decline is only 6% and not significant in the age-state regression, meaning that the growth in internet usage can explain about three-quarters of the total decline in term life prices.

We next turn to the Forrester question about whether the individual with online access has ever researched insurance online. We compute the share of each group that has done so (as of 1998) and multiply it by the share with online access in each year. This gives us a measure of the share of the group that has researched insurance online.

The results from using this insurance measure by age-state-year as the explanatory variable are presented in column 1 of table 4. The coefficient is negative and significant. Raising the share of the group using the internet to research insurance online by 1% lowers prices by about 2.5%. We will use this insurance research variable in the remaining results.

Given the observed impact of the internet on term life prices, we can make a rough calculation of the gain in consumer surplus from the price declines generated by growth of the online comparison sites. We do this by multiplying the change in the price generated from the increase in internet usage over the period (8% to 15% in our specifications) by the total amount of term life that was sold in 1995 (the year prior to the introduction of these sites). In 1995 the total annualised premiums for new term policies came to $1.44bn. This generates an annual increase in consumer surplus of about $115m to $215m, quite large for a service used by only a small number of people. If renewals were included, this figure would rise by an additional $560m to $1bn, although this is the upper bound since we would expect renewals to be much less price sensitive because of switching costs. Thus, insurance consumers are clearly the beneficiaries of the internet-enabled search technology.5

Alternative explanations
A. Mortality changes
The most straightforward alternative explanation is that changes in internet use are spuriously correlated with changes in the mortality rates. Mortality has declined over most of the 20th century and, unsurprisingly, so has the price of term life insurance. However, mortality improvement from 1992-1997 was gradual and will have a hard time explaining the sharp price declines witnessed at the end of the sample, and the larger drop for groups with a high propensity to use the internet.

As a test of the importance of mortality changes, in column 2 of table 4 we compute the log mortality rate for each age-state-year using data from the Department of the Census and the National Center for Health Statistics. The coefficient on log mortality is positive and significant on prices as expected but the coefficient on the internet term is not significantly different from the previous regression.6

Another piece of evidence against the spurious life expectancy correlation alternative is seen in whole life prices. Changes to life expectancy should influence both term and whole life policies. Since the comparison sites did not cover whole life policies, however, we do not predict any reduction in search costs in that arena and the internet should have no effect on prices. The results are presented in column 3. The coefficient on the internet variable is positive and insignificant, suggesting that rising shares of the group using the internet to research insurance is not associated with any reduction in whole life prices.

B. Spurious correlation of the growth of internet usage with other factors
Fundamentally, any alternative explanation of the results we have found must be based on the idea that the growth in internet use for a group is correlated with some other unobserved factor that is reducing prices for that group.

One way to check this hypothesis is by estimating the effect of internet usage on insurance prices during the period when there were no online insurance sites (i.e. 1992 to 1995). During this early period, there is no reason for rising internet usage to be correlated with lower insurance prices unless it is spuriously correlated with some other factor. In column 4 we add a variable that is equal to the share of the age-state-year with internet access for 1992 to 1995 interacted with the share having researched insurance online and then zero in 1996 and 1997 (in addition to our standard measure that is zero from 1992 to 1995 and then positive in 1996 and 1997). The results show that prices fell significantly with the rising use of the internet during the period when the insurance sites existed and with approximately the same magnitude as before, but that rising internet usage had no significant effect on prices before the sites existed. This helps rule out the alternative hypothesis that these results are simply due to spurious correlation with internet use.

Much of the existing empirical literature about the internet and search theory has examined whether price dispersion falls when search costs are lowered. We have noted, however, that the theory does not have a monotonic prediction for price dispersion, especially when the starting share of fully informed consumers is low, as it is here. Instead, dispersion may at first rise, and then fall as the fraction of consumers with low search costs rises.

Using our regression results, we can examine the amount of price dispersion within observable groups and correlate it to the share of people using the internet to research insurance (our proxy for having no search costs). We take the residuals from the price specification in column 1 of table 2 and compute the standard deviation within age-state group for each year. This is the amount of price dispersion within a group that cannot be explained by the observable characteristics of the people or the policy types. The standard deviation in the residuals for the median age-state-year group is about 26.

In column 1 of table 5, we regress these measures of price dispersion on online insurance use by age-state-year, as well as the square and the cube of the measure to allow for non-linearity. We graph the predicted values as a function of the share for the specification without fixed effects to show the direction of the non-linearity.

The evidence indicates that price dispersion within groups is actually rising with the share of people researching insurance online for low shares and then falling with the share online once that share exceeds about 5%. This is consistent with the theoretical predictions of the literature. When no one has access to full information, giving the information to a small number of people tends to increase the amount of price dispersion.

In this paper we have examined the US market for term life insurance from 1992 to 1997 and documented that the growth of internet price comparison sites appears to have made the market significantly more competitive. Controlling for policy characteristics and a variety of individual and group controls, we find that as the share of people in a group that uses the internet and research insurance online increases, the more their quality adjusted prices fall. The data also show, consistent with the theory, that increasing the probability of using the internet tends to raise price dispersion initially and then reduce it as internet usage continues to grow. The results seem robust: the growth of internet use does not appear to reduce the price of whole life policies (which were not covered by the internet insurance comparison sites), the growth of internet use before 1996 (when insurance comparison sites did not exist) did not reduce prices, and the results are not affected by adding detailed controls for changes in group specific mortality.

Overall growth of internet usage can potentially explain a significant share of the large price declines of the 1990s. The rise of the internet from 1995 to 1997 appears to have reduced term life prices by about 8% to 15%, leading to an increase in consumer surplus of $115m to $215m per year, and perhaps as much as $1bn.

In this sense, our results show that the ability of the internet to reduce search costs can have a significant impact on markets. When it does so, it may lead to large consumer welfare gains, potentially at the expense of supplier profits. The implications for the market value of online and offline companies could not be more important.

1 Calculated using data from Forrester's Technographics 2000 and Clemmer et al (2000).

2 For controls, we include standard mortality proxies, state and occupation fixed effects, and indicators for whether the policy was purchased from an own agent and whether it was a participating policy. We use the monthly CPI as the deflator and the inverse of one plus the Baa bond rate for the interest rate term. We include information on policy lengths and policy size, as well as year indicators, the coefficients on which provide a price index in log terms for the cost of identical term-life insurance over the period.

3 The regression has an R2 of .837.

4 Because of measurement error in the occupation and income variables, we will concentrate our results below on the age-state.

5 There are several caveats to the conclusion that this technology represents a pure transfer from insurance companies to consumers. First, when prices are reduced for participating policies, policy dividends may also be reduced, partially mitigating the gain to consumers. Second, we cannot completely rule out the possibility that consumers switched to policies that are of lower quality along some unobservable dimension, although evidence below suggests this is unlikely to be the case. Third, while the life insurance industry as a whole will suffer when prices decline, some individual companies (e.g., low price competitors that see substantially increased volume due to the internet) may gain. Finally, if information technology simultaneously serves to reduce costs for insurers, this would also mitigate the net profit reductions that arise from the internet.

6 An alternative mortality-based explanation is to argue that the sample of life insurance buyers changed in 1996, with less healthy individuals purchasing less insurance. We tested for this using data from the 1992 and 1998 Surveys of Consumer Finances and found no evidence that the probability of owning term life insurance changed differentially by age, income, education, or occupation groups.

By Jeffrey R Brown and Austan Goolsbee

Jeffrey R Brown is Assistant Professor of Public Policy at Harvard University's John F Kennedy School of Government. Dr Brown received a PhD in economics from the Massachusetts Institute of Technology, an MPP degree from Harvard University and a BA in economics from Miami University in Ohio. He is a Faculty Research Fellow of the National Bureau of Economic Research, a Research Associate of the Center for Retirement Research at Boston College and a Faculty Research Associate of the Center for Business and Government at Harvard University. Dr Brown is also a member of the American Risk and Insurance Association, the Risk Theory Society and the American Economic Association. During the 2001/02 academic year, Dr Brown is on leave from Harvard to serve as Senior Economist at the Council of Economic Advisers and the President's Commission to Strengthen Social Security.

Austan Goolsbee is Associate Professor at the University of Chicago Graduate School of Business. Dr Goolsbee received a PhD in economics from the Massachusetts Institute of Technology, and MA and BA degrees from Yale University. He is a Faculty Research Fellow of the National Bureau of Economic Research, a Research Fellow of the American Bar Association, and a member of the American Economic Association. Dr Goolsbee has published extensively in the areas of internet commerce, executive compensation, and taxation. Dr. Goolsbee was named by i-Street Magazine in 2001 as a `Leading Academic in Technology', and was named by Business Week as a `star' professor among the nation's top business schools.