The demand for life insurance in OECD countries

Journal of Risk and Insurance, Sept, 2007 by Donghui Li, Fariborz Moshirian, Pascal Nguyen, Timothy Wee

Inflation figures (percentage change in the consumer price index of the respective countries) and M2 stock are obtained from the IMF's International Financial Statistics CD-ROM. M2 figures are converted into U.S. dollars. Population figures are from OECD Economic Outlook: Annual and Semi-annual Data, Vol. 2003, Release 01. Education levels (measured by tertiary GER) are obtained from the UNESCO Statistical Yearbook. Social Security data are from the OECD Public Expenditure Vol. 2001, Release 01. Average life expectancy figures are from the World Competitiveness Yearbook, various issues. GDP figures and exchange rates are taken from the OECD Annual National Accounts (volume 1) Main aggregates, Vol. 2002, Release 04. The GDP figures are on a per capita basis and are expressed in U.S. dollars. The exchange rates for countries that switched to the Euro in 1999 have those years adjusted by using the fixed exchange rate between the Euro and the domestic currency as obtained from the European Central Bank. Real interest rates are based on benchmark government bond yields obtained from Datastream, with missing values replaced by average overnight interbank lending rates (also from Datastream).

Table 1 presents the summary statistics for the regression variables. Life insurance consumption in OECD countries appears to represent a substantial portion of per capita income, with an average of 5.9 percent corresponding to average purchases of about US$1,267 over an average income of US$21,500. Average demand exhibits a large dispersion across OECD countries. One quarter of the sample presents an average demand less than US$300, whereas another quarter presents an average demand higher than US$1,340. Average income is slightly less dispersed, the higher quartile being only 3 times the lowest quartile. On the other hand, social security expenditure exhibits a very large dispersion. Sociodemographic characteristics such as life expectancy and dependency ratio are seen to be quite similar across the sample.

OECD countries present significant variations in their levels of financial development. Foreign participation in a country's life insurance market can vary substantially, with one-fourth of the sample displaying foreign market share below 6.4 percent and one-fourth having foreign market share above 30.3 percent. Inflation rates are seen to be generally moderate, with a median around 3 percent, close to the median and average of real interest rates.

EMPIRICAL RESULT

Univariate Analysis

Table 2 presents the average values of the explanatory variables classified by life insurance consumption quartiles. Differences between the upper and lower quartiles are also provided with t-test statistics in the last column. The first observation is that income is highly correlated with life insurance purchases. Average income also increases monotonically across life insurance quartiles. There appears to be a combination of wealth and preference effects as life insurance consumption almost doubles over the two central quartiles, whereas little difference is observed in terms of average income. On the other hand, the difference between the upper and lower quartiles reflects a strong wealth effect. In fact, average demand in the upper quartile is 3 times higher than in the next quartile for average incomes only 50 percent higher. Likewise, life insurance demand falls to an average of only USS71.8, revealing a high income elasticity at low income levels.