Many Households Are Underinsured

03/01/2000
Summary of working paper 7372
Featured in print Digest

Almost one third of wives and 10 percent of husbands would have suffered a decline in living standards of more than 20 percent had their spouse died in 1992. And 15 percent of wives would have suffered a decline in living standards of 40 percent or more.

A significant proportion of U.S. households do not have sufficient life insurance to protect them in case of the death of the primary bread-winner, according to new research by Douglas Bernheim, Lorenzo Forni, Jagadeesh Gokhale, and Laurence Kotlikoff. Secondary earners, primarily wives, and dependent children can face severe hardship if the main earner in the household dies and they are not covered by enough insurance.

Couples where one partner earns much more than the other and those with dependent children - and thus with more family members to protect - are most likely to be underinsured, Bernheim et al show in The Adequacy of Life Insurance: Evidence from the Health and Retirement Survey (NBER Working Paper No. 7372). Older couples in the study's sample, with household heads close to retirement (51-61 years old in 1992) are more likely to have adequate insurance. Also, secondary earners are less likely to be underinsured if their level of education is higher, and when they have pension rights. Underinsurance is also less common among couples that own their home.

To analyze life insurance coverage in U.S. households, Bernheim and his colleagues introduce a number of benchmarks: for example, life insurance is deemed inadequate if it does not allow individuals and their children to sustain their standard of living, in financial terms, upon the death of one spouse. Severe underinsurance refers to a situation where the individual (and children) would face a drop in living standards of 40 percent or more. Significant underinsurance refers to a decline of 20 percent or more. The researchers stress that these are only benchmarks; for example, couples quite rationally may regard life insurance as too expensive.

This study shows that a sizeable minority of the 7,500 couples in the 1992 Health and Retirement Survey sample was significantly underinsured. Almost one third of wives and 10 percent of husbands would have suffered a decline in living standards of more than 20 percent had their spouse died in 1992. And 15 percent of wives would have suffered a decline in living standards of 40 percent or more. Among some groups, the level of underinsurance exceeds two thirds and the extent of severe underinsurance exceeds one quarter.

Household life insurance needs are calculated using ESPlanner, a financial planning software package. This allows the researchers to account for a broad array of economic, demographic and financial factors. The average level of recommended life insurance for husbands, taking the sample as a whole, is put at $88,000. This is around 50 percent more than the average actual level shown in the results, of $60,000. Under-insurance tends to decline with household income at low levels of income, though, and then to level-off at moderate levels of income, the study shows. Among some groups, however, the degree of under-insurance increases with income. One quarter of secondary-earners in households with incomes of at least $100,000 was severely under-insured.

Couples do not increase their life insurance to cover this increased risk, the researchers show. Non-earners in single-earner households are particularly vulnerable. More than one in five non-earners is severely under-insured and another one in seven is significantly under-insured. Nor do couples take full account of the needs of their kids. Families with children are more vulnerable than childless couples or those with adult children but do not buy more insurance to compensate. More than two-thirds of secondary earners in households with dependent children is under-insured, and more than a quarter is severely underinsured. The comparable figures for couples without children are much lower.

Younger households are more vulnerable than older households, but again do not adequately compensate. The degree of under-insurance exceeds 70 percent for 40-something secondary earners, with nearly half the group severely or significantly under-insured. By contrast, the frequency of under-insurance is just over one-third for 60-something secondary earners, with only one in four significantly or severely underinsured.

The researchers also demonstrate a strong relationship between under-insurance and race or ethnicity. The occurrence of under-insurance is more than three times higher for non-white husbands and nearly twice as high for non-white wives as for their white counterparts. Among non-white households, more than one in four secondary-earners are severely under-insured and nearly half are severely or significantly under-insured.

-- Andrew Balls