The Life Insurance Deficit of American Families: a Pilot Study

Citation:

E. Scott Maynes and Loren V. Geistfeld (1974) ,"The Life Insurance Deficit of American Families: a Pilot Study", in NA - Advances in Consumer Research Volume 01, eds. Scott Ward and Peter Wright, Ann Abor, MI : Association for Consumer Research, Pages: 66-82.

Advances in Consumer Research Volume 1, 1974    Pages 66-82

THE LIFE INSURANCE DEFICIT OF AMERICAN FAMILIES: A PILOT STUDY

E. Scott Maynes, University of Minnesota

Loren V. Geistfeld, Purdue University

[This paper has been tested and improved by comments from William E. Kingsley of the Institute of Life Insurance and from Professors Richard A. Easterlin, James N. Morgan, Craig Swan, and C. Arthur Williams. The authors are appreciative. The complete paper will appear in the Summer, 1974 issue of the Journal of Consumer Affairs.]

[The authors are, respectively, Professor of Economics, University of Minnesota and Assistant Professor of Home Management and Family Economics, Purdue University.]

I. THE PROBLEM

Life insurance plays an important role in contemporary American society. Ninety-one percent of husband-wife families in the United States carry some life insurance. (2, p. 14) For those families which comprise this group, the average amount of coverage in 1972 was $25,200 per family which compares to a mean disposable income of $11,200 per family. (2, p. 26) Even more striking is that for those with life insurance, premiums ate up 3.4% of disposable personal income. (2, p. 61) As is apparent from these figures, life insurance performs a major function in the financial planning of the modern American family.

Given the importance of life insurance, we were puzzled by a question which has not been addressed to date in the literature: To what extent are families under- or overinsured in terms of the income flows they intend to provide for those surviving the death of the chief income earner in the family? This question is of importance for two reasons. First, insufficient insurance implies, at best, a less than adequate level of after-death income and, at worst, economic hardship for survivors. It is this aspect of the question to which this pilot study is addressed, but there is another, perhaps equally important, aspect of the question which is excess insurance. Overinsurance implies the squandering of scarce income on a product people really do not want. In an era of increasing scarcity which is manifesting itself in higher prices, this dimension of the question is assuming greater importance daily.

II. THE PILOT STUDY

Our approach to the problem of over- or underinsurance follows that developed in the Consumers Union Report on Life Insurance. They define the life insurance deficit to be the difference between the expressed need for after-death income and the after-death financial resources of a family. If this difference is negative, it indicates that no life insurance deficit exists. Should a family with a "negative life insurance deficit" have life insurance, they would be overinsured.

The expressed need for after-death income (or financial resources) has the following components:

1. Family Pre-Independence Income Fund. The funds required, net of social security benefits, to maintain a desired level of living for the surviving spouse, children and other dependent until the youngest child is 18.

2. Education Fund. The funds required, net of social security benefits, to provide the level of support parents prefer for their children's higher education.

3. Widow's Pre-Retirement Fund. The funds required to provide the desired level of living for the surviving widow between the child-rearing period and the widow's old age or retirement. The child-rearing period ends when the youngest child becomes 18. Widows begin to draw social security retirement benefits at age 62 or 65.

4. Widow's Retirement Fund. The funds required, net of social security or other pension benefits, to provide a desired level of living for the surviving widow after her retirement, or in her old age. These funds would then be used to purchase a life annuity which would guarantee the desired monthly income for life.

With the exception of the "Widow's Retirement Fund ' the calculation of these "Funds" requires the conversion of future income flows to a current lump sum. A discount rate of 3 1/2% was employed in the calculations. (This is a discount rate which Consumers Union in 1967 hoped would allow for a shrinkage of purchasing power due to taxes and inflation. This is also the reason for not discounting the "Widow's Retirement Fund.)

The existing after-death financial resources include the following: [The categories used are, with changes in wording, those employed in the "Life Insurance Planning Worksheet" in the Consumers Union Report on Life Insurance (1, pp. 113-125). Omitted from this list are funeral expenses and expenses arising from a final illness. Funeral expenses, because they are thought to be "small" relative to other after-death needs. And expenses from a final illness, since they are expected to be covered by Medicare and other health insurance. One other post-death liability -- uninsured consumer debt -- is accounted for as a subtraction from the total of after-death financial resources.]

1. Life Insurance Policies. The face value (less any policy loans) of all policies covering the chief income recipient.

2. Liquid Assets. The sum of cash, demand deposits, savings accounts, and United States Savings Bonds held by the family.

3. Equity in Real Estate. The market value less any outstanding debts on all real estate owned.

4. Securities. The market value of stocks, mutual fund shares. and marketable bonds.

5. Widow's Pension Benefits. A death benefit for an employee's beneficiary arising from the employee's pension plan which is paid in a lump sum to the beneficiary.

6. Other Assets. The market value of art works, antiques, etc.

7. Uninsured Non-Mortgage Debt. The principal of non-mortgage debt not covered by credit insurance. As "negative assets", these must be subtracted from the total of after-death resources.

This list includes only private after-death resources. In addition, most families have after-death rights to various benefits under the Social Security System.

III. METHODOLOGY

This study represents a by-product of a classroom assignment to students in "Consumer Economics" (Economics 20) at the University of Minnesota in Spring 1968, Spring 1969, and Fall 1969. Each student was instructed, if possible, to ascertain the life insurance deficit of a family with an insurable interest [An "insurable interest" exists when the death of the potential insure would place a considerable financial burden on those surviving him.] to whom he had easy access -- his own, his parents or other relatives, or close friends. Students who lacked access to a real family were asked to calculate the deficit for a hypothetical family. [Of a starting sample of 264 families, 93 or more than one-third, were hypothetical, leaving a potential sample of 171 real families.] The population (and sample) for which statistical results are reported here consists of 95 real families with an insurable interest. Each family passed three "tests" pertaining to the quality of data: (1) key assumptions regarding such matters as the desired after-death income levels were provided by either the husband, the wife, or by both jointly; (2) the set of numbers reported was internally consistent; (3) the student obtaining the information reported its "realism" as five to nine on a scale with possible values from zero ("least realistic") to nine ("most realistic"). This screening process reduced the sample size from a potential 171 to an actual 95 families. The degree to which this sample differs from all United States families is discussed in the next section.

To ascertain the life insurance deficit of a family, students used the "Life Insurance Planning Worksheet" which appears in the Consumers Union Report on Life Insurance. (1, pp. 113-1253 In outline, the CU Worksheet asks the user to specify desired after-death income goals for various time periods. From these are subtracted rough estimates of social security benefits to which the survivors are entitled. [Families estimated their social security benefits from a table similar to those provided by the Social Security Administration in the booklet, "Your Social Security". Exact benefits to each family could have been obtained from the Social Security Administration in Baltimore. In view of the uncertainty affecting all types of future income, including Social Security benefits, the gain was not thought to justify the effort required.] Where appropriate, two conversion tables were used to convert future income streams to necessary lump sums, and future lump sums to present lump sums.

That the estimates of the life insurance deficit reported were made with care is probably assured by (1) the interest most families have in the adequacy of their life insurance coverage and by (2) the systematic thinking about insurance needs imposed by the Worksheet itself.

The estimate of the life insurance deficit obtained here is cost free, the results of a single iteration. This constitutes an important limitation of this pilot study. Let us explain. Suppose that you ask a husband-wife to state how much monthly income they want the survivors to have in the event of the chief income earner's death. Suppose the answers to this and similar questions imply, after taking account of the current provision of after-death resources, a deficit, of $30,000. Now suppose further that the husband-wife price $30,000 worth of life insurance and find it will cost them an additional $300 of net premiums per year. They might, upon comparing this with their income, say that it costs too much, and then revise downward their estimate of desired after-death level of income. The opposite reaction, of course, is also possible.

What we wish to stress is that the deficit estimation procedures in this study did not allow the size of the deficit to be modified in the light of information on the cost of additional insurance. [It is possible that some sample families were aware of the cost of additional insurance for them and that the reported deficit reflects this fact.]

IV. THE FINDINGS

The Pilot Study Families vs. United States Families

Of no intrinsic interest, the population which formed the basis for this study becomes interesting only when we see how it compares with United States families in general. Tables 1.1 - 1.5 tell the storY.

From Table 1.1 we see that the youngest child in the pilot study families tends to be older than the youngest child in typical United States families; while Table 1.3 shows that families in the pilot study tend to have a larger net worth than do United States families in general. The effect of both of these is to bias downward the estimates of the life insurance deficit.

TABLE 1

COMPARISONS BETWEEN PILOT STUDY SAMPLE AND UNITED STATES FAMILIES

When comparing the typical United States family to the typical pilot study family, the number of children tends to be the same for each. (Table 1.2) So on this account, no difference between the two in the life insurance deficit, would be expected.

Table 1.4 shows that family income tends to be higher for pilot study families than for United States families, $10,300 versus $9,200; while Table 1.5 shows that pilot study families tend to be younger than families for the United States as a whole. The effect of these two factors on the size of the life insurance deficit is uncertain.

When taking these factors into consideration, we expect our estimate of the life insurance deficit to be an under-estimate of the "actual" deficit; likewise, we expect our estimate of the life insurance surplus, to be an over-estimate of the "actual" surplus.

The Life Insurance Deficit

Table 2 reveals the pilot study families, life insurance deficit. For those pilot study families with a deficit, the average deficit was found to be $40,000. When a serious deficit is defined to be one of $15,000 or more, 46% of the pilot study families have a serious deficit.

According to their own computations, 42% of the pilot study families had no life insurance deficit. In most instances this implies they have too much insurance. The average surplus for those families with a surplus was $48,000 with 38% of these pilot study families having a serious surplus.

TABLE 2

THE SIZE OF THE LIFE INSURANCE DEFICIT

Determinants of the Life Insurance Deficit

Tables 3 and 4 show the effect of various factors on the life insurance deficits of our pilot study families. Table 3.1 and Equations 4.1-4.4 support the contention that for those families who paid attention to the life insurance problem in the past, a negative relationship exists between the present life insurance coverage and the size of the deficit. The regression results of Equations 4.1-4.4 indicate that the deficit decreases about $760 for each $1000 increase in life insurance coverage.

Tables 3.2 and 3.3 along with Equations 4.1-4.4 demonstrate the effect of the economic capacity of the families (normal family income and net worth) on the size of the life insurance deficit. In the regression analysis normal family income proved to fail tests of statistical significance in all but one equation. (These equations are not reproduced in Table 4.) A possible reason for this is its high correlation with net worth ant the size, both absolute and relative, of the pre-independence income goal.

TABLE 3

FACTORS ASSOCIATED WITH THE LIFE INSURANCE DEFICIT

3 A. EFFECT OF CURRENT LIFE INSURANCE COVERAGE

3 B. VARIABLES RELATED TO ECONOMIC CAPACITY OF FAMILY

3 C. VARIABLES RELATED TO THE NEED FOR INSURANCE

3 D. VARIABLES MEASURING GENEROSITY OF VARIOUS AFTER-DEATH INCOME GOALS

As expected, there is a negative relationship between net worth and the size of the life insurance deficit. This most likely obtains since life insurance is a death contingent substitute for net worth.

Tables 3.4-3.6 and Equations 4.1-4.4 show the effects of the need for life insurance deficit. These variables include: (1) the number of dependent children, (2) the age of the youngest child, ant (3) the age of the heat of the family.

All of these variables have a negative influence on the size of the life insurance deficit, but it should be noted that this negative effect consistently fails conventional tests of statistical significance. But in spite of this, the only need variable for which the negative relationship is unexpected is the number of dependent children.

The next set of variables is collectively referred to as after-death financial responsibilities. These are "yes or no" type variables - that is, what is the effect of the family's decision to make provision for a particular after-death responsibility? Table 3.8 and Equation 4.2 show that the decision to provide for the children's education appears to affect the frequency of the life insurance deficit and not necessarily the size of the deficit. Regression results not reported in Table 4, indicate there may be some effect on the size of the deficit: therefore. no conclusion can be drawn.

With respect to the effect of the widow's pre-retirement income, there is conflicting evidence. Table 3.9 says the decision to make provision for this is associated with a smaller deficit while the regression results (Equations 4.1, 4.2, 4.4) say the relationship is positive but not statistically significant. Clearly, the effects of providing for the widow's pre-retirement income on the life insurance deficit is not certain.

The effect of generosity variables on the life insurance deficit is reported in Tables 3.8 and 3.9 as well as Equations 4.1 -4.4. The size of the monthly income goal has a positive and statistically significant effect on the size of the life insurance deficit. The regression results indicate that as the monthly income goal increases by $1,000, the size of the life insurance deficit increases by $120. Variable 9 of Equations 4.2 and 4.3 indicates that this effect is independent of the number of dependents in pilot study families.

The amount of the education goal also has a significant and positive impact on the size of the life insurance deficit. The small size of the beta-coefficients reported is in Table 4 is probably due to the fact that this variable is a stock rather than a flow.

Both the frequency tables and the regression analysis indicate that the pre-retirement income goal and the widow's retirement goal have no apparent effect on the size of the life insurance deficit.

V. CONCLUSIONS

The life insurance deficit is large for many pilot study families. Since this is likely to be an underestimate of the life insurance deficit of United States families, the magnitude of a life insurance deficit of $40,000 cannot be over-emphasized. Upon generalizing, this says that if the head of an average family with a life insurance deficit should die, his family would be $40,000 short of being financially self-sufficient according to their own computations.

A natural question to ask at this point is: Why does a serious deficit (or surplus,for that matter) develop? There are several factors which could explain this, The first is inertia. Families have postponed for too long a re-assessment of life insurance needs, and their coverage has been rendered inappropriate by any of several factors: (1) inflation, (2) major changes in dependency, or (3) a major change in financial status such as (a) the level of living they regard as "desirable", (b) a large change in net worth, and (c) aging.

A second factor which could give rise to a deficit is the use of a faulty framework or rule of thumb for calculating life insurance needs.

Third, the purchase of expensive non-term policies when young can also contribute to this problem. Since non-term policies are relatively costly when young, young families may purchase less insurance than they would have purchased if they had been aware of cheaper policies--that is, due to budgetary constraints imposed by purchasing non-term insurance, the family was unable to purchase sufficient insurance to meet its needs.

Since this study has left many unanswered questions and since there is conflicting evidence on many of the relationships between the life insurance deficit and other variables, further work is required to answer the question posted earlier: Do many families have deficits (surpluses) of "serious magnitude"? If they do, what factors explain the size of the deficit (surplus)?

TABLE 4

MULTIPLE REGRESSION ESTIMATES OF THE LIFE INSURANCE DEFICIT

REFERENCES

Consumers Union, The Consumers Union Report on Life Insurance, Consumers Union, Mount Vernon, New York, 1967.

Institute of Life Insurance, 1973 Life Insurance Fact Book, Institute of Life Insurance, New York, N.Y., 1973.

Maynes, E. Scott and Loren V. Geistfeld, "The Life Insurance Deficit of American Families: A Pilot Study," The Journal of Consumer Affairs, (forthcoming, Summer 1974.)

----------------------------------------