Assuming that the exemption provisions were proper on the basis of 1942 values, then the transposition of the exemption provision from a gross-income to a net-investment-income basis should take account of the changed price level situation.

THE DEDUCTIBILITY OF POLICYHOLDER DIVIDENDS

The recently enacted life insurance tax bill provides that dividends to policyholders cannot be used to create an underwriting loss in excess of $250,000. Two reasons were advanced for this provision. The first was that if a life company were to create an underwriting loss by paying dividends to policyholders and this dividend-created underwriting loss then were substracted from taxable investment income, the result would be the distribution to policyholders of some interest earnings that were not taxed. In phase 1 of the life insurance tax bill the great bulk of interest earnings escape taxation, going into reserve to mature policy obligations. There is no comparable escape of taxation of the mutual interest earnings in fire and casualty taxation. The only escape is tax exempt interest.

The second reason was that in general the larger life companies are mutual companies and that these companies in an effort to reduce their tax burden might declare larger policyholder dividends. The larger resources of the mutual life companies might permit of such a practice and thus create for them a market advantage over companies writing on a net basis.

The factors which led to the limitations upon policyholder dividend deductions in the life field are either not present at all in the fire and casualty field or they are relatively unimportant. In the fire and casualty field, the total share of the market going to the mutuals is only about 25 percent. Also, the mutuals are typically much smaller companies than are the typical stock companies.

Mutuals in the fire and casualty field do not dominate that field as do mutual life companies in the life insurance field. Neither is there any evidence that their surplus is such as to permit them to disburse larger amounts in policyholder dividends. Indeed, a major problem of many mutuals is to maintain the proper ratio of surplus to net premiums written as public demand for their policies increases. Participating stock companies, likewise, are probably now paying about as much in policyholder dividends as their surplus and underwriting results justify.

In the stock formula, dividends are deductible from underwriting results. This is quite comparable to the dividend deduction in phase 2 of the life bill. Does this deduction, when losses exceed the net premiums after dividends, result in a return to the policyholder of untaxed interest earnings? Possibly. Under the total income theory, any underwriting loss resulting from an excess of outgo over premium income offsets interest earnings. This is an inevitable result of the total income approach.

However, the magnitude of investment income is a less important element in the total income of most fire and casualty companies than is true in the case of life insurance. All of the taxable investment in

come of both stocks and mutuals is subject to tax. There is no portion of investment income set aside, tax free, to create reserves. Any escapement of investment income tax, as the result of a policyholderdividend-created underwriting loss, is minimal.

The situation under the present mutual tax formula is somewhat different. The law permits a deduction of dividends from the 1-percent gross income tax floor. There is no deduction of dividends for companies paying on the investment income basis. Hence, under the mutual formula (in which a mutual pays at either corporate rates on its taxable investment income and realized taxable capital gains, or 1 percent of gross income, whichever is the greater), a mutual will never pay less than full tax rates on its entire taxable investment income. It may pay more. That would be when 1 percent of gross income exceeds the tax on investments.

Since taxable investment income is fully taxed and since there is no offset from this portion of income because of policyholder dividends, there are no untaxed interest earnings flowing to policyholders in a mutual policyholder dividend.

Any participating company, whether stock or mutual, could without too much difficulty avoid any limitation on dividends by shifting from a participating to a deviating basis. Writing on a deviated basis is equivalent to paying a dividend at the inception of the policy period rather than at the end. The relative short-term nature of fire and casualty insurance contracts makes this far more possible for fire and casualty companies than it would for life insurance companies, whose existing contracts cannot be changed and which do not terminate except in a far-distant future.

The full deduction of policyholder dividends should be allowed for both stocks and mutuals; otherwise, there would be discrimination between participating and deviating companies.

Assume the case of two otherwise identical companies A and B. A operates on a participating basis, writing $1 million of net premiums and returning $100,000 in dividends. B operates on a deviating basis at a net premium of $900,000. Both have the same net retention, $900,000. If the dividends of A were subject to a tax, A would be discriminated against as compared to B.

CONCLUDING COMMENT

In the preparation of this paper the effort has been to point out those significant differences between stock and mutual fire and casualty companies which would require consideration should the Congress decide to change the present basis of taxing those companies. These differences do, it is submitted, justify separate tax treatment for stocks and mutual fire and casualty companies.

The present method of taxing these basically different types of insurance organizations has remained unchanged in any important respect since 1942. During all of that rather long period, there has developed no conclusive evidence that any segment of the fire and casualty business has suffered because of this dual system of taxa

tion. Competition remains keen and the portion of the business going to the different segments has not changed to any appreciable extent. Differences of opinion can exist regarding the conformity of the present method of fire and casualty company taxation to some preconceived notion of a proper theoretical tax base. Regardless of such conformity or lack of it, the fact remains that the mutual tax provisions have worked out well and with as little inequity as could be found in any tax plan.

In view of the heterogeneous nature of the companies affected, there is no reason to suppose that some other plan would be more equitable than the present one.

In approaching the subject of fire and casualty insurance taxation, the author has elected to confine his discussion to that phase of the problem with which, as a result of his experience, he is most familiar and is, therefore, in a position to speak with a reasonable degree of authority.

FEDERAL TAXATION OF INSURANCE COMPANIES

t

S. Alexander Bell

THE INSURANCE BUSINESS

Before proceeding with the discussion of taxation of insurance companies, it is important to define clearly the nature of the insurance business and the several methods developed of rendering insurance protection to the public. Unlike any other business or economic activity, insurance is not concerned with the production or exchange of tangible goods or the rendering of specific individual services. Insurance, other than life, is concerned with assuring protection against loss from the hazards incident in business or daily life which may result from accidents such as fires, explosions, earthquakes, automobile accidents, etc. Life insurance is concerned with providing, to the beneficiaries, compensation for the loss of the earning power of a victim of premature death. In either case, no specific individual service is rendered to the policyholder other than affording an opportunity to spread the risk insured among a large group of similarly situated individuals and thus minimize the impact of a possible large loss. Insurance is in essence a cooperative venture of large numbers of individuals who band togethehr for the purpose of spreading the risks to which they are subject. They each contribute periodically a small amount to the common fund out of which those few who sustain the losses draw the indemnities provided.

The individual contributions to this common fund, generally termed insurance premiums, are determined, in principle, by dividing the total cost of losses, plus administrative expenses, sustained by the whole group in a given period, by the total number of members of the group known as policyholders. This process known as ratemaking is rather complicated and involves a great many statistical and actuarial implications. This results from the fact that the premiums paid today are calculated to provide the funds to pay losses which will occur in the future during the period for which protection is sought. In some types of insurance, payments may be delayed for months and years and their amount may hinge upon events which may not take place until after protracted investigation and often litigation has been concluded.

It is thus clear that the price or premium for insurance protection cannot be fixed with anywhere near the finality and assurance as in any other business concerned with production of tangible goods or specific services, whose costs are known or can be reliably ascertained beforehand. Safety or contingency reserve additions must be provided to assure that the premiums charged today will be adequate to accumulate the funds which will be required to pay tomorrow's losses. These contingent or safety reserves must be accumulated out of the premiums paid by the policyholders if the total fund is to be