In attempting to reach a conclusion as to which of these four alternative approaches is the most desirable, it should be observed at the outset that the third alternative (1942 House bill) is the most consistent in its theory. If even one tax is to be collected on all net investment income, there can be no deduction allowed for policyholder dividends paid out of investment net. Both the second and the third of the alternatives listed above confine the deduction for policyholder dividends to sums paid from underwriting income. However, only the third alternative would be fully effective to prevent any investment income dollar from creeping into the deductible category.

It is true, of course, that one tax could be collected on investment net by requiring the recipient of the policyholder dividend to include in his income that part of the dividend received which stemmed from his company's investment net. When considerations of enforcement. are recognized, however, it seems far better that the deduction be disallowed to the company, rather than that the policyholder be expected to include portions of his policyholder dividend in his taxable income. A choice among the possible treatments for policyholder dividends requires some note of revenue yield comparisons. On this score there is little to distinguish the first and second alternatives noted above. The second alternative does impose a maximum limitation on the deduction allowed for policyholder dividends paid; however, in the normal course of events a casulty insurer is unlikely to pay in policyholder dividends an amount large enough to reach that limit. For example, over the 15-year period, 1943-57, the total of the policyholder dividends distributed by all mutual companies was slightly less than the total of the net underwriting income of those companies. Thus, on an industry aggregate basis even the mutual companies would have suffered no disallowances of deduction for policyholder dividends paid if the approach of the second alternative had been law during that 15-year period.27 It is true, of course, that for a particular company in a particular year the amount distributed in policyholder dividends may exceed the company's underwriting net.28 However, with only minor exceptions, most of the mutual casualty insurers (and all of the stock casualty insurers) appear to have made it a general practice to distribute in policyholder dividends an amounts not in excess of their underwriting net. There is, therefore, probably little or no real revenue estimate distinction to be drawn between the first and second alternatives noted above.

As to the third and fourth alternatives listed above, it is clear that either all or a substantial part of the sums distributed as policyholder dividends would not qualify as deductible. At first blush this might seem to indicate that there would be a higher effective tax yield (particularly from the mutual insurers) if either the third or fourth alternatives were adopted. However, I believe that this apparent possible increase is only apparent. The almost certain change in the premium charge practices which would follow adoption of either the third or

27 For the 15-year period referred to in the text above, the mutual companies had net underwriting income of $2,899,934,000, and have distributed in policyholder dividends a total of $2,678,307,000.

Of course, the underwriting net of the stock companies exceeded policyholder dividends by stock companies with a much greater margin. For the same 15-year period, the stock companies had aggregate underwriting net income of $1,095,418,000, and distributed in policyholder dividends $447,869,000.

28 See, for example, comments about Liberty Mutual Insurance Co., supra, at note 24.

fourth alternative proposals would probably forestall any noticeable increase in taxes actually paid.

On balance, then, I doubt that either the third or fourth alternative proposals (the 1942 House bill or H.R. 198) would yield in revenues any more than would be yielded by the first or the second alternative approach. In addition, I doubt that there would be any appreciable difference in the revenue consequences between the first and second alternatives. Consequently, it appears to me that the Federal revenues would be just as well served by the simple approach of H.R. 7671 and 7672 as by the more complex approaches of the second and third alternatives, or the simple but drastic fourth alternative.

If I am correct in concluding that considerations of revenue yield may be set aside, then it would appear that the approach of H.R. 7671 and 7672 is the most desirable of the differing alternatives. It is true that the 1942 House bill approach was more nearly correct in its theoretical structure. At the same time, if that 1942 proposal would yield no appreciably greater revenues than could be expected under the scheme of H.R. 7671 and 7672, then the much greater ease of compliance offered by H.R. 7671 and 7672 should be determinative. In sum, then, my own conclusion is that the theoretical niceties of the other alternative proposals should yield to the practical advantages of the approach taken in H.R. 7671 and 7672.

PROBLEMS IN THE TAXATION OF FIRE AND CASUALTY INSURANCE COMPANIES

George D. Haskell,1 American Mutual Insurance Alliance, Chicago, Ill.

We are honored to be invited by the Committee on Ways and Means to express our individual views on this panel with respect to the problems in the taxation of fire and casualty companies. In accepting the invitation we subscribe to the statement of your chairman that—

A revenue system such as our own *** must be understood if it is to be supported * ** Many of the existing * * * differentials contained in our tax laws can be justified as serving an essential public purpose *** The individuals and groups which enjoy such *** differentials deserve an opportunity to make the public aware that they are not enjoying unjustified tax benefits.2

As one long associated with the mutual branch of the fire and casualty insurance industry, we do not conceive that this industry enjoys preferences; but we believe that differentials are justified and must be taken into consideration in any equitable system of income taxation. Our task, if we are to be helpful to your responsibility of providing such a system, is to point out the differences in structure, operation, and results which we believe must be taken into consideration.

THE EVOLUTION OF INSURANCE

Insurance is a practical device by which civilized man protects himself against the various contingencies to which his person or his property are subjected in a private ownership type of economy. Basically, it does not eliminate risk but rather transfers the major portion of the risk from the individual to the group organized for sharing risks.

Insurance had its origin in the development of maritime trade. The development of commerce subjected traders to hazards which in time gave rise to a demand for insurance.

As man moves forward in an ever increasingly complex society, he tries, whenever new or important hazards arise to threaten his economic security, to shift as much of the burden as he can to a riskbearing institution. He prefers the certainty of making a small payment to the possibility of incurring a great financial loss.

Although sporadic attempts at the organization of fire insurance companies were undertaken in the American Colonies, the first successful American fire insurance company was organized in 1752. In that year, Benjamin Franklin and his associates in Union Fire Co., a firefighting company in Philadelphia, organized the "Philadelphia

1 Mr. Haskell is economist and director of education for the American Mutual Insurance Alliance, a trade association of advance premium mutual fire and casualty insurance companies. The views expressed are his own, and not necessarily the views of any group or organization.

2 Wilbur D. Mills. Virginia Law Review, October 1958. Preface, pp. 835, 837.

Contributionship for the Insurance of Houses from Loss by Fire." This mutual company is still in existence today and is the oldest business corporation of any type in the United States.

The year 1792-40 years later-saw the formation of the first American stock insurance company-The Insurance Co. of North America. This company expanded fire insurance coverage to cover not only buildings, but also contents.

The number of fire insurance companies increased rapidly in the early 19th century. Most of them were small local companies. But in 1835 a great fire swept New York City. Much of the insurance coverage was held by local companies whose concentration of risks resulted in the failure of 23 of the 26 local companies. As a result, the small companies began to fade from the scene in terms of importance. The period after 1835 is characterized by the growth of large fire insurance companies, writing business throughout the Nation and thus avoiding the danger of overconcentrating risks.

In 1835 some New England textile manufacturers organized the first mutual company to insure factory risks. Existing companies either did not insure such properties because of the extrahazardous nature of these risks-or, when they did, the insurance rates were very high. The new company was successful, and as a result other New England factory owners set up similar companies. Thus the Factory Mutual System was established. By research into the causes of and ways to prevent fire, by developing the science of fire protection engineering, by selectivity of risks through providing insurance only for those risks willing to improve their property, and by economy in op erating costs, fire insurance costs were reduced to a small fraction of the old levels. Of equal importance, this competition caused the whole fire insurance industry to change methods and to study and develop the science of fire protection engineering and underwriting. The result benefited all buyers of factory insurance.

The Farm Mutual System came into being because of the special' needs of farmers. The only available insurance was provided, at very high cost, by general writing stock fire insurers. The rates were inequitable as well as high. There was no effective competition. So, like the property owners in the cities and towns of the East, those who needed protection organized mutual insurance companies to provide protection that fitted their requirements as to coverage and price. These companies have made an enormous contribution to our agricultural system by providing sound protection at greatly reduced cost through economy of operating methods and mutual self-interest in reducing losses. The farm mutuals generally were organized on a local basis, and prorated losses among their members. These companies still retain their local character, and they still operate largely on an assessment basis.

With the great agricultural expansion during the 19th century, flour mills and grain elevators became substantial risks with serious hazards. The only available fire insurance was provided by stock companies. Rates varied from $3 to $10 per hundred, the forms of coverage were not satisfactory, and the insurance companies did not want the business even at their own rates and on their own terms. The millowners and the elevator owners found a solution to their insurance problems by organizing their own mutual companies. For both the manage