It is now urged in some quarters that both public and private rate controls should be abandoned, leaving prices to be kept at reasonable levels by the competitive forces at work in the industry-except that a public watch-dog agency might act to reduce the likelihood, and mitigate the effects of, insolvency, and to restrain unfairly discriminatory practices. If this were done, the investment income question would melt into the general pricing system. Investment income, like any other resource, would enter the rational, self-interested calculations of any firm seeking an advantage against its rivals.

Rating bureaus and advisory organizations are seen by many as having great potential for adversely affecting rate competition and for doing so through the concentrated power of a relatively small group. Their existence and their potential has been a stimulant to countervailing regulation intended to depress what would otherwise presumably rise. It would take a most fortuitous cancellation effect for these two regulatory forces to neutralize each other. More likely would be a federal supervention of them both.

The departure of some leading companies from the bureaus to which they once subscribed testifies to a decline in bureau power. The proliferation of forms and policy variations where once uniformity was imposed shows increased freedom to deviate from old bureau positions. As bureaus have become more liberal under pressure, the dangers of experimental rates that flirt with insolvency reassert themselves.

Price variation can always seem to be either greater or less than it is--the reason being that at any given price there may be variations in coverage, or in attendant services, or in the willingness of the seller to accommodate a riskier class of customer. Similarly. apparent variation in price may be unreal in that details of coverage or service may differ sufficiently to justify differences in price within a general category of coverage. One thing is clear, however: a far broader menu of choice-both as to price and product-is available to the insurance consumer today than was open to him twenty years ago. This does not always make the buyer happy. Though liberated from controlled alternatives, he sometimes founders in a bewildering sea of not always very meaningfully differentiated choices. Confusion may be better than constriction, but it can be improved upon.

The degree of innovatory freedom in property and liability insurance marketing is quite pronounced. Changes along such lines as installment-payment plans, 6- and 3-month policies, package policies (combining various coverages), and noncancellable insurance have occurred with increasing frequency in recent years. 22

There is evidence to suggest that the buyers of insurance are not so attentive to the competitive activity of the sellers as they might be if they were to derive the greatest benefit from it. Many buy insurance because they have to or feel obligated to; but they shrink from the complexity of it, hesitate to expose their lack of sophistication, and take apathetic comfort from the assumption, "After all, it's regulated, isn't it?" The choice among symbols of stability, prudence, dependability, imagination, and similar abstractions does not excite them. They tend to give their business to a friend or neighbor, or to

22 Crane, p. 41.

whomever a lawyer or banker may suggest. When lower prices are strongly emphasized, however, they can be attracted and pried loose from old ties, and when prices generally rise they become quite vocal.

The success of direct-writing companies has been accompanied by intense competition among them and has led the bureau companies to strike back with more flexible and less u šiform pricing, policies of their own.

EFFICIENCY AND PRODUCTIVITY

Having dealt with concentration, entry and exit conditions, and the extent of price and product variation, we shall now inquire into the general degree of efficiency and productivity in the insurance field. The forces of competition, where they are effective, are supposed to spur the various firms in an industry to cut costs in order to secure advantages in pricing (or at least keep up with the pacesetters).

It is most important to measure the insurance industry carefully in terms of its costs and expenses, because even when rates of return are low, monopolistic power can express itself in sluggish, expensive, inefficient costs of operation. It has been remarked that "the greatest of all monopoly profits is a quiet life." There are those who believe that this is the form of profit which appeals to some insurance managements. They will have to be shown to be wrong if the case for a competitive industry is to hold up. All of our findings about risk-related rates of return could be counterbalanced by a picture of slow, costly and unimaginative performance.

Consequently, we have examined the behavior of expense ratios of a sample of large, medium, and small companies in the active, subsidiary, and mutual sectors to see what trend exists. 23 It is a readily discernible trend of falling expense ratios in the years from 1935 to 1965 (see Figure 2). Although this may imply previous abuses, it also suggests that the competitive forces described above have been exerting their customary downward pressure on costs and upward pressure on efficiency. It cannot be said with certainty exactly what part of this trend is attributable to competition and what part derives from the economies of scale available to large, expanding firms. We have noted, however, that economies of scale are not as pronounced in insurance as in many other industries; and we have seen the expense ratio moving down in medium- and smaller-sized firms.

Expense ratios of larger firms are quite uniformly less than for smaller ones, and expense ratios of mutuals are lower throughout the size range than for stock enterprises. This latter difference reflects in part the effect of price differentials, but it also reflects the fact that average commission rates paid by mutuals are less than those paid by stock companies. In 1958, 370 mutual companies paid average commissions one-half the size of those paid by 730 stock companies 24 Competition has often taken the form of bidding up commission scales for agency selling, thus triggering direct-selling efforts as a cost-avoidance technique.

23Best's Fire and Casualty Aggregates and Averages, 1956-1966 (Morristown, N.J.: A. M. Best Co.), passim. 24 Hensley, p. 180.

The increasing share of the market cornered by firms with lower selling costs, whatever it implies for the future of the agency system, testifies to the operation of competitive forces in insurance markets. In 30 years that share has doubled. Firms combining lower costs with lower premiums in a classically competitive pattern have enjoyed a spectacular growth. The five leading specialty groups increased their net premiums written by 463% from 1949 to 1959; while the five largest capital stock agency groups increased theirs by 175%; and the industry as a whole grew by 116%.25

Even if insurance production costs are higher than they need to be, there is abundant evidence that they are coming down for more than just the economies-of-scale

reason.

The expense ratio trend gives us a reading as to relative efficiency and the direction of change, but it does not show its absolute level or its relationship to efficiency in other fields. There is some reason to believe that performance in this comparative sense is not too impressive in insurance. One reason may be that salaries in the insurance industry do not give it great leverage on the nation's managerial resources. It is usually quite expensive to have inexpensive management, and there is always the long-term risk that quality will subside to a point where it justifies comparatively low rates of remuneration.

Attempts to measure productivity in the insurance business are fraught with difficulties which are compounded when comparisons are made with other types of activities. Our approach was to measure output in terms of numbers of policies (weighted by 1957 share of total industry) to input in terms of full-time equivalent employees. Using this fixed weight index, we uncovered a picture of productivity in the insurance industry which seemed fairly sluggish from 1937 to 1957 but which by 1964 had picked up notably-the compound annual rate of productivity growth having more than trebled (from 0.5% in 1957 to 1.7% in 1964).

SUMMARY

Taking all of our data together-on concentration, on conditions of entry and exit, on variations in price and product, and on levels of efficiency and productivity--what do they tell us? They tell us that a strong case exists to show that insurance prices in general are neither loaded down with the weight of excess profits nor puffed up with excessive costs. Rather they emerge from market processes in which the standard criteria of effective competition are quite reasonably satisfied.

In summary then, after looking at various measures of concentration, some of the historical results on entries into and exists from the industry, price and product variations and productivity measures, we find little evidence of uncompetitive or restrictive behavior in the property and liability insurance industry. By these measures at least, there is evidence of good form and properly competitive behavior in the industry.

25 Hensley, p. 189.

V. CONSIDERATIONS OF SOLVENCY

THE CONCERN WITH SOLVENCY

We have determined that price is low by return/risk standards in the American niny, but, is it too low by inadequacy or solvency standards? Indeed, one can ask, whatever the prices and profits of the industry are, "Is the product-the insurance contract of acceptable quality? Can the insurer meet future obligations at the price he charges for this contract?"

SOLVENCY AND PRICE REGULATION

Early in this investigation of the level of price and the treatment of investment income in the property and liability industry, it was necessary to consider the reasons for regulation and the ways in which regulation might be focused more efficiently on its real business. Behind these questions was the basic problem: Can a definition of the goal of price regulation be laid down which leads to meaningful and empirically verifiable guideposts to action?

A standard definition of the correct price is that price should not be "excessive or inadequate." Once made sufficiently precise, this definition seems to represent adequately the opposing forces at work. We have found what we consider to be workable definitions of the criteria that price be neither excessive nor inadequate. Whether price is excessive may be determined chiefly by examination of the performance characteristics of the industry, with a most serious scrutiny of the rate of return in the industry. If an investigation finds that the performance characteristics evidence a workably competitive market structure, then we may tentatively conclude that price is not excessive.

The determination of the adequacy of price is a more difficult question, partly because adequacy is neither generally understood nor firmly grounded in the economic and juridical mode of discourse. After careful consideration, it appears that the proper way to determine the adequacy of price is through an examination of the quality of product. especially the solvency of the insurer. If the quality of the insurance product can be significantly raised by a higher price--specifically, if the social cost of insolvency can be alleviated with low cost by raising price-then the price of insurance is inadequate. Similarly. if it is found that the number of insolvencies and their effects cannot be significantly reduced by a moderate rise in price, then we cannot say that price is inadequate.

The findings on the question of adequacy of price consist of three parts. First, we have developed a mathematical model of the insurance industry showing the relation between insolvency and its determinants. This model is necessarily a simplified representation of the real process to the extent that it invokes the assumption that "all other things remain equal"; that is, that nothing else is affected except the variables contained in the model. This oversimplification ignores the other responses of an insurance company to contracting profits, rising losses, and impending insolvencies, such as more

selective underwriting, more stringent settlement of claims, and shifts in investment strategies. Though simplified, this model can still be relied upon to shed light on the primary effects of price on solvency. Second, we have applied this model to predict the rate of insolvency and the effect of changing price and imposing a different treatment of investment income on the rate of insolvency. Finally, we have drawn from all available independent evidence to test the validity of the conclusions drawn from the model. (See Appendix E.)

Before developing the mathematical model of the insurance process, it is necessary to review the findings on the appropriateness of including solvency of companies as one of the two principal considerations in price policy.

SOLVENCY AND PRICE POLICY

In most of American industry the regulatory environment has been to a considerable extent laissez-faire; within a broad range the performance characteristics have been determined by the forces of the marketplace. The property and liability industry, however, has been regulated by state commissioners, such regulation becoming increasingly strict over time. Recently, the goals of regulation have narrowed to ensuring an acceptable level of quality of the insurance product. This dimension of product is chiefly reflected in the solvency of insurance companies.

The premium on solvency is a singular phenomenon in American industry. The failure of individual companies is not generally recognized as a social indicator of significance. Occasional insolvencies are the indispensable stick of the marketplace-just as profits are the carrot-necessary to attain the efficient allocation of scarce resources. Super-normal profits indicate that an industry needs an inflow or entry of resources. Conversely, low profits and a high rate of business failure give notice that there should be an outflow of resources from an industry. In most industries insolvency does not cause serious social costs. Those affected are the creditors of the deceased company, usually individual security owners or financial intermediaries who have accepted the inherent risks of the business and should properly be prepared to take the consequences of their risk-taking activities. The purchaser of goods and services is generally not affected, since the transactions have been completed.

The insurance industry is not, in this regard, a typical industry. Its main product is a promise to pay a contingent claim in the future. The quality of this promise is probably the most important component of the product: Will the company pay the claims of the policyholders? Will they be paid promptly, fairly, and without excessive litigation? The quality of the promise is the essence of the product.

At present many observers feel that a complete laissez-faire environment would lead to a level of insolvency which is socially undesirable; many observers maintain that a significant change in. either the level of price or the degree of competitiveness would have drastic effects on the costs of insolvency. It is our purpose to test the validity of these assertions.