Because of the long-term nature of a life insurer’s obligations and the charge of fixed premiums, premiums need to be established on a conservative basis. Barring unusually adverse experience, such conservative premiums should normally generate gains from operations which increase the size of the insurer’s surplus. The allocation of this gain in surplus depends upon whether the insurer writes participating or nonparticipating policies. Nonparticipating polices are those which guarantee the amount insured but do not entitle the insured to receive any benefits other than those explicitly set forth in the contract. Traditional participating policies are also written at a fixed premium. However, they commonly require a relatively larger premium to enable the insurer to meet its obligations on each class of policies under any plausible future adverse experience. Due to the more conservative (higher) premiums, the gains to surplus tend to be greater on participating policies. Most participating life insurance is written by mutual insurers, whereas nonparticipating insurance is written by stock insurers.

The gain of a stock (nonparticipating) company may arise either from a specific provision in the gross premium for profit and/or from favorable actual, as compared to assumed, experience with respect to mortality, interest and/or expenses. This surplus is available to be distributed to stockholders, for the use of their capital funds, in the form of stockholder dividends. Or the gains in surplus may be retained for use in the insurer’s operations. Or such gains may be added to and retained in surplus to provide greater security to policyholders against future adverse contingencies.

The increase of surplus for mutual (participating) companies normally is greater since the premiums are intentionally established on a more conservative basis. However, the higher premium cost to the policyholders is adjusted by dividends paid to the policyholders. The source of these dividends is the extra margins built into the conservative premiums as well as actual mortality, interest and/or expense experience being better than that assumed.

In the absence of legislation or some other form of regulatory constraint, management may use its discretion in the amount and frequency of policyholder dividends. In making such a determination, management must balance such considerations as (a) maintaining or improving financial soundness of the insurer by retaining gains from operations in surplus as protection against adverse future contingencies, (b) the need in some companies to provide investors with an acceptable return on their investments, (c) financing company growth by utilizing such funds in current or new business activities and/or (d) distributing such funds or a portion thereof to policyholders not only as a matter of equity but also to establish a liberal dividend scale for competitive purposes in selling new business.

Once management determines to pay dividends to policyholders, it must apportion the surplus to be so distributed among the various groups of policyholders. It is said that the system of distribution should be simple in operation, flexible, understandable by both policyholders and agents, and equitable, with equity being served by a dividend formula that allocates to each policy its share of surplus in the proportion it has contributed to such surplus. Although management possesses considerable discretion in the distribution of surplus, the states have found it appropriate to impose some constraints or requirements including (1) mandated dividend distributions to policyholders, (2) limitations on stockholder charges and (3) prohibitions against unfair discrimination.

The initial source of regulatory concern arose out of the early use of tontine and semitontine life insurance policies which became very popular in the late 1800s. Basically, the tontine was a standard participating policy with the payment of dividends deferred, for example, for 10, 15 or 20 years. Those who died forfeited any interest in the dividends not yet paid. Those who lapsed their policies forfeited both dividends and policy reserves. The accumulated dividends and forfeitures for each class of policies were paid to those policyholders whose policies were still in force. Based upon estimates that two out of three policies would lapse, the predicted "jackpot" for the surviving policyholders was very substantial. The original tontine was modified to a semitontine under which the dividends but not the reserves were forfeited. However, as was the case with the full tontine policies, only those whose policies were still in force at the end of the designated period shared in the accumulated dividends. The success of the tontine and semitontine policies was phenomenal. However, abuses were not long in emerging and were publicly highlighted by the Armstrong investigation.

It became quite evident that even those policyholders who survived would not receive an amount close to their expectations for three reasons. First, there was a decline in interest rates, a factor beyond insurer control. Second, agents confidently predicted dividend results substantially beyond what actually were achieved. Third, in the absence of appropriate controls, the accumulated funds proved to be easy prey for costly, extravagant and sometimes dishonest management practices. Too often, the large accumulated surpluses were neither properly accounted for nor allocated to the policyholders. Instead, they were often squandered in extravagance and corruption.

As a result of the Armstrong investigation, New York enacted a statute requiring the annual apportionment and distribution of dividends to policyholders. Many states followed New York’s lead. A few states required distribution no less than every 5 years. And a few states prohibited tontine and/or semitontine policies by departmental regulation. Currently, most states require that an insurer make an annual apportionment of divisible surplus. Commissioners review insurer dividend formulas from time to time, often in conjunction with the examination of the company. Furthermore, a few states limit the amount of aggregate surplus which an insurer may accumulate to an amount not to exceed (typically) 10 percent of the legal reserve. Thus, as a consequence of loss of public favor due to publicity and regulatory mandate in many states for annual distribution of dividends, deferred dividend plans have been almost wholly superseded in this country by the annual payment of policyholder dividends.

The requirement for annual distribution of policyholder dividends is inapplicable to stock companies issuing nonparticipating policies. However, a potential conflict of interest arises in those stock companies issuing both participating and nonparticipating policies. In several states, as well as Canada, laws have been enacted or regulations promulgated limiting the amount of unused margins from participating policies which the company can allocate to the benefit of the stockholders, commonly referred to as the stockholders’ charge. For example, in New York and Wisconsin, stockholders are limited to 10 percent of the "profits" on participating business or $0.50/$1,000 of insurance in force, whichever is larger. Illinois uses the 10 percent rule with no alternative, New Jersey the fifty cents per thousand rule and Nebraska requires that all of the surplus arising from participating policyholders accrue to their benefit.

Although annual policyholder dividends are mandated and the dividend formulas are subject to periodic review by the insurance departments, insurer managements possess wide latitude in establishing dividend policy. However, such discretion is constrained by the NAIC Model Unfair Trade Practices Act which defines as an unfair trade practice the making of or permitting any unfair discrimination between individuals of the same class and equal expectation of life in the policyholder dividends payable. Commissioner review and enforcement are the same here as with respect to other unfair trade practices.