Titles Titles & descriptions

It might appear that spending a great deal of time attempting to define insurance is no more than an academic exercise. It is extremely important and relevant, however, in many instances to determine whether an activity is legally insurance or not.

First, many types of regulation, both state and federal, apply specifically to insurance. For example, the federal McCarran-Ferguson Act created an anti-trust exemption for those in the insurance business, to the extent that state law regulated their business. So courts have been requested many times to rule on whether an activity is insurance. States must determine what is and is not insurance in order to carry out the regulation of insurance activities to which they are charged by their own state laws. It has long been established that product warranties are not insurance and, therefore, are not subject to state insurance regulation even though they transfer risk and utilize the law of large numbers (Keeton and Widiss). Certain insured employee retirement and health care benefit plans enjoy an ERISA exemption from state regulation. Again, controversy arises in some instances as to what constitutes insurance for purposes of these exemptions (Patek).

Second, various taxation laws, both state and federal, apply differently to insurers or insurance activities than to others. Insurance premiums, for example, are subject to state premium taxes, but not to sales taxes. Insurance proceeds are sometimes treated differently by inheritance laws. Insurers are treated differently by federal income tax laws. It has been claimed, however, that "legislation to define insurance for tax purposes is both impractical and unnecessary" (Lenrow). This will be discussed later in this article.

Third, there are various other situations in which insurers or insurance activities are treated differently under the law than others. For example, some state laws have waived the common law doctrine of governmental immunity to the extent that a municipality purchases liability insurance. Whether or not certain risk management programs constitute insurance has become an issue (Coffey).

The legal definition of insurance, then, is extremely relevant and practical. This legal definition is often established with the aid of opinions and writings of experts, many of whom are economists or academicians. These experts, along with authors of law textbooks, put forth the economic, mathematical, historical, and academic definitions, which are then typically embodied in the legal definition.

It is the purpose of this study to examine the concepts and definitions of insurance--both legal and economic--in light of the tremendous changes in the ways risk managers and insurers have sought to reduce their risks over the past decade or two. Self-insurance arrangements, risk pools, catastrophe future contracts, offshore captives, HMOs and other managed health care arrangements, and other risk management devices have emerged as major competitors to traditional insurers and reinsurers. The nature of these and other transfer devices is examined in light of the concepts and definitions of insurance. The goal is to provide (1) a synopsis of the issues and (2) a framework for making decisions about the definition of insurance applicable for the various parties interested in the definition.

The goal of this study will be accomplished in the following manner. The authors will examine:

1. The need for defining insurance--to whom and for what reasons is it important?
2. The published observations, opinions, and theses of various academic and legal authorities on the subject.
3. The state statutes defining insurance.
4. The court cases defining insurance, categorized by the issues and by the context in which the issue arises (i.e., why the issue arose).

The first two objectives will be accomplished from a thorough review and analysis of the literature in the subject--both academic and legal. The third and fourth objectives will involve a survey and systematic analysis of the relevant state statutes and court decisions.

In the classic thesis always referred to by those researching this topic, The Economic Theory of Risk and Insurance, Allan Willett defines insurance as "that social device for making accumulations to meet uncertain losses of capital, which is carried out through the transfer of the risks of many individuals to one person or to a group of persons. Wherever there is accumulation for uncertain losses, or wherever there is a transfer of risk, there is one element of insurance; only where these are joined with the combination of risk in a group is the insurance complete" (Willett, 72). He further states that "the essential features of economic insurance as we have defined it are the accumulation of capital to meet uncertain losses and the transfer and combination of risks" (Willett, 79). According to an excellent thesis on insurance theory by Irving Pfeffer, however, Willett's definition would exclude retrospective rating, insurance assessment plans, and self-insurance from being considered insurance (Pfeffer, 32).

Professors Robert Mehr and Robert Osier, in an early edition, followed Professor Albert Mowbray's definition, stressing the law of large numbers in their definition: "Insurance may be defined as a social device for reducing risk by combining a sufficient number of exposure units to make the loss predictable. The predictable loss is then shared proportionately by all those in the combination" (Mehr, 17). Professor C.A. Kulp, of the Wharton School, followed an actuarial approach to the definition, claiming that "Insurance ... exists whether the cost of insurance is distributed before or after the period of protection ... and whether benefits are paid in money or services or property. If averaging is applied to risks or hazards on an agreed basis the result is technically insurance" (Kulp, 267).

Pfeffer's classic work suggests that different groups have wrestled with the need to define "insurance" and that their definitions have depended on their particular viewpoints and their reasons for seeking a definition. Thus, there are definitions derived by the courts, historians, economists, governments, and academicians. He provides a convenient summary of these viewpoints:

The courts, guided by the search for administratively convenient rules of law, have stressed the indemnity or insurable interest aspects. The historians, seeking to date the origins of the institution, emphasize the independence of the insurance transaction from loan agreements. The economists, adapting the insurance concept to their oversimplified analytical apparatus, high-light the uncertainty-reducing feature. Government insurance institutions, because of access to the fiscal power of the state, suggest the risk-bearing capacity of the insurer. The textbook writers, finally, because of their concern with underwriting and actuarial considerations, tend to accent the importance of the averaging principle (Pfeffer, 33-34).

Pfeffer's treatise critically analyzes insurance definitions under the following elements: (1) the form (contract, promise, agreement, social device, etc.), (2) the purpose (loss distribution, risk distribution or sharing, risk transfer, loss indemnification, etc.), (3) the means (contribution to a common fund, combination of risks, distribution of costs, etc.), (4) the insured's interest (insurable interest, independent risk of economic loss, material interest, etc.), (5) the insurer's offer (indemnification, compensation, assumes the risk of economic loss, restores capital, etc.), and (6) the insured hazards (specified perils, fortuitous events, contingent events, loss, independent events, etc.) (Pfeffer, 34-36). The generic definition that emerges from his analysis is as follows:

Insurance is a device for the reduction of the uncertainty of one party, called the insured, through the transfer of particular risks to another party, called the insurer, who offers a restoration, at least in part, of economic losses suffered by the insured (Pfeffer, 53).

Interestingly, in this definition he states, "... contains what I believe to be the minimum necessary and sufficient conditions for the existence of the device," while "... some principle is required for the insurer to be able to bear the risk, (but) this principle need not be an application of the laws of large numbers. In addition, the concepts of insurable interest and indemnity are unnecessary and the event insured against need not be specified as `accidental,' but it must be a cause of the insured's uncertainty" (Pfeffer, 52-53). However, he does say, "for the business or institution of insurance, certain additional elements are required either for safety, equity, or both" (Pfeffer, 53). Nevertheless, most current textbooks on risk and insurance hold that the law of large numbers and the principles of accidental loss, insurable interest, and indemnity (for property insurance) are fundamental prerequisites for insurance (Pritchett, Trieschmann, Vaughan).

It is also instructive to review the literature that relates more to the legal definition of insurance. Keeton and Widiss, in their text on insurance law, state that insurance "is generally understood to be an arrangement for transferring and distributing risks" (Keeton, p. 3). They then note that, "The definition of insurance that is either explicit or implicit for purposes of the statutes regulating entities engaged in an insurance business, may be quite different from the definition of insurance used in an estate tax law concerned with determining whether it is appropriate to tax payments made to a beneficiary.... In a complex commercial society, it is both appropriate and desirable that insurance concepts, including definitions of insurance, remain flexible enough to be adapted to changing and differing circumstances rather than being so rigid that they become shackles to thought, expression, or innovation" (Keeton, pp. 4-5). Finally, they state the dilemma as follows: "All insurance contracts concern risk transference, but not all contracts involving risk transference are insurance" (Keeton, p. 12). Interestingly, we get our first indication that under the law insurance may be defined differently depending on the purpose for the definition.

Eric Holmes, in his recent legal text on insurance, suggests there are three tests that may be used in defining insurance:

1. Fortuity and substantial control test:

a. an insurable interest

b. a risk of loss

c. an assumption of risk

d. a general scheme to distribute the loss

e. a payment of premium for the assumption of risk

2. Principal object and purpose test:

Were the elements of risk transference and distribution of a fortuitous insured event a central and relatively significant feature of the commercial transaction or simply incidental and ancillary to the other elements that give the transaction its distinctive nature?

3. Regulatory value test:

Each commercial transaction should be examined to determine if that transaction ought to be regulated in the public interest; e.g., auto clubs, vehicle protection plans, collision damage waivers, homeowner warranty plans, and prepaid service plans. Protection of the public from schemes, deceptions, and insolvency of third parties who make insurance-type promises justifies application of state insurance regulatory laws (Holmes, 1996).

A 1963 treatise was one of the inspirations for this study. Entitled "The Legal Definition of Insurance," it examined the state laws at that time that defined the term insurance (Deneberg). He concluded that, "Many of the difficulties could be eased by the correct application of basic insurance principles in this legal context."

While many textbooks on risk and insurance contain a definition of insurance, most are derivatives of those discussed above. One recent paper by a professor of insurance, however, proposes a new definition of insurance: "An insurance transaction occurs when any number of risks from individuals and/or firms is transferred efficiently to an insurance organization operating subject to the forces of the insurance marketplace" (Porat, 1999). The primary change sought by the authors is the addition of the word "efficiently" to the transfer of risk, which they claim is what the IRS requires to make premiums to captives deductible. They do not define "insurance organization," however. This paper reveals once again that the definition seems to be dependent somewhat on the nature of the legal question posed.

In the next section we will examine state statutes regarding the definition of insurance. This literature review was necessary as a prerequisite to the examination of state statutes and court cases because both legislatures and courts appear to draw heavily upon the testimony--direct and via texts and other written works--of experts when seeking a legal definition of insurance.

In this section, we will first review the statutes and regulations of the 50 states by examining the individual state insurance codes to determine which states specifically define the term "insurance" and how they define it. Second, we will examine certain National Association of Insurance Commissioners' (NAIC) model laws where a definition of insurance is necessary and some attorneys' general opinions in which the issue of insurance has arisen.

An increasing number of states have a risk retention law containing a definition of "insurance" to include "primary, excess, reinsurance, surplus lines, and any other arrangement for shifting and distributing risk which is determined to be insurance under the laws of this state." However, that definition begs the question of what is insurance "under the laws of this state."

More helpful is the basic definition, usually found in the beginning of states' insurance codes along with other defined terms. Thirty-four states have a definition of insurance in their codes (16 states and Washington, DC have no definition). In Table 1, we provide an illustration of the various elements included in the different state definitions.

Of the 34 states that define insurance, 31 describe it as an "agreement" or "contract" (column 1). Twenty-eight specify that there is a promise to "indemnify another" or "promise to pay" (column 2), and 24 state that the triggering event is a contingent or unknown event (column 3). (Arkansas and Wyoming use the terms "fortuitous event," which we have interpreted broadly as being equivalent.) While the wording differs slightly among states, a "standard" definition, which contains all three of the above elements, might be stated along the following lines:

Insurance is a contract whereby one undertakes to indemnify another or to pay or provide a specified and determinable amount or benefit upon determinable contingencies.

Nine state laws/regulations specifically require a finding of "insurable interest" in the definition of insurance (although such also could be implied in the references to "damages"). These states are identified in column 4.

A requirement that insured risks be pooled with other similar risks is spelled out (or strongly implied) only in the laws of Connecticut, Georgia, Utah, and Wisconsin. Interestingly, leaving out the requirement of pooling (application of the law of large numbers) would agree with Pfeffer's minimum requirements and disagree with most of the textbook authors. The same is true concerning the inclusion of "insurable interest" in the definition.

New York is notable in that, while the insurance code contains no definition of "insurance," the term "insurance contract" is viewed by some authorities (Keeton, pp. 942-943) as being equivalent. The term incorporates many of the provisions described above in that an insurance contract is deemed to be "any agreement or other transaction whereby one party ... is obligated to confer benefit of pecuniary value upon another party ... dependent upon the happening of a fortuitous event in which the insured or beneficiary has, or is expected to have at the time of such happening, a material interest which will be adversely affected by the happening of such event."

What is interesting is that the same law (section 1101 of the New York Insurance Law) says that a contract of warranty, guaranty, or suretyship is also an insurance contract. But, "... only if made by a warrantor, guarantor, or surety who or which, as such, is doing an insurance business within the meaning of this chapter." In other words, the contract is an insurance contract only if the entity standing behind it is "doing an insurance business." If it is not, that contract would NOT be an insurance contract!

There are two Model Laws established by the NAIC relating to "unfair practices" that have suggested definitions of insurance.

The first is entitled the Unfair Trade Practices Act. This law results from states' efforts to have in place laws, pursuant to the McCarran-Ferguson Act, to preempt federal law. It covers such topics as misrepresentations in describing coverages, false advertising, defamation, boycott or coercion, false financial statements, unfair discrimination, failure to maintain proper procedures and records, and failure to provide claims history. This model law has been adopted in most states. In fact, the only one that hasn't adopted it in some form is the District of Columbia. Five states have a "variation" of the model act--Alabama, Illinois, Oregon, Utah, and Wisconsin.

The second model law is entitled the Unfair Claims Settlement Practices Act, drafted by the NAIC in 1991. Before 1990, this law was part of the aforementioned Unfair Trade Practices Act in most states. However, the NAIC decided that provisions relating to claims settlement practices should be incorporated into a freestanding model law. This new act established standards for the investigation and disposition of claims under insurance policies. While these standards are not intended to create or imply a private cause of action for the violation thereof, nevertheless, for purposes of the applicability of the act, NAIC did define "insurer."

It was defined as follows in that act: Insurer means a "person, reciprocal exchange, interinsurer, Lloyd's insurer, fraternal benefit society, and any other legal entity engaged in the business of insurance, including agents, brokers, adjusters, and third-party administrators. Insurer shall also mean medical service plans, hospital service plans, health maintenance organizations, prepaid limited health service plans, dental, optometric, and other similar health service plans as defined in Section [insert applicable section]. For purposes of this Act, these foregoing entities shall be deemed to be engaged in the business of insurance." Interestingly, this definition does little to describe the characteristics of insurance, but rather takes the easy way out by simply listing entities deemed to be insurers.

In questions relating to auto servicing, home warranties, and even collision damage waivers, the courts and attorneys general in some states have applied differing tests to determine whether a particular arrangement constituted "insurance." For example, the Kansas state attorney general said in an opinion that there were five elements to insurance: (1) insurable interest; (2) risk of loss; (3) assumption of that risk by an insurance company; (4) distribution of loss among a larger group of persons bearing similar risks; and (5) payment of premium (Attorney General Opinion No. 80-8, 1/11/80).

However, in separate opinions, rendered years apart, the attorneys general in Kansas, Massachusetts, and Maryland all concluded that extended auto warranties were not insurance even though most or all of the enumerated elements were present. Maine was slightly more direct, stating that auto service contracts are not insurance unless there is a contract issued by a licensed insurer. In other words, with respect to auto service contracts and warranties, what might otherwise be viewed as "insurance" might not be insurance because of the absence of a licensed insurer.

The NAIC has a model law that defines "service contracts" as follows: "to perform the repair, replacement, or maintenance of property or indemnification for repair, replacement, or maintenance for the operational or structural failure [of such property]" (NAIC Model Acts, Volume IV, page 685). Notable is the exception--"does not include mechanical breakdown insurance which is a policy issued by an authorized insurer that provides for repair, replacement, or maintenance...."

Further complicating the issue is the above-cited Kansas attorney general opinion addressing the definition of "insurance" (which is not a defined term in Kansas statutes or regulations). While quoting from a "standard" definition like that noted earlier, the attorney general said, "However, every contract indemnifying or providing some measure of protection from loss or damage resulting upon specified contingencies may not necessarily be insurance."

Case law is a strong determinant of the legal definition of insurance because it applies state laws and previous court decisions to specific controversies. The following is a list of contexts in which the term "insurance" is defined:

1. Is an entity subject to regulation by a state insurance department?
2. Is an activity exempt from the federal anti-trust law?
3. Can payments to a captive insurer be deducted for income tax purposes?
4. Are the activities of a business subject to a state's unfair claims settlement practices law?
5. Does a state's insurance common law apply to a specific insurance activity?

The following is a discussion and itemization of cases categorized by the context in which the issue of the definition of insurance arose.

Based on statutory authority, a state's department of insurance (DOI) typically regulates those entities engaged in "the business of insurance." Many entities resist the regulations by asserting they are not in the business of insurance. As indicated earlier in this article, the level of clarity in a statute's definition of "insurance" varies greatly. The DOI and, ultimately, the courts are then left to apply the state's statutory definition, if any, to the variety of businesses that resist regulation by the DOI. If the statutory definition does not provide the basis for a decision, the courts will in turn refer to other appellate decisions, treatises, or legal dictionaries for guidance. One treatise--Holmes's Appleman on Insurance, 2d--has suggested that courts should consider:

a. Whether the insurance aspects of a transaction (risk shifting and pooling) are a dominant part of a transaction or only an incidental or ancillary part; or

b. Whether the public would benefit from regulation of the activity.

Griffin Systems v. Ohio Department of Insurance (DOI), 575 N.E.2d 803 (Ohio 1991)

Under Ohio law, a company shall engage in the business of insurance, or enter into a contract substantially amounting to insurance, only if licensed. The law being applied to this case was R.C.3905.42, which requires an entity to be authorized as an insurer if it engages "... in the business of insurance, or enter(s) into any contracts substantially amounting to insurance."

An Ohio business sold a car repair agreement to an individual. The agreement covered mechanical breakdown, but did not cover external causes of loss, such as collision, weather, negligent maintenance, or vandalism. The agreements were not incidental to selling the cars or repairing the cars; the agreements were the principal business of the contract seller.

Should this business be licensed with the Ohio DOI? Ruling: No, because the coverage limits itself to covering defects within the product itself--i.e., a warranty.

The crucial factor is not the status of the party offering the agreement, but rather the type of coverage within the agreement. The fact that the seller of the contract does not sell or produce or repair the subject of the contract is not important. The DOI alleged that since Griffin was an independent third party to the transaction, the "warranty" was in reality a "contract substantially amounting to insurance." However, the court decided that since the warranty only covered repairs necessitated by mechanical breakdown of defective parts, the coverage constituted a warranty, but not insurance. This court cited to another Ohio case, Dully v. Western Auto (16 NE2d 256), in making the distinction that "A warranty promises indemnity against defects in an article sold, while insurance indemnifies against loss or damage resulting from perils outside of and unrelated to defects in the article itself."

Courts in other cases reached a different conclusion. One example follows.

Mein v. United States Car Testing, 184 N.E.2d 489 (Ohio App. 1961)

Under Kansas law, a business engaged in the business of insurance in Kansas must be licensed.

An unlicensed company from Ohio sold a car maintenance contract to Kansas customers. The Kansas DOI charged that the company was illegally engaged in the business of insurance.

Is the car maintenance contract equivalent to a contract of insurance or is it merely a warranty? Ruling: It is equivalent to a contract of insurance because it contains elements of insurance and, therefore, the company is subject to licensure.

The court referred to:

1. Kansas statute 40-1102 declares that licensure applies to a business that insures "against liability, loss, or damage from any other risk, hazard, or contingency which may lawfully be the subject of insurance..."
2. Couch Cyclopedia of Insurance Law (2d Ed.) 42, Section 1.15 says, "Whether a warranty amounts to insurance depends on its terms ... [A]n agreement to indemnify against loss or damage resulting from perils outside of and unrelated to inherent weaknesses in the goods themselves constitutes a contract substantially amounting to insurance within the purview of a statute regulating the right of a foreign corporation to do business in the state."

The McCarran-Ferguson Act provides that federal anti-trust laws do not apply to the business of insurance to the extent that such business is regulated by state law. Because this law fails to define the "business of insurance," the courts are left to perform this task. The U.S. Supreme Court uses three criteria:

a. Does the practice have the effect of transferring or spreading a policyholder's risk?

b. Is the practice an integral part of the policy relationship between the insurance company and the policyholder; not an agreement or transaction between the insurance company and an outside entity?

c. Is the practice limited to entities within the insurance industry, such as shared data for ratemaking?

Group Life & Health v. Royal Drug, 99 S.Ct. 1067, 440 U.S. 205, 59 L.Ed.2d 261 (U.S. 1979)

"Business of insurance" is exempt from federal anti-trust law.

Insurance company contracted with a group of pharmacies to provide prescription drugs to its insureds. Nonincluded pharmacies claimed illegal "boycott."

Was this activity exempted by McCarran-Ferguson because it was the "business of insurance"? No, an arrangement to buy services may be the business of insurers, but it is not necessarily the business of insurance. There are three criteria in determining whether an activity involves the business of insurance:

1. Does the practice have the effect of transferring or spreading a policyholder's risk?
2. Is the practice an integral part of the policy relationship between the insurance company and the policyholder, not an agreement or transaction between the insurance company and an entity outside the insurance?
3. Is the practice limited to entities within the insurance industry, such as shared data for ratemaking?

"Since the law does not define the `business of insurance,' the question for decision is whether the pharmacy agreements fall within the ordinary understanding of that phrase, illumined by any light to be found in the structure of the Act and its legislative history."

Union Labor Life v. New York State Chiropractic Assn., 102 S.Ct. 3002, 458 U.S. 119, 73 L.Ed.2d 647 (U.S. 1982)

"Business of insurance" is exempt from federal anti-trust law.

An insurance company used a group of chiropractors to advise it on whether a chiropractic treatment was necessary and the bill was reasonable. A nonparticipating chiropractor sued.

Was this arrangement a part of the "business of insurance," and thereby entitled to an exemption from the anti-trust prohibition against conspiracies? No. Conduct that is perfectly legal may nevertheless not be entitled to an exemption under the anti-trust laws. Every exemption, whether explicit or implicit, from the anti-trust laws must be construed narrowly.

Based on a review of the three criteria cited in Royal Drug, this arrangement was not the "business of insurance."

The Internal Revenue Code sections 951-953 allow businesses to deduct "insurance premiums," but fail to define "insurance." The IRS resists allowing deductions for payments to captive insurance. The federal courts analyze these cases on the basis of whether there is risk shifting. Without risk shifting, captive insurance is merely self-insurance and is not deductible. A captive can achieve risk shifting by selling a significant percentage of its insurance to unrelated businesses.

Amerco, Inc. v. Commissioner, 96 TC 18, (1991)

Premiums paid to the insurance subsidiary were deductible by the parent company under applicable tax laws because the insurance subsidiary wrote a substantial amount of "unrelated" insurance and was licensed under state insurance laws.

Amerco was a holding company for about 250 subsidiaries (part of the U-Haul rental system) and formed a wholly owned insurance subsidiary licensed as an insurance company in 45 states. That subsidiary wrote insurance for, and received premiums from, unrelated entities, as well as providing insurance for Amerco's subsidiaries. The amount of premium received from unrelated entities varied from 54 percent to 74 percent of the total premium received by the insurance subsidiary.

Were the insurance premiums paid by Amerco to its insurance subsidiary deductible as ordinary business expense?

The court looked at whether there was a true insurance transaction between Amerco and its insurance subsidiary (the IRS alleged that there was no real risk transfer). The three tests applied by the court were: (1) whether there was an "insurance risk," (2) whether there was risk shifting and risk distributing, and (3) in the absence of a statutory definition, did the transaction fit the commonly accepted definition of "insurance"? The court concluded that all three tests were satisfied.

The court focused on whether there was a transfer of risk to a pool with unrelated parties. In applying the three tests above, it found that each insured faced some potential hazard, and concluded that there was an insurance risk. There was risk shifting and risk distributing in both the technical sense and an economic sense due to the substantial unrelated business written. The court addressed the definition of insurance by stating that based on the technical indicia of insurance, insurance existed in its commonly accepted sense.

Humana v. Commissioner, 64 AFTR2d 89-5142

The IRS denied the deductibility of insurance premiums paid to an insurance subsidiary by its parent because of the relationship between the parties (using the "economic family theory").

The parent of an affiliated group of hospitals formed an insurance subsidiary, which charged the parent premiums for insurance, which the parent then deducted as ordinary business expenses. The IRS, relying on previous decisions, denied the deduction because the insurance transaction was between "brother-sister" companies.

Were the premiums paid to an insurance subsidiary by its parent not deductible simply because of the relationship between the two parties?

The court found the IRS's reliance on the "economic family theory" to be unjustified. Treating the parent and its subsidiary as separate legal entities, the court concluded that the effect of an insurance claim on the assets of each insured affiliate did effectively shift the risk of loss from the affiliates to the insurance subsidiary: thus satisfying the test used by earlier courts that required such transactions to involve both risk shifting and risk distribution.

Rationale The court held that the "... test for deductibility of premiums is whether there has been risk shifting and risk distribution; risk shifting exists between subsidiaries of parent corporation and parent corporation's captive insurer, and thus payments by a subsidiary to captive insurer owned by parent are deductible."

Sears, Roebuck & Company v. Commissioner, 972 F2d 858 (7th Cir. 1992)

Insurance subsidiary, not technically a captive because it was not formed initially to provide insurance to its parent, reduced reserves based on the premiums it received from its parent. If the transaction did not qualify as insurance, the taxable income of the insurer would increase.

Allstate, the insurance subsidiary of Sears, received more than 99 percent of its insurance premiums from unrelated insureds. The IRS, following the "economic family theory," concluded that it was not relevant whether Allstate also insured unrelated parties; the sole test was whether Sears had divested itself of the adverse economic consequences of a claim. Since Allstate was included in Sears's consolidated tax return, the adverse economic consequence test was not satisfied.