This entry summarizes the economic understanding of insurance regulation witha focus on solvency regulation, underwriting regulation, contract regulation andcompetition law exemptions. Insurance industry solvency regulation is viewedas a solution to a collective action problem that might otherwise induce insurersto take excessive risk. The entry analyzes the economics of regulationsaddressing adverse selection, including traditional doctrines of warranty andmisrepresentation as well as recent laws relating to genetic testing. Traditionaldoctrines examined from an economic perspective include contra proferentum,"extra-contractual damages," and resolution of conflict of interest problemsarising out of typical liability insurance policies. Competition law exemptions areseen as perhaps an overly broad vehicle to foster the economies of scale inproduction of information needed to prevent adverse selection and insolvency.

JEL classification : K20, K29

Keywords : Insurance Regulation, Solvency Regulation, UnderwritingRegulation, Adverse Selection, Genetic Testing, Warranty, Misrepresentation,Competition Law, Contra Proferentum, Contract Remedies

A. Introduction and Scope Note

In almost all jurisdictions, government heavily regulates the business ofinsurance. To varying degrees in different states and nations, it uses itslegislative and judicial branches to regulate solvency, underwriting practices,and contract structure. It likewise often distinguishes the business of insurancefrom other businesses regarding the extent to which separate enterprises mustbehave competitively rather than cooperatively. While economics-orientedscholarship has not yet spanned the breadth of these varying forms of insuranceregulation or plumbed the intricacies of every doctrine thereunder, an increasingmass of literature has begun to review these regulatory practices. This entryseeks to develop the main contours of the existing body of research.

It should be noted that several critical topics are addressed only briefly inthis entry. The focus is on insurance regulation through legislative action andjudicial rulings as it has developed in the United States. The entry does notdiscuss purely legal analyses of insurance regulation or analyses that draw onother disciplines such as political philosophy. Nor does the entry discuss purelyeconomic analyses of the insurance industry, including the extensive empiricalwork in the field. Rather, the focus is on analyses that have attempted to bringthe tools of economic analysis to bear on the laws regulating insurance.Economic analysis of laws requiring purchase of insurance is addressed in aseparate entry.

B. Solvency Regulation

1. Traditional Economic Understanding

Economic theory has often understood efforts by government to monitor andregulate the solvency of insurers as a solution to a collective action problem notunlike that arising with banks and other financial intermediaries. Winter(199b) Insureds in this view are analogous to fixed debt investors putting moneyinto insurers who reinvest that money and return sums to the insured in theevent certain events come to pass. Hansmann (1985) This beneficialintermediation of investment requires an expectation by the investor/insured thatthe insurer will actually be able to pay those sums when the events occur atsome point in the relatively near future (property/casualty insurance) or in thedistant future (life insurance). Given the incentive of insurer/owners with limitedliability to invest in high risk ventures, however, the likelihood of repayment ishardly assured. This incentive exists because insurer/owners capture all theupside of a favorable materialization of the risk while they are able to shift partof the downside of an unfavorable materialization onto the investor/insureds.While economic theory suggests that investor/insureds may be able to constrainthis "agency cost" or "moral hazard" by use of monitored conditions incontracts Smith & Warner (1979) Mayers & Smith (1981) or through bonding(mutual ownership), Fama (1980) few individual investor/policyholders canstructure the complex arrangements needed to achieve this end. Free ridershipproblems hinders groups of investor/policyholders from coalescing toaccomplish this aim. These problems are particularly weighty with respect toinsurance because non-payment to the investor/insured occurs after the insuredhas suffered an often catastrophic loss at times when prompt payment isparticularly valuable.

It is this understanding of the insurance industry that motivates thepredominant regulatory structure: governments intervene by licensure andpublic insurance mechanisms often known as "guaranty funds" that provide atleast partial indemnity to selected policyholders of an insolvent, licensed insurer. Winter (1991b) The licensure mechanism generally imposes severe punishmentfor sale of insurance without a license and conditions licensure on regulationand monitoring of insurer finances. The public insurance mechanism transfersthe risk that this system will fail in the case of a particular insurer away from theindividual insured to a collective of all insurers and, derivatively, their insureds,although this backup form of risk transfer, if effective, minimizes the incentiveof insured/policyholders to monitor insurer solvency.

Although this mode of regulation is prevalent throughout the world, differentsystems of shared sovereignty differentially allocate responsibility for solvencyregulation. In the United States, a complex system exists under which, for mostinsurers, each of the states may regulate the solvency of insurers "doingbusiness" in that state. Regulations need not be consistent among the states. Lewis & Coates (1991) The danger of excessive regulation posed by this system,particularly with respect to sophisticated insureds, is tempered, however, by twoco-existing alternative forms of insurance regulation: captive regulation andsurplus lines regulation. So-called captive insurers have their access restrictedto relatively sophisticated policyholders capable of banding together to monitorthe insurer. With some exceptions, they are generally regulated only by onestate even when policies are sold outside that state. Surplus lines insurers,instead of being regulated by all states into which the insurer is admitted to dobusiness, likewise have their solvency seriously regulated by only one of thestates. With surplus lines insurers, however, states in which these insurers dobusiness without a license may nonetheless regulate their market conduct.Moreover, most states make their own regulated surplus lines brokersresponsible for appropriate investigations and disclosures with respect to thesolvency of the surplus lines insurers. These alternative modes of insuranceregulation in which a home state has greater responsibility for solvencyregulation of the insurer, resembles, in certain respects, that applying in theEuropean Community. Hogan (1995)

2. Recent Economic Analysis of Solvency Regulation

Economic scholarship regarding the laws governing insurance regulation havedivided into three categories: studies of rules relating premium revenue tosurplus and studies ofprice regulation.

3. Rules Relating Premium Revenue to Surplus

Either through the so-called Kenney Rule or its more modern equivalent via the"IRIS" system of insurance solvency monitoring used in much of the UnitedStates, most jurisdictions attempt to enhance solvency by restricting the ratiobetween the volume of premiums an insurer can write in some given period andits stock of capital surplus. The notion is that restrictions on the rate at whicha financial intermediary such as an insurer can engage in new borrowingdecreases the likelihood of insolvency and preserves the ability of monitoringagencies to provide early warning of problems. Mayerson (1969) Severalscholars have challenged this implementation of solvency regulation by arguingthat it creates or exacerbates positive feedback cycles in insurance pricing andavailability. Winter (1988) Winter (1991b) Gron (1994a) Gron (1994b) Theseinstabilities hamper the economy.

The Winter model assumes that this method of regulation creates a topologyknown as a "fold catastrophe." Winter argues that the regulatory constraintestablishes a peculiar short-run supply curve for insurance which, when coupledwith a relatively inelastic demand curve for insurance, can establish a dynamicrelationship between surplus and insurance price. This dynamic process, whileoften equilibrating insurance markets in a relatively smooth fashion, can alsolead to almost instantaneous increases in insurance price and a correlativedecrease in insurance supplied - a phenomenon last observed in the late 1970sand early 1980s.

4. Price Regulation

Other scholars have considered the economic effects of government priceregulation. This regulation may occur either through the direct setting of a pricefor various forms of insurance or by constraining the prices an insurer maycharge to some range. By requiring that rates be "adequate," insurers areprotected from "trembling hand" instabilities Selten (1975) in the systemwhereby the perhaps irrational decision of one insurer to underprice insurancecan drive out more rational insurers and lead to the collapse of an insurancesystem. Kimball (1992) While history suggests that fear of insuranceunderpricing are not irrational, the ability of pricing regulation to combatinsolvency is questionable. If insurers compete, suppression of pricecompetition merely bolsters non-price competition in areas such as marketing.

Stigler (1968) It lowers consumer welfare because consumers no longer get theproduct they desire most. The loss to the United States economy from thisnon-price competition was estimated in 1980 at $1 billion. Frech & Sambrone(1980) Moreover, to the extent the insurers do not compete, price regulationbased on assuring an adequate return may deter efficient methods of productionby insurers.

Price regulation is also troublesome to the extent that permitted rates do notencompass those which would clear the market. If permitted rates fall below themarket clearing rate, insurance shortages may exist that harm insureds andinsurers alike. Winter (1991a) While the existence of legal "black markets" suchas the surplus lines market and the captive market may reduce the impact of thedistortions and hardships created thereby, the transaction costs associated withpurchase of these sorts of policies, coupled with their limited availability, rendersthese alternatives at best only a partial cure. If permitted rates stay atop themarket clearing rate, a deadweight loss occurs resulting from a reduction inprofitable transfers of risk. Insurers, under such a system, obtain economic rents.Costs incurred by insurer in obtaining or preserving such a hospitableregulatory state may also constitute economic waste.

C. Underwriting Regulation

5. Adverse Selection with Costly Sorting

Adverse selection, the tendency of persons with private and accurate beliefsabout their high risk to transfer risk more extensively than persons with low riskor without private and accurate beliefs, is the central economic problem ofinsurance contracting. Unchecked adverse selection can destroy an insurancemarket and the benefit conferred thereby. Akerloff (1970) The law has thusdeveloped various mechanisms to support "underwriting," the process ofcurbing the "asymmetric information" between insured and insurer that causesadverse selection. Doctrines such as representation and warranty have beencentral in this effort. These mechanisms have long come in occasional tensionwith concerns regarding insurer power and the misallocation of risk that resultswhen these doctrines are interpreted by imperfect courts. These tensions haveled only to adjustments to the central legal doctrines. In recent years, however,it has become increasingly common for laws in support of private adverseselection control to collide with other social goals such as wealth redistributionor antidiscrimination entitlements. Accordingly, the law of underwriting has, inrecent years, undergone significant chance.

As shown by Borenstein (1989) , Rea (1992) and others, when individualswith different expected losses are "pooled" and charged the identical premium,the result is a loss that depends on the elasticity of demand for insurance. Ingeneral, however, the loss results from (1) the consumption of insurance byhigh-risk insureds beyond the point where the marginal value they attach to risktransfer equals the marginal cost of their risk transfer and (2) the consumptionof insurance by low-risk insureds in an amount less than that where the marginalvalue they attach to risk transfer equals the marginal cost of their risk transfer.Although, as noted by Rea (1992) and Chamberlain (1985) , this efficiency lossfrom pooling should perhaps be offset by the gain that pooling - which itselfamounts to insurance of a particular "commodity" price - creates for those who,at some earlier position Harsanyi (1967) behind a "veil of ignorance" Rawls(1971) , do not know their risk class and are unable to adjust measured indicia oftheir expected loss.

Separation of otherwise pooled classes is generally considered to be efficientwhen the cost of the separating program is less than the gains resulting fromenhanced allocative efficiency through pricing that more closely reflects themarginal costs of risk transfer posed by each potential insured. Separation thusbecomes more sensible as the (1) cost of the developing and implementing theseparation algorithm decreases; (2) the homogeneity of expected losses posedby the separated risk classes increases; and (3) the elasticity of demand forinsurance increases. Reduction of the cost of a separation program in turndepends on (a) technological advances in testing, (b) reduction of variable costsby administration of binary tests only to those likely to be in the numericallysmaller of the two separated classes, and (c) legal rules supporting separation.

As also shown by Borenstein (1989) , Schmalensee (1984) and Rea (1992) ,however, low risk individuals and insurers may develop a separation programeven when these efficiency conditions are not met. The low risk insureds, thisargument contends, fail to take into account the interests of the high-riskinsureds in not implementing a testing program that, while it restores allocativeefficiency, ends the cross-subsidization they enjoyed under pooling. Coaseanbargaining by the high risk insureds to pay off the low risk insureds not to betested is even more difficult than usual here, because, almost by definition, thehigh risk insureds do not know the identity of the low risk insureds. Theunregulated market equilibrium in favor of separation is also troubling, assuggested by Abraham (1986) , because it favors separation techniques thatexamine readily ascertainable characteristics such as race or gender that mayhave no causal relationship to expected loss but that bear a statistical correlationwith the difficult-to-observe factors that indeed determine loss.

In theory, the unregulated market equilibrium described above operates byestablishing different pricing terms for identical contracts and achievesseparation through verification mechanisms that reliability and validly classifyindividuals by risk. A second mechanism described first by Rothschild andStiglitz (1976) and explicated by various scholars, Rasmusen (1994) achievesseparation by having insurers offer a menu of contracts with varying shortfallsin the indemnity to be paid - and commensurately lower prices. This menu ofcontracts induces only low-risk insureds (who know themselves to be low risk)to purchase the most incomplete contracts. The mechanism saves the cost ofverification but results in inefficiently low levels of risk transfer to the low-riskinsureds.

6. Classification of Legal Responses To The Problem of Adverse Selection

Legal responses to the tensions between adverse selection control and othergoals, such as non-discrimination, differ in the way that they control the problemof "error" and in the zones of legality they establish according to classificationcost and classificatgion benefit. One prevailing response is "libertarian," to givethe insurer and insured complete freedom to use whatever classification devicesthey choose. Perhaps hoping that the market will punish inaccurate classifiers,the government itself makes no effort to control error. The government likewisefails to evaluate whether the classification method employed is inefficient, eitherclassifying too coarsely or too finely. In some ways, this response, which atleast prevents the possibility of government-created error in classification, hasbeen the prevailing legal response in the United States and elsewhere for manyyears.

Laws that focus primarily on error control represent a second response modeto the tension between adverse selection control and other goals. Laws, forexample, that prohibit classification for automobile insurance based on maritalstatus of the driver or that prohibit different health insurance charges fordisabled individuals absent "sound actuarial principles" for doing so exemplifythis sort of law. Such laws act as a safeguard against the cooperative behaviorgenerally permitted in the insurance industry by prohibiting "irrational"collective discrimination by insurers against disfavored groups. They fail,however, either to require classification in this way or, apparently, to prohibitoverly coarse classification in which separated groups have apparently differentexpected losses but in which that difference is actually caused by an underlyingvariable that greater classification expense would have revealed. Such lawslikewise generally fail to criticize insurance classifications that result inreasonably accurate separation of applicants into groups with equivalentexpected losses but that do so too finely through testing methods that are socostly as to make the whole classification enterprise inefficient.

A third response mode makes classification impermissible, except when it isextremely fine, even where the cost of fine separation exceeds the economicbenefit. Laws generally prohibiting use of gender as a basis for classifyingindividuals for life insurance or annuity policies exemplify this mode ofresponse. Arizona Governing Committee v. Norris (1983) In the western world,gender correlates with life expectancy and it is unclear whether many otherunderlying variables that determine expected life expectancy may be ascertainedat any reasonable cost. Gender is nonetheless sometimes prohibited as aclassification device. Laws of this sort protect both against "irrational" collectivediscrimination but also against the "accident" of certain variables that correlatewith expected losses being cheaper to observe that others that actuallydetermine expected loss. They also may serve as a safeguard against theperpetuation of irrational collective prejudices whose existence is difficult toprove.

A fourth response ("non-actuarialism") is to prohibit classification even if itis efficient and even if it is accurate. Laws barring use of race for classifyingapplicants for life insurance and annuities exemplify such laws. Such lawsredistribute wealth according to political and social preferences apart fromnotions of economic efficiency.

Interestingly, laws that implement the "efficiency concept" described above- do not appear yet to exist, although, as noted below, restrictions on use ofgenetic information may fall into this category of regulation. This absence of"efficient actuarialism" from the law as it actually exists may in part be due to theextraordinarily high legal and administrative costs that would be incurred indetermining accurately whether any particular classification system met theefficiency standard.

7. Traditional Regulation via Legal Doctrines of Misrepresentation andWarranty

As part of a libertarian impulse, the law has long facilitated separation and theeconomic benefits it often entails by permitting the insurer to investigate the riskcharacteristics of a potential insured even where this investigation probesgenerally private information about the insured. Moreover, subject to recentrestraints on use of cheap observables such as race or gender, the law has notintervened in insurer's selection of characteristics used to determine expectedloss, even where this selection - often done in concert - causes substantial lossto identifiable social groups.

Pre-contractual investigations of all potential insureds can, in manyinstances, be quite costly. The doctrines of warranty and representationrepresent potentially cheaper vehicles to create symmetric information andthereby reduce adverse selection. Under both doctrines, the insured is given anopportunity prior to contract formation to make statements that, if true,demonstrate a lower risk and thereby lower premiums in the event a contract isformed. The insured accepts some penalty in the event that some subset ofotherwise insurable events come to pass and the statements made prior tocontract formation turn out to have been false to some pre-specified extent asdetermined by some post-event investigation.

The economic concept behind both warranty and representation is torestructure the "game" being played by the insured and insurer into one wherethe insured's dominant strategy is to reveal truthfully all private informationabout his risk and therefore to render the information symmetrical. The successof this concept becomes more likely as (a) the amount of the potential penaltyincreases, (b) the scope of the subset of events triggering the potential penaltyincreases, (c) the degree of falsity required to trigger the penalty decreases; and(d) the likelihood that falsity will be detected increases. The joint benefit to bothinsurer and insured increases as the expected cost of this post-eventdetermination decreases. Measurements of this expected cost should reflectboth the cost of the investigation itself and the risk cost imposed on the insuredas the result of a determination that is either incorrect or that occurs absent trulyasymmetric belief structures.

The traditional doctrine of warranty is a structure that results in stronginducements to tell the truth but creates substantial costs. Under a polar versionof the traditional doctrine, the penalty (defeasance of an otherwise existinginsurance obligation) is exacted whenever any claim for insurance benefits arisesand any statement made by the insured is shown to have been false to anydegree. Thus, with warranty, a claim for insurance is defeated by the making ofeven minimally false statements even when the complete truth would not havelowered the premium for a policy providing indemnity only for the event thatactually occurred. Because warranty bars claims even when the insured cannotbe shown to have known that the statement was false, the doctrine protects theinsurer (and, in a competitive market, insureds) from statements that the insuredin fact knew to be false in conditions where the insurer cannot prove thatknowledge. The downsides of the warranty mechanism include the frequentinvestigations into the truthfulness of pre-contractual statements and theundesirable transfer of risk to the insured in situations where belief structureswere indeed symmetric but the insured was careless in his pre-contractualstatements.

The doctrine of representation represents another method of controllingadverse selection. Although many variants of the doctrine exist, it may generallybe idealized as a rule under which the insured is penalized (defeasance of anotherwise existing insurance obligation) whenever the insured knowingly madea false statement, a claim for insurance benefits arise, and the premium for apolicy providing indemnity for the event that in fact occurred would have beenmaterially higher had the insured been truthful. Since fewer events trigger aninquiry into the truthfulness of pre-contractual statements and since exaction ofthe penalty depends on the insured's being shown to have knowingly lied, theexpected penalty for false statements is less than that under traditional warranty.The incentive to be truthful and the correlative reduction in adverse selectionmay be tempered as well. On the other hand, representation rather than warrantylowers expected investigation costs and substantially reduces the chance ofimposing a penalty on the insured in situations where beliefs were trulysymmetric.

During the early nineteenth century, many American and English courts sawconsiderable virtue in a legal regime where warranties were unfettered in theirability to control adverse selection. Horwitz (1992) In recent years, however,legal authorities, apparently believing representation either to be a superiorvehicle for controlling adverse selection or in an effort at protecting insuredsfrom the costs of warranty, have generally imposed a preference forrepresentation, at least in situations where the legal status of pre-contractualstatements has been ambiguous. Some legislatures have thus enacted statutesthat deem statements to be representations absent clear statements to thecontrary. Others have enacted "contribute to cause" statutes that permitpenalization of the insured only where the premium for a policy providingindemnity for the event that in fact occurred would have been materially higherhad the insured been truthful. Few jurisdictions have considered a contribute tocause standard coupled with a significant penalty (beyond mere defeasance ofotherwise existing contractual obligations) for discovery of a material falsehood,although such a mechanism with low policing costs and high penalties wouldfit well with economic theory regarding control of illegal behavior. Polinsky &Shavell (1979) .

Also reducing the efficacy of the pre-contractual statement mechanism,though arguably decreasing its undesirable side effects, has been the doctrineof incontestability. Incontestability, which acts as a kind of statute of limitationsagainst insurer defenses of misrepresentation of breach of warranty, generallyaddresses situations where there is a high risk of an erroneous determinationthat a pre-contractual statement was false: the insured is dead and thus unableto testify regarding knowledge of falsity or falsity itself against a life insurerseeking to avoid an indemnity obligation. Thus, at least for the subset of riskfactors that the insurer could reasonably have determined by an investigationeither before the contract or during some period of time (1 or 2 years) afterformation of the contract, the law refuses to let post-event investigations resultin a penalty to the insured. Simpson v. Phoenix Mutual Life Insurance Company(1969) .

Laws reducing the penalty for false statements likewise potentially weakenthe use of post-event investigations as an underwriting tools for controllingadverse selection. These laws reduce, however, the costs associated witherroneous determinations of falsity and determinations of falsity in the face ofsymmetric beliefs. Legislative distrust of the market's determination of optimalmethods of adverse selection control may explain statutes limiting the penaltyfor false declarations of age on a life insurance policy to a reduction in theindemnity equal to the difference between the face amount of the policy and theamount of insurance that could have been purchased for the same premium hadthe insured been truthful about age. Similar judicial distrust of the market mayexplain occasional rulings limiting the penalty for false representations to areduction in indemnity equal to that which the insured would have been entitledto had it told the truth in the application process.

8. Genetic Testing

Legal support for both investigatory and post-event methods of separation hasrecently collided with the growing interest in acquisition of informationregarding the genetic make-up of individuals. The result has been the emergenceof new laws or "voluntary" agreements on the part of insurers restricting use ofgenetic information. The Health Insurance Portability and Accountability Act ofthe United States, laws recently passed in several American states, and lawsenacted in Austria, Belgium, all restrict the ability of insurers either to investigategenetic information or to rely on traditional doctrines such as representationwith respect to genetic information. Indeed, the Belgian law would appear toprohibit even voluntary disclosure of favorable genetic information. Insurers inAustralia, France, the Netherlands, Norway and Switzerland have restricted useof genetic information through industry trade associations.

Most scholars who have examined "genetic information privileging laws"from an economic perspective have been critical of these efforts. They fear it willlead to serious problems of adverse selection in health and life insurance. Christianson (1996) , Pokorski (1995) , Pokorski (1997) . Those with privilegedinformation of a genetic defect that increases the probability of early death butwithout other symptomatic manifestation might, for example, purchase largeamounts of life insurance, knowing it to be a tremendous investment under thosecircumstances. The investment could be liquidated prior to death throughviatical settlements and accelerated death benefit provisions now becomingmore prevalent within insurance policies. Similarly, those with knowledge of agenetic predisposition to debilitating illnesses such as Alzheimer's disease mightstock up on long term disability insurance.

Others, however, have suggested that permitting insurance underwriting todeter genetic testing may lead to delayed diagnosis and treatment of disease.The harm caused thereby outweighs dislocations to the insurance market causedby privileging genetic information. McGoodwin (1996) This economiccomponent of this argument rests, however, on the value to the insured (andperhaps related parties) regarding foreknowledge of genetic defect . Moreover,this argument for privileging genetic information has greater weight with respectto health insurance and possibly long term care insurance than it does for lifeinsurance. With respect to the latter, it is hard to develop a traditional economicargument supporting genetic privileging laws, except perhaps one that sees theprohibition as the solution to a collective action problem under whichindividuals with "good genes" would benefit from testing by costly methodseven when the harm done to those with bad genes, coupled with the cost oftesting, outweighed the total benefit created by the classification scheme.

Still others Stone (1996) Rothstein (1993) Rothstein (1997) Jacobi (1997) haveargued that genetic testing poses a perilous dilemma for a system that reliesheavily on private risk-based insurance for basic health insurance. If genetictests are not privileged, few will take them and those who do may findthemselves denied financial access to good health care. If genetic tests are notprivileged, the private insurance market is seriously endangered. An intriguingescape from the dilemma is suggested by one scholar, however, who proposesto bar individuals from undergoing genetic tests absent prior purchase ofinsurance to indemnify the insured for the higher health or life insurancepremiums that would result from a finding of defect. Tabarrok (1994) Such asystem might itself be subject to some adverse selection, however. Peoplerationally suspecting a genetic defect based either on family history or testingnot covered by the law (such as testing in a foreign nation), purchasing thisbundle of genetic testing and "insurance premium insurance" with greaterfrequency than those without suspicion.

D. Contract Regulation

9. Introduction

Two significant themes pervade regulation of insurance conflicts: the first is theextent to which government intervention is desirable to protect the insured fromthe presumed power of the insurer; a second is the extent to which governmentintervention is desirable to protect non-parties to the transaction from thecoalition created by insurer and insured.

10. Insured Protection

10.1 Contra Proferentum

Construing ambiguous contracts against the drafter, usually the insurer, haslong been part of American insurance law. One economic theory forunderstanding this doctrine of "contra proferentum" sees it as beneficiallyinternalizing the social costs fostered by ambiguity. Imposing a "penalty" forambiguous drafting in the form of expanding the enforceable promises made bythe drafter or contracting the conditions protecting the drafter theoreticallyinduces the drafter to avoid ambiguity. An overlapping economic theory seescontra proferentum as correcting information problems in the insurance market:interpreting ambiguities in the contract in favor of the insured provides theinsured with the optimal expected coverage; the insurer's foreknowledge of thedoctrine insures that the insured pays the correct premium for the expandedcoverage. Abraham (1996)

While the theory behind contra proferentum, so stated, may not beparticularly controversial, its application in sculpting existing insurance law is.To begin with, if the non-drafting party can understand the terms of the contractwithout incurring significant cost, as is the case when the insured is sop-histicated or a "repeat player" in the transactional form, the assignment ofliability to the drafting party appears arbitrary. Indeed, under such circumstancesthe labeling of one party as "the drafter" appears arbitrary as well. Many courtsnonetheless hold the doctrine still to apply in this circumstance. Ostrager &Newman (1995) This arbitrariness might be tempered, to be sure, if the partieswere costlessly able to reallocate the private burdens of ambiguity through theinsurance contract itself, but impasse costs, bargaining costs, and the risk thata court would not recognize the reallocation of risk would hinder such a cure. Rappaport (1995)

Second, while the burdens of ambiguity fall substantially on society in theform of higher dispute resolution costs, the benefits of the doctrine accruelargely to the non-drafting party (at least in the case of ambiguity unrecognizedby the drafter). The doctrine thus creates a disincentive for non-drafting partiesto become educated and, indeed, creates a possible incentive for non-draftingparties to prefer contracts that have unrecognized ambiguities. Moreover, somehave argued that the doctrine of contra proferentum discourages innovativeterms in contract drafting because new terms are subject to the doctrine whereasold terms, already having been interpreted by the courts, are better understood. Goetz & Scott (1985) The social costs of ambiguity - more judicial disputes - arethus not reduced as much as might initially be thought by the existing judicialdoctrine of contra proferentum. Examination of contract terms by governmentofficials before litigation arises, though creating its own difficulties, may thushave some place in an economic theory of insurance regulation.

Scholars have also profitably used the tools of economics to classify thevariants of contra proferentum found in American courts. Professor Abraham,for example, drawing on traditional law and economics scholarship, describes theoptimal degree of clarity as one that would minimize the sum of maldrafting costs(uncertainty, unwarranted reliance) plus maldrafting avoidance costs (lawyertime and lengthier policies Rappaport (1995) ) Abraham (1996) . Negligence-likeregimes are those in which contra proferentum is invoked only where the degreeof clarity created by the drafter falls short of the optimal level. Strict liabilityregimes are those in which contra proferentum is invoked whenever anambiguity is found. Drawing on recent scholarship regarding contract defaultrules Ayres & Gertner (1989) , Abraham further classifies legal regimes regardingcontra proferentum as penalty or majoritarian-oriented. "Majoritarian"approaches alter the terms of a contract deemed ambiguous (by one of the abovemeasures) to that which a majority of policyholders would prefer - and be willingto pay for. Penalty approaches alter the terms of the contract to that which thepolicyholder would prefer if the expanded coverage were provided for free.While negligence/majoritarian approaches to the doctrine of contra proferentumhave some appeal, the popular strict liability/penalty approach has the virtue ofavoiding expensive fact-finding on issues that courts are unlikely to resolvecorrectly much more often than chance.

10.2 Extracontractual Damages

A significant development in American insurance law has been the willingnessof courts and legislatures to grant insureds injured by breach of an insurancecontract damages in excess of that which would have been paid by the insurerhad it honored its contractual obligations. These "extra-contractual damages"have been controversial since their inception, have been the subject of someretrenchment in recent years Macintosh (1994) , and have the subject ofconsiderable economic analysis.

There are two sorts of arguments made for augmenting damages for breachof an insurance contract. One, not the subject of examination in this entry, restson the general inadequacy of traditional contract damages and applies thosegrounds in the insurance arena. Sebert (1986) A second type of argumentjustifying extracontractual damages is that otherwise the winning strategy forinsurers is frequent "opportunistic breach" with respect to meritorious claims.This strategy is particularly effective against those who show no signs ofsophistication and with respect to smaller claims. Abraham (1986)

Economic arguments in favor of extracontractual damages generally rest onforms of information failure. Unsophisticated insureds, often ignorant of theiractual rights under the policy, particularly as expanded by doctrines such ascontra proferentum, fear that if they sue and lose they will be out significant timeand money. Attorneys are unlikely to correct any misinformation becauseinsureds have imperfect access to the legal marketplace even for advice. Curtis(1986) Moreover, even insureds accessing the legal market are likely to stall therebecause attorneys will have difficulty cheaply determining the merits of theinsured's case, particularly where the insurer has failed to create a detectablepattern signaling wrongful denials. And even an attorney determining that thecase has merit will be unlikely to bring it because of the small rewards relative tothe cost of the case. The reward is small because the insurer pays only what itowed in the first place, plus any positive difference between court-systeminterest rates and the money earned by holding on to the money, plus someattorneys fees, possibly including those of the insured. Jerry (1986)

According to this theory, the "equilibrium" under a legal regime denyingextracontractual damages is for the insurer to breach frequently and the insuredfrequently to do nothing about it. The consequence is lesser entry intoinsurance transactions and less risk transfer than desirable. Extracontractualdamages thus creates a Pareto superior contract as a default rule. Perlstein (1992)

Against this theory rests a basic question: if extracontractual damages areso superior why do they not exist in a competitive insurance market absent legalintervention? If insureds operating in a competitive insurance market really feltit were necessary and desirable to have extracontractual damages, some sort ofmechanism would be built in to the contract, such as a formula damage clausefor some subset of failures by the insurer timely to pay a meritorious claim. Yetsuch a system was seldom if ever seen. In part, economic critics ofextracontractual damages say, this absence may exist because the harmfulreputation an insurer would acquire through consistent bad faith denials of aclaim would lead to market punishment at least as harsh as any the legal systemcould provide. And even if the market failed, public regulators and licensuremechanisms could succeed. It may not exist because insurers and insureds knowthat the legal system would be unable to distinguish "bad faith" breach fromsimple mistake. This inability would lead to precisely what has occurred:perversions of desired coverages, induced breach Richmond (1994) Schmidt(1994) , and needlessly higher premiums for insureds. And it may be absent inpart because the legal market functions better than proponents ofextracontractual damages assert. Sykes (1994a) Awards of legal fees toprevailing parties substantially correct the attorney incentive problems withoutresort to radical reformulation of contract law. The ability of some insured's to"sell" their rights against insurers in a competitive market to others whose needfor funds is less immediate may sometimes hinder the ability insurers wouldotherwise possess to force cheap settlements out of insureds. Finally, someeconomic scholars have argued that (ex ante) insureds subject to suffering acatastrophic loss of an irrepaceable commodity want less than full compensationfor their financial loss rather than the extra compensation provided by muchcurrent law on extra-contractual damages. Cook & Graham (1977) Fenn (1987) This attack on extracontractual damages is not without problems. Deterrenceof opportunistic breach through reputational injury rests on the dubious abilityof the market accurately to distinguish breach from lawful claims denial. Publicregulators are often underfunded, incompetent or captured by insurers. Thecompetitive market that supposedly exists in fact consists of insurers who maylawfully collude. An award of simple legal fees to prevailing insureds may notentice attorneys uncertain of their clients' prospects. The supposed cure of theinsurer's ability to exploit the weakness of an insured through assignment ofclaims relies on markets that seldom exist precisely because such transactionsare often difficult and, in some instances illegal. State Farm Fire & Casualty Co.v. Gandy (1996)

10.3 Conflicts of Interest in Liability Insurance Policies

Protection of the insured is likewise illustrated by judge-made law in the UnitedStates governing conflicts of interest created by the contract between a liabilityinsurer and its insured over private settlement of legal actions brought againstthe insured. The conflict, which resembles that arising in the corporate contextbetween various forms of claimants against corporate assets, arises from theextraordinarily complex decomposition of responsibility liability insurancepolicies almost always create.

The responsibilities created by typical liability insurance policies inconjunction with legal doctrine may be crudely represented by positions alongthree not entirely orthogonal axes. One axis distinguishes responsibility for"discharge" payments such as settlements and judgments made to reduce theliability of the insured from "defense" payments made on behalf of the insuredin defending against claims of liability. Certain discharge payments such asthose that "exhaust" certain limits of the policy may eliminate future defenseobligations of the insurer. This feature of conventional liability insurancecontracts effectively gives the insurer a "put" on future defense obligationsprovided it can completely discharge its insured's liability to one potential victimof the insured. Thus, a conflict occasionally arises between insurers, who wantto exercise this put (at a price negotiated with a victim of the insured) and aninsured, who does not want to take on defense responsibilities even when theinsurer's failure to exercise the put increases at least one insured's expectedpayments in discharging liabilities to various victims. Texas Farmers Ins. Co. v.Soriano (1994) The law is not yet well developed on whether the insurer will bepermitted unfettered discretion in exercising its apparent contractual authorityin this area.

One axis that creates frequent conflicts distinguishes responsibility for"discharge" payments made to reduce the liability of the insured based on theamount of the payment in question. Often, the insured takes the first level ofresponsibility pursuant to deductible or "self-insured retention" provisions inits policy. A primary insurer will often take the next level of responsibility up tosome policy limit, but with payments thereunder reducing its responsibility forother certain other judgments against the insured and, potentially, litigationexpenses in other cases. An excess insurer may take the next level ofresponsibility up to some policy limit, with the insured having responsibility forjudgments beyond the final layer of excess. All of these responsibilities can beshifted in whole or in part to reinsurers. To the extent that litigation strategies,such as reactions to settlement offers, can shape the probability distribution offuture discharge and defense obligations, which distribution in turn determinesthe expected loss from the case suffered by each party, the decomposition alongthis axis creates conflicts of interest among the various parties regardinglitigation strategy. Chandler (1993) Sykes (1994a) Sykes (1994b) Silver &Syverud (1995) Pryor (1997)

In the absence of clear language to the contrary in the insurance contract,courts typically react to this conflict by simplifying the decomposition in a waysomewhat analogous to a judically-imposed corporate restructuring. In the faceof an opportunity to settle for some "reasonable amount" with the plaintiff, theinsurer that fails to tender the requisite amount or who acts in "bad faith" withrespect to settlement of the case, is held to "own" all layers above the one itnominally owns by contract. The faulty layer is thus liable directly to the victimfor judgments piercing that layer or, through subrogation and other mechanisms,to higher layers who discharged the responsibilities that would have been theirsabsent the fault of a lower layer. Stuhr (1992) And, in most Americanjurisdictions, this liability exists even where the higher layer would not actuallyhave the resources to pay the judgment.

Economic theory might well view this restructuring as one vehicle (amongothers) for correcting inefficient behavior through internalizing of the costscreated when parties in control of litigation strategy are otherwise able to foistthe costs of certain strategies on to others. The restructuring often inducesinsurers with settlement responsibility to make decisions that maximize the jointwealth of insurer and insured, thereby leading to optimal consumption of liabilityinsurance. The problem, as noted by two scholars Sykes (1994a) Logue (1994) occurs when the higher layer (usually the insured) would not have the financialresources to pay the judgment. Under these circumstances, the conventionallegal rule induces settlements that fail to maximize joint wealth, which in turnleads to underconsumption of liability insurance, which in turn leads toundercompensation of victims. A minority legal doctrine has recently re-emergedthat limits the liability of the faulty layer to no more than the higher layers wouldactually have been able to pay. While this "Michigan rule" may cure thedistortions in the settlement process caused by the traditional legaldecomposition of responsibility, it may create other problems such asundercompensation of victims and underdeterrence of wrongful conduct.Whatever the merits of the various rules, however, there is always the theoreticalpossibility of Coasean bargaining to restore efficiency Sykes (1994b) but it isunlikely to be particularly useful here given (a) the high costs of bargaining inthis extraordinarily complex field and (b) a likely judicial reluctance to acceptprivate party modification of legal rules.

Interestingly, consonant with economic theory, liability insurance contractstend not to grant the insurer authority to settle cases without trial absent theconsent of the insured when the insured's reputation and future income dependon exoneration from the allegations made in the lawsuit. Mayers & Smith (1981) Courts, concerned about the major change in expected discharge responsibilitiescreated by such a provision, generally refuse, however, to imply it into contractsotherwise silent on the matter. Dear v. Scottsdale Insurance Company (1997) ; McKinley, Moody & Barlow (1997) Insurers and insureds protect againstexploitation of this consent clause by frequently providing that a settlement heldup by the insured's non-consent re-layers discharge responsibilities such thatthe insured takes on responsibility for discharge payments in excess of theamount of the rejected settlement. Rauch (1986)

A third axis in which conflicts of interest arises surrounds the liabilityinsurance contract's decomposition of responsibility for various types of events.The insurer, for example, generally has discharge obligations arising out ofliability from negligence but generally does not have discharge obligationsarising out of more serious types of fault. This conflict leads to an intricate balletof litigation strategies and counterstrategies described well by one recentscholar. Pryor (1997) In essence, however, the injured party may "underlitigate"the dispute in an effort both to induce the insurer to incur heightened expecteddefense and discharge obligations, which increase may be partly appropriatedby the injured party through offering to withdraw the triggering underlyingdispute against the insured in exchange for a payment mostly from the insurer.

11. Public Protection

11.1 Anti-Diminution

In the past, some liability or indemnity insurers sought to reduce the amountthey paid victims of their insureds based on the insolvency of their insureds.The insurers' theory was that their own liability derived from actual payment orthe possibility of payment by the insured, which possibility was reduced by theinsured's insolvency. "Anti-diminution laws," or "bankruptcy provisions" nowfrequently prohibit this practice Jerry (1987) and are a classic example ofinsurance regulation acting to the detriment of the insured and in favor of thepublic at large. Assuming with Huberman that an insured's utility of wealth isconstant below some "insolvency" level, but that the indemnity owed as afunction of insured loss does not have to be non-decreasing, an insured wouldprefer a policy whose indemnity obligation disappeared above some threshholdlevel of loss to a policy that had some constant upper limit regardless of theamount of the loss. Indemnity obligations on the part of the insurer that fail toincrease the utility of the insured in the event of some serious adverse increaseneedlessly decrease wealth in all other states by increasing the actuariallyjustified premium owed by the insured for the insurance. And, indeed, during theearly years of indemnity/liability insurance, such "diminution clauses" existedin insurance policies.

The problem with such "diminution clauses" is that they result in inadequatecompensation to victims of the insured and exacerbate the moral hazard problemcaused by the existence of insurance. Shavell (1986) Many American states nowrequire that insurance policies not diminish on account of the insolvency of theinsured.

11.2 Treatment of Proceeds in Insolvency

The conflict between the interests of the insured and those of potential strangervictims of the insured shows itself again in the legal treatment of liabilityinsurance proceeds upon the insolvency of the insured. Treating these proceedsas property of the estate and thus available to all creditors of the insured,including contract creditors, has the potential generally to enrich the insured bypermitting it to borrow at lower interest rates, which inures to the benefit of theinsured in the event a subsequent insolvency does not occur. Treating theproceeds as dedicated to the victims of conduct covered by the insurancepolicy, as the law now often does, means that uncovered claimants such ascontract creditors or victims of tortious conduct excluded from the insurancepolicy's indemnity obligations must look solely to the insured's non-insuranceassets in the event of an insolvency. This treatment thus has the potential toincrease the interest rate the insured must pay. This result likewise may occurwhere the proceeds of the policy are nominally part of the insured's estate butwhere the law permits victims of conduct covered by the insurance policy to"cut through" and bargain with the insurer for a direct payment in exchange fora reduction of the claim against the insured.

11.3 Extreme Conduct Limitations

Government limitations on the ability of insureds to protect themselves againstliability either for their intentional misconduct to others or for conduct deemedappalling or sadistic enough to be worthy of "punitive" or "exemplary" damagesfurther illustrates intervention in the market to protect potential victims of theinsured's and insurer's cooperation. Chandler (1996) Generally, these limitationsrestrict coverage for appalling conduct. Other times, however, notably in the areaof automobile insurance, government requires coverage for appalling conductso as not to defeat a "compulsory insurance" regime State Farm Fire & CasualtyCo. v. Tringali (1982)

Many commentators applaud limitations on the insurability of punitivedamages or damages for intentional conduct on grounds that such restrictionsprevent undesirable moral hazard. Ellis (1982) These efforts generally fail toconfront the questionable need for such governmentally imposed constraints,however, given the ability of insurers who fear moral hazard and who will paythe price for failure to control it on their own through exclusions and conditionsin the insurance contract. One explanation is that the refusal to enforce contractspermitting indemnification for severely deviant conduct lets insurers andinsureds, apparently confused at the time of contracting, "come to their senses."Such a legally created "reprieve" from a faulty contract may be particularlyimportant given the possibility that indemnity for punitive damages will increasethe frequency of wrongful behavior by insureds and the frequent inability of thelaw, even with punitive damages, to restore their victims to the status quo exante. This argument explanation draws support from simulations conducted byone scholar suggesting that few rational contracting parties would want toprovide indemnity for conduct grossly below an efficient level of care, even withcourt systems prone to error. Chandler (1996) Perhaps the simplest explanationmade with respect to the non-insurability of punitive damages is that deterrencemodels are irrelevant to punitive damages and that punishment is destroyedwhere punishment is indemnified. Ellis (1982)

E. Competition Law Exemptions

In the United States and in Europe, insurers have been exempt from many"competition" or "antitrust" laws prohibiting forms of cooperative behavior bydistinct insuring entities. In the United States, this exemption occurs via theMcCarran-Ferguson Act (15 United States Code ¤ 1011-15 insulating insurersfrom most of federal antitrust laws, and by "mini McCarran-Ferguson Acts"enacted by some states insulating insurers to varying degrees from the states'own laws barring combinations and conspiracies in restraint of trade. In Europe,it occurs via Commission Regulation promulgated in 1992 under Article 85(3) ofthe Treaty of Rome. This regulation generally permits certain cooperativepractices in the "insurance sector" such as cooperation with respect toestablishment of premiums, standard policy forms and coverages that mightotherwise be prohibited by Article 85(1) of the Treaty and the regulationsthereunder.

An economic theory has emerged to justify this departure from the generalnotion that competition rather than cooperation maximizes welfare: thesignificant economies of scale in production of information needed to preventadverse selection and insolvency. Miller & Macey (1993b) Danzon (1992) Achampong (1991) Without significant data on the experience of policyholders,it becomes difficult reliably to model the expected loss posed by individualpolicyholders and therefore accurately to establish actuarially sound premiums.Individual insurers, even large insurers operating multiple lines in manyjurisdictions, may lack adequate internal historic data to perform the necessaryanalyses with confidence. Some have argued that this inability extends even toprojections of future losses on future policies. Collective analysis on this pointshould thus be exempt from antitrust scrutiny. Danzon (1992)

Advocacy of cooperation rather than competition in the business ofinsurance extends to collective drafting of policy forms. Without common forms,the data needed to make reliable predictions would be "noisy" and thus lessamenable to accurate, fine-grained statistical analysis. Macey & Miller (1993a) Miller & Macey (1993b) Consumers are thought better able to shop whenpolicies have only a few varying parameters, such as price. Uniformity in policyprovisions thus actually favors competition, this argument proceeds. Moreover,reinsurance, which serves an economically beneficial specialization function,would be hampered were collective policy forms barred and certain lines ofinsurance made more custom and less of a "commodity." Ayres & Siegelman(1989)

Two other economic arguments also frequent debates about application ofcompetition laws to the insurance industry. First, scholars such as Robert Bork Bork (1978) generally question whether collusive behavior between entitiesshould ever be illegal so long as the law would permit those entities to merge.Limited cooperation between competitors might well be better than theelimination of competition through merger. Bork (1978) Exemption fromcompetition laws thus fosters the existence of numerous small insurers and,derivatively, competition. Danzon (1992) Second, and somewhat relatedly, somescholars have suggested that legally supported cooperation helps prevent"cut-throat" or "anarchic" competition that, at least historically, appears to havejeopardized the solvency of insurers. Carlson (1979) McDowell (1987)

Several scholars have noted, however, the scope of the American exemptionfrom competition laws is considerably broader than the arguable economicjustification. Macey & Miller (1993a) Achampong (1991) Collective datagathering would almost certainly be protected by modern interpretations ofexisting American competition laws, even were the McCarran-Ferguson Act notin place. Moreover, a need for collective data gathering would not seem tojustify collective data analysis, nor would it justify collective pricing ofinsurance policies, both of which are frequently permitted in the United Statesthough practiced somewhat less frequently that in years past. Moreover, if thereare economies of scale in data collection, it is not clear why for insurance but notfor other industries the law would "prop up" small companies operating at lessthan an efficient scale by permitting them to collude with each other in a wayother than lawful merger or joint research permitted for all industries underfederal laws enacted in the 1980s. This critique would appear to be supported byempirical research suggesting that the economies of scale in the insuranceindustry are limited and that many insurers could survive at the minimumefficient scale. Geehan (1986) Joskow (1992) It has also been suggested that thecollusive rating practices facilitated by the exemption from competition lawsinduces sloppy and inefficient cost control measures by insurers. Angoff (1988) Finally, collective data gathering may trap underwriting practices in some localoptimum whereas the greater experimentation facilitated by barring collusionmight lead to more accurate methods of assessing policyholder risk.

Interestingly, the more recent regulations of the European Community appearto echo at least some of these criticisms of the broad American exemption fromcompetition laws. Under Regulation 3932/92, for example, collaborativecompilation of statistics on expected losses is permitted as is statistical analysisof that data. Collective determinations of actual premiums, after expense loadingand other cost factors are taken into account, is prohibited, however.Obligations among competitors to adhere to a joint pricing mechanism are void.While development of common policy forms and coverages are permitted,insurers are not permitted to enter into contracts with each other than wouldcompel use of these forms and coverages.

F. Conclusion

Sensible regulation of insurance requires an understanding of a broad span ofeconomics including finance theory, price theory, game theory, dynamics,international trade, and econometrics as well as ancillary subjects such asstatistical and information theory. It also requires an understanding ofjurisprudential issues such as the appropriate specificity of rules, the appropriatecentralization of legal authority across networked cultures, the appropriateallocation of law-making authority between legislature and judiciary, and thecosts of various forms of dispute resolution. While the broad contours ofinsurance regulation have been profitably and extensively examined using toolsfrom each of these disciplines separately, and while this entry evidences someconsiderable progress in a more interdisciplinary approach, many of theintricacies of legal doctrine that determine the actual course of insurancetransactions remain unconquered by the powerful combination of law andeconomics.

Bibliography on Insurance Regulation (5700)

Abraham, Kenneth S. (1985) , 'Efficiency and Fairness in Insurance Risk Classification', 71 Virginia Law Review , 403-451

Abraham, Kenneth S. (1986) , Distributing Risk: Insurance, Legal Theory and Public Policy ,New Haven, Yale University Press.

Abraham, Kenneth S. (1991) , Environmental Liability Insurance Law , Prentice Hall Law &Business.

Abraham, Kenneth S. (1996) , 'A Theory of Insurance Policy Interpretation', 95 MichiganLaw Review , 531-569.

Achampong, Francis (1991) , 'The McCarran-Ferguson Act and the Limited InsuranceAntitrust Exemption: An Indefensible Aberration', 15 Seton Hall Legislative Journal ,141-169.

Akerloff, George (1970) , 'The Market for Lemons: Quality Uncertainty and the MarketMechanism', 84 Quarterly Journal of Economics 488-500.

Angoff, Jay (1988) , 'Insurance Against Competition: How the McCarran-Ferguson ActRaises Prices and Profits in the Property-Casualty Insurance Industry', 5 Yale Journalon Regulation , 397-415.

Ayres, Ian & Siegelman, Peter (1989) , 'The Economics of the Insurance Antitrust Suits:Towards an Exclusionary Theory', 63 Tulane Law Review , 971-997.

Borenstein, Severin, (1989) , 'The Economics of Costly Risk Sorting in CompetitiveInsurance Markets,' 9 International Review of Law and Economics 25-39.

Carlson, Larry D. (1979) , 'The Insurance Exemption from the Antitrust Laws', 57 TexasLaw Review 1127 ff.

Chamberlain, John R. (1985) , 'Assessing the Fairness of Insurance Classifications ', 7 Research in Law and Economics , 65-87.

Chandler, Seth J. (1993) , 'Reconsidering The Duty to Settle', 42 Drake Law Review ,741-806.

Chandler, Seth J. (1996) , 'The Interaction of the Tort System and Liability InsuranceRegulation: Understanding Moral Hazard', 2 Connecticut Insurance Law Journal, 91-177.

Christianson, David J. (1996) , 'Genetic Testing: Risk classification and adverse selection ', 15 Journal of Insurance Regulation , 75-79.

Cook, P.J. & Graham, D.A. (1977) , 'The Demand for Insurance and Protection: The Caseof Irreplaceable Commodities', 91 Quarterly Journal of Economics ,143-156.

Crocker, Keith J. & Snow, Arthur (1986) , 'The Efficiency Effects of CategoricalDiscrimination in the Insurance Industry', 94 Journal of Political Economy , 321-344.

Cummins, J. David (1988) , 'Risk-Based Premiums for Insurance Guaranty Funds' 43 Journalof Finance , 823-839.

Curtis, Linda (1986) , 'Damage Measurements for Bad Faith Breach of Contract: AnEconomic Analysis', 39 Stanford Law Review , 161-185.

Danzon, Patricia M. (1992) , 'The McCarran-Ferguson Act: Anticompetitive orProcompetitive', Regulation , Spring 38-47.

Ellis, Dorsey D. (1982) , 'Fairness and Efficiency in the Law of Punitive Damages', 56 Southern California Law Review, 1-78.

Fama, Eugene F. (1980) , 'Agency Problems and the Theory of the Firm', 88 Journal ofPolitical Economy , 288 ff.

Fenn, Paul (1987) , 'Evaluating Irreplaceable Loss: Some Implications for Insurance andLiability Rules', 12 Geneva Papers on Risk and Insurance, 158-167 .

Finsinger, Jorg & Pauly, Mark (1992) , 'Reserve Levels and Reserve Requirements forProfit-maximizing Insurance Firms' in: Dionne, George & Harrington, Scott E., eds., Foundations of Insurance Economics , Boston, Kluwer.

Finsinger, Jörg, Hammond, Elizabeth & Trapp, Julian, eds. (1985) , Insurance: Competitionor Regulation?, London, The Institute for Fiscal Studies.

Frech, H. E. & Samprone, Joseph C. (1980) , 'The Welfare Loss of Excess NonpriceCompetition: The Case of Property-Liability Insurance Regulation' 23 Journal of Lawand Economics , 429-440.

Geehan, Randall (1986) , 'Economies of Scale in Insurance: Implications for Regulation', in:Wasow, Bernard & Hill, Raymond D., eds., The Insurance Industry in EconomicDevelopment , New York, New York University Press, 137-159.

Goetz, & Scott (1985) , The Limits of Expoanded Choice: An Analysis of the InteractionsBetween Express and Implied Contract Terms, 73 California Law Review , 261-322.

Gron, Anne (1994a) , 'Evidence of Capacity Constraints in Insurance Markets', 37 Journalof Law and Economics , 349-377.

Gron, Anne (1994b) , 'Capacity Constraints and Cycles in Property-Casualty InsuranceMarkets', 25 RAND Journal of Economics , 110-127.

Hansmann, Henry (1985) , 'The Organization of Insurance Companies: Mutual versus Stock', 1 Journal of Law, Economics and Organization 125-153.

Hogan, Arthur M. B. (1995) , 'Regulation of the single European insurance market', 13 Journal of Insurance Regulation , 329-358.

Huberman, Gur, Mayers, David & Smith, Clifford W. (1983) , 'Optimal insurance policyindemnity schedules', 14 Bell Journal of Economics , 415-425.

Jacobi, John V. (1997) , 'The Ends of Health Insurance,' 30 University of California DavisLaw Review 311-404.

Jerry, Robert H., II (1987) , Understanding Insurance Law, ¤ 95c.

Jerry, Robert H., II (1986) , 'Remedying Insurer's Bad Faith Contract Performance: AReassessment', 18 Connecticut Law Review , 271-321.

Jones, Brian A. (1996) , 'Insurance Regulation in the European Union', 47 Federation ofInsurance & Corporate Counsel Quarterly , 75-93.

Joskow, Paul L. (1992) , 'Cartels, Competition and Regulation in the Property-LiabilityInsurance Industry' in: Dionne, Georges & Harrington, Scott E., eds., Foundations ofInsurance Economics , Boston, Kluwer.

Kimball, Spencer L. (1961) , 'The Purpose of Insurance Regulation: A Preliminary Inquiryin The Theory of Insurance Law', 41 Minnesota Law Review 471-524.

Kimball, Spencer L. & Denenberg, Herbert S. eds. (1969) , Insurance, Government, AndSocial Policy , Homewood, Illinois: Richard D. Irwin.

Logue Kyle D. (1994) , 'Solving the Judgment-Proof Problem', 72 Texas Law Review ,1345-1394.

Macey, Jonathan R. & Miller, Geoffrey, P. (1993a) , 'The McCarran-Ferguson Act of 1945:Reconceiving the Federal Role in Insurance Regulation, 68 New York University LawReview , 13-88.

Macey, Jonathan R. & Miller, Geoffrey P. (1993b) , Costly Policies: State Regulation andAntitrust Exemption in Insurance Markets , Washington, D.C., American EnterpriseInstitute Press.

Macintosh, Kerry L. (1994) , 'Gilmore Spoke too Soon: Contract Rises from the Ashes of theBad Faith Tort', 27 Loyola of Los Angeles Law Review , 483 ff.

Mayers, David & Smith, Clifford W., Jr. (1981) , 'Contractual Provisions, OrganizationalStructure, and Conflict Control in Insurance Markets', 54 Journal of Business , 407-434.

Mayerson, Allen L. (1969) , 'Ensuring The Solvency of Property and Liability InsuranceCompanies' in: Kimball, Spencer L. & Denenberg, Herbert S. eds. Insurance,Government, And Social Policy , Homewood, Illinois, Richard D. Irwin.

McDowell, Banks (1987) , 'Competition as a Regulatory Mechanism in Insurance', 19 Connecticut Law Review , 287-310.

McGoodwin, Wendy L. (1996) , 'Genetic Testing: A Tool for Doctors, not for Insurers', 15 Journal of Insurance Regulation , 71-74.

Perlstein, Barry (1992) , 'Crossing The Contract-Tort Boundary: An Economic ArgumentFor The Imposition of Extracompensatory Damages For Opportunistic Breach ofContract', 58 Brooklyn Law Review , 877-911.

Pokorski, Robert J. (1995) , 'Genetic Information and Life Insurance', Nature , July 6, 1995,13-14.

Pokorski, Robert J. (1997) , 'Insurance Underwriting in the Genetic Era', 60 AmericanJournal on Human Genetics , 205-216.

Pryor, Ellen S. (1997) , 'The Stories We Tell: Intentional Harms and the Quest for InsuranceFunding', 75 Texas Law Review , 1721-1764.

National Association of Insurance Commissioners (1996) , 'Report of the NAIC GeneticTesting Working Group', 15 Journal of Insurance Regulation , 7-65.

Rappaport, Michael B. (1995) , 'The Ambiguity Rule and Insurance Law: Why InsuranceContracts Should Not Be Construed Against the Drafter', 30 Georgia Law Review 171-257.

Raviv, Artur (1992) , 'The Design of an Optimal Insurance Policy' in: Dionne, George &Harrington, Scott E., eds., Foundations of Insurance Economics , Boston, Kluwer.

Rea, Samuel A., Jr. (1992) , 'Insurance Classifications and Social Welfare' in: Dionne, Georgesed., Contributions to Insurance Economics , Boston, Kluwer.

Rea, Samuel A., Jr. (1993) , 'The Economics of Insurance Law', 13 International Review ofLaw and Economics , 145-162.

Rothschild, Michael & Stiglitz, Joseph E. (1976) , 'Equilibrium in Competitive Insurancemarkets: An Essay on the Economics of Imperfect Information', 90 Quarterly Journalof Economics, 629-649 .

Rothstein, Mark A. (1993) , 'Genetics, Insurance and the Ethics of Genetic Counseling', 3 Molecular Genetic Medicine, 159-177.

Rothstein, Mark A. (1997) , 'Genetic Secrets: A Policy Framework' in Rothstein, Mark A.ed., Genetic Secrets: Protecting Privacy and Confidentiality in the Genetic Era , NewHaven, Yale University Press.

Schmalensee, Richard (1984) , 'Imperfect Information and the Equitability of CompetitivePrices', 99 Quarterly Journal of Economics 441-460.

Sebert, John A., Jr. (1986) , 'Punitive and Nonpecuniary Damages in Actions Based UponContract: Toward Achieving The Objective of Full Compensation', 33 University ofCalifornia Los Angeles (UCLA) Law Review 1565-1678.

Shavell, Steven (1986) , 'The Judgment Proof Problem', 6 International Review of Law andEconomics, 45-58.

Shavell, Steven (1987) , An Economic Analysis of Accident Law , Cambridge, HarvardUniversity Press.

Silver, Charles & Syverud, Kent (1995) , 'The Professional Responsibilities of InsuranceDefense Lawyers', 45 Duke Law Journal 255-363.

Smith, Clifford W. & Warner, Jerold B. (1979) , 'On Financial Contracting: An Analysis ofBond Covenants', 7 Journal of Financial Economics , 117-161.

Stigler, George J. (1968) , 'Price and Nonprice Competition,' 76 Journal of PoliticalEconomy , 149 ff.

Stone, Deborah A. (1996) , 'The Implications of the Human Genome Project for Access toHealth Insurance', in The Human Genome Project and the Future of Health Care,Indiana University Press.

Sykes, Alan O. (1994a) , 'Bad Faith' Refusal to Settle By Liability Insurers: Some Implicationsof the Judgment-Proof Problem', 23 Journal of Legal Studies , 77-110.

Sykes, Alan O. (1994b) , 'Judicial Limitations on the Discretion of Liability to Settle orLitigate: An Economic Critique', 72 Texas Law Review 1345-1374.

Sykes, Alan O. (1996) , 'Bad Faith Breach of Contract By First-Party Insurers', 25 Journalof Legal Studies , 405-444.

Tabarrok, Alexander (1994) , 'Genetic testing: an economic and contractarian analysis', 13 Journal of Health Economics , 75-91.

Van Den Berghe, Lutgart, (1990) '(De)Regulation of Insurance Markets' in: Henri Loubergeed., Risk, Information and Insurance: Essays in the Memory of Karl H. Borch , Boston,Kluwer, 1990, 199-220.

Winter, Ralph A. (1988) , 'The Liability Crisis and the Dynamics of Competitive InsuranceMarkets', 5 Yale Journal on Regulation , 455-499.

Winter, Ralph A. (1991a) , 'Solvency Regulation and the Property-Liability 'InsuranceCycle'', 29 Economic Inquiry, 458-71.

Winter, Ralph A. (1991b) , 'The Liability Insurance Market', 5 Journal of EconomicPerspectives, 115-136.

Wortham, Leah (1986) , 'The Economics of Insurance Classification: The Sound of OneInvisible Hand Clapping', 47 Ohio State Law Journal, 835-890.

Other References

Ayres, Ian & Gertner, Robert (1989) , 'Filling Gaps in Incomplete Contracts: An EconomicTheory of Default Rules,' 99 Yale Law Journal , 87-130.

Bork, Robert H. (1978) , The Antitrust Paradox: A Policy At War With Itself , New York, BasicBooks.

Harsanyi, John (1967) , 'Games with Incomplete Information Played by 'Bayesian' PlayersI: The Basic Model', 14 Management Science , 159-182.

Horwitz, Morton J., (1992) , The Transformation of American Law , New York, OxfordUniversity Press.

Howard, Lisa S. (1996) , 'Genetic testing debate starts to bubble in U.K.', NationalUnderwriter , August 5, 45-46.

Kimball, Spencer L. (1992) , Cases and Materials on Insurance Law , Boston: Little Brown.

Lewis, Stephen & Coates, Katherine (1991) , 'European Perception of U.S. State Regulation'in: Semaya, Francine & Vitkowsky, Vincent J., eds., The State of Insurance Regulation (American Bar Association).

Ostrager, Barry R. & Newman, Thomas R. (1995) , Handbook of Insurance CoverageDisputes , Prentice Hall Law And Business, ¤1.03.

McKinley, Hall F., III, Moody, G. Randall & Barlow, Peter B. (1997) , 'Issues in the Selectionof Counsel and Control of Litigation When The Insured Has a Self-Insured Retention', 32 Tort and Insurance Law Journal , 769-786.

Polinsky, A. Mitchell & Shavell, Steven (1979) , 'The Optimal Tradeoff Between theProbability and Magnitude of Fines', 69 American Economics Review , 880 ff.

Rasmusen, Eric (1994) , Games and Information: An Introduction to Game Theory , 2d. ed.,Cambridge, Massachusetts, Blackwell.

Rauch, Michael, et. al. (1986) , 'Settlement of Professional Liability Claims' in: ProfessionalLiability Insurance for Attorneys, Accountants, and Insurance Brokers 1986 , PracticingLaw Institute.

Rawls, John (1971) , A Theory of Justice , Cambridge, Massachusetts, Belknap Press ofHarvard University Press.

Richmond, Douglas R. (1994) , 'The Two-Way Street of Insurance Good Faith: UnderConstruction, But Not Yet Open' 28 Loyola University Chicago Law Journal 95-140.

Riegel, Robert (1916) , 'Fire Insurance Rates: Problems of Cooperation, Classification,Regulation', 30 Quarterly Journal of Economics , 704-737.

Schmidt, Stephen R. (1994) , 'The Bad Faith Setup', 29 Tort and Insurance Law Journal 705-739.

Selten, Reinhard (1975) , 'Re-examination of the Perfectness Concept for Equilibrium Pointsin Extensive Games', 4 International Journal of Game Theory , 25-55.

Sankey, Keith (1997) , 'UK: Insurers Measure Up Genes', Reinsurance (April 30, 1997)

Stuhr, V. Eileen (1992) , 'Excess Carrier Tort Suits: Are Primary Insurers Up The CanalWithout a Paddle?', 29 Houston Law Review , 661-717.

Cases

Arizona Governing Committee v. Norris (1983) , 463 United States Reports 1073-1011.

Dear v. Scottsdale Insurance Co. (1997) , 947 Southwestern Reporter (Second Series) 908-919.

Simpson v. Phoenix Mutual Life Insurance Co. (1969) , 24 New York Reporter (SecondSeries) 262-268.

State Farm Fire & Casualty Co. v. Gandy (1996) , 39 Texas Supreme Court Journal696-719.

State Farm Fire & Casualty Co. v. Tringali , 686 Federal Reporter (Second Series) 821-825.

Texas Farmers Ins. Co. v. Soriano (1994) , 881 Southwestern Reporter (Second Series) 312-318.