Albright, Justice, dissenting:

        I respectfully dissent from the majority opinion and would answer the certified question in the affirmative. The issue of whether a release or covenant not to sue an agent is tantamount to the release of the principal is an issue of first impression in West Virginia.

            A release to, or an accord and satisfaction with, one or more joint trespassers, or tort-feasors, shall not inure to the benefit of another such trespasser, or tort-feasor, and shall be no bar to an action or suit against such other joint trespasser, or tort-feasor, for the same cause of action to which the release or accord and satisfaction relates.

Prior to the 1931 amendment, that statute did not contain the term “tort-feasor.” The statute provided only as follows:

            A release to, or an accord and satisfaction with one joint trespasser, hereafter executed or had, shall not inure to the benefit of another such trespasser, and shall be no bar to an action or suit against such other joint trespasser for the same cause of action to which the release or accord and satisfaction relates.

W.Va. Code Ann., c. 136 § 7 (Barnes' Code 1923).

        The 1931 addition of the term “tortfeasor” was explained in the Revisers' Note   (See footnote 2)  as follows: “The words 'or tort feasor,' wherever occurring, are new, and are added to indicate that this section applies to all joint wrongdoers, as is held in Leisure v. Monongahela Valley Traction Co., 85 W. Va. 346, [101 S.E. 737 (1920)].” (emphasis supplied.).   (See footnote 3)  Thus, while the statute   (See footnote 4)  clarifies that a release of one tortfeasor does not release another tortfeasor, the issue of whether the release of the tortfeasor functions as a release of a non-wrongdoer such as a vicariously liable employer is not resolved thereby. The statute applies strictly and exclusively to joint wrongdoers, as elucidated in the Revisers' Note.   (See footnote 5)  Vicarious liability is a doctrine imposed as a matter of public policy, allowing inclusion of a non-wrongdoer as a party. Vicarious liability is not premised upon the commission of a tort, and there is no joint wrongdoing of the employer/principal and employee/agent.   (See footnote 6)  Where a principal is liable only through vicarious liability, such principal did not commit a negligent act and would consequently not constitute a “tortfeasor” under West Virginia Code § 55-7-12. Consequently, that statute does not provide a foundation for a conclusion that release of an agent cannot function as a release of the principal.

        West Virginia case law similarly fails to provide a definitive answer to the certified question posed. We tangentially addressed issues of the doctrine of respondeat superior in conjunction with a determination of the moral obligation of the State in State ex rel. Bumgarner v. Sims, 139 W. Va. 92, 79 S.E.2d 277 (1953),   (See footnote 7)  and acknowledged that a joint action may be maintained against a master and servant in a case in which the plaintiff's injuries were occasioned solely by the negligence of the servant.   (See footnote 8)  Id. at 111, 79 S.E.2d at 289. We explained as follows:

        Subject to the rule set forth in points 3 and 4 of the syllabus of the Humphrey case that where the master's duty is absolute and nondelegable, or where the liability of the master is not predicated solely upon the negligence of the employee impleaded, but upon the negligence of another employee, or that of the master himself, an acquittal by a jury of the servant in an action instituted against the master and servant to establish liability based solely on the servant's negligence will not release the master. The relation of master and servant in those cases, in which the doctrine of respondeat superior applies, is joint, and the parties should be regarded as though they were joint tort-feasors. Wills v. Montfair Gas Coal Co., 97 W.Va. 476, 125 S.E. 367. In some respects, however, the relation may be regarded as joint and several.

Id. at 111, 79 S.E.2d at 289.   (See footnote 9)  Syllabus point eight of Bumgarner provided:

            Subject to the rule that an acquittal by a jury of the servant in an action instituted against the master and servant to establish liability based solely on the servant's negligence, will release the master, the relation between the master and servant, the latter acting within the scope of his employment, is joint and several in the sense that both master and servant are liable for injuries caused by the negligent wrongdoing of the servant, acting within the scope of his employment, and liability for such injuries may be asserted in an action at law against the master and servant jointly or against each of them in a separate action at law.
Id. at 94, 79 S.E.2d at 280-81. The statements in Bumgarner regarding the manner in which the principal/agent relationship may resemble the joint tortfeasor relationship were made only in the “sense that both master and servant are liable for injuries caused by the negligent wrongdoing of the servant” as clearly articulated in syllabus point eight of Bumgarner. Id. Indeed, they are both liable, as joint tortfeasors would both be liable. 139 W. Va. at 94, 79 S.E.2d at 280-81. However, they are not liable as joint wrongdoers; one is culpable and one is not. Thus, while the principal/agent or master/servant relationship does resemble the joint tortfeasor relationship in limited degree, the parallels are not boundless, and the Bumgarner court neither encountered nor resolved the question presently before this Court.

        The Bumgarner court also stated that the common law rule that a valid release of the servant releases the master from liability was abrogated, in part,   (See footnote 10)  by West Virginia Code § 55-7-12. 139 W. Va. at 112, 79 S.E.2d at 290. In reviewing the Bumgarner opinion in its entirety, it appears that this statement was pure dictum. There was neither an actual release nor a covenant not to sue   (See footnote 11)  executed by the plaintiff in Bumgarner. The civil action against the wrongdoer, Mr. Coiner, had proceeded to trial, and a jury verdict had been rendered. It was simply not satisfied due to Mr. Coiner's bankruptcy. Any statements regarding the effect of a release of the wrongdoing agent upon the principal must therefore be considered dicta. Such statements were not necessary to the conclusion reached and were not restricted to the facts before the Bumgarner Court. This Court has clearly stated as follows in In re Assessment of Kanawha Valley Bank, 144 W.Va. 346, 109 S.E.2d 649 (1959):

        The rule of stare decisis does not apply where the former decisions have misunderstood or misapplied the law or are contrary to reason. Simpkins v. White, 43 W.Va. 125, 27 S.E. 361. “. . . no legal principle is ever settled until it is settled right.” Weston v. Ralston, 48 W.Va. 170, 36 S.E. 446, 450. “ . . . it is better to be right, than to be consistent with the errors of a hundred years.” Lovings v. Norfolk & W. R. Co., 47 W.Va. 582, 35 S.E. 962, 965.

Id. at 382, 109 S.E.2d at 669.   (See footnote 12) 

        This Court has also repeatedly cautioned against establishing precedent based upon dicta. As explained in Kanawha Valley Bank, “[o]biter dicta or strong expressions in an opinion, where such language was not necessary to a decision of the case, will not establish a precedent.” 144 W. Va. at 382-83, 109 S.E.2d at 669. Regarding the rationales of prior holdings and their inclusion within the doctrine of stare decisis, we distinctly stated as follows in Banker v. Banker, 196 W.Va. 535, 474 S.E.2d 465 (1996): “This doctrine concerns the holdings of previous cases, not the rationales[.]” Id. at 546 n. 13, 474 S.E.2d at 476 n. 13.

        Stare decisis is the policy of the court to stand by precedent. It is different from the doctrine of stare rationibus decidendi_“to keep to the rationes decidendi of past cases.” Rather under the doctrine of stare decisis, a case is important only for what it decides--for the “what” not for “why” and not for “how.”

Id.

        A judicial precedent attaches to a specific legal consequence to a detailed set of facts in an adjudged case or judicial decision, which is then considered as furnishing the rule for the determination of a subsequent case involving identical or similar material facts and arising in the same court or a lower court in the judicial hierarchy.

Allegheny Gen, Hosp. v. N.L.R.B., 608 F.2d 965, 969-70 (3rd Cir. 1979) (footnote omitted).

        An engaging commentary upon the value of dicta was provided by Justice Neely in his dissent to Pittsburgh Elevator Co. v. West Virginia Board of Regents, 172 W.Va. 743, 310 S.E.2d 675 (1983). Justice Neely surmised that “a dissent to dicta is like the sound of one hand clapping.” “[L]aw must be written with care. It is meant to be an exercise of the mind, not a venting of the spleen.” Id. at 758, 310 S.E.2d at 690 (Neely, J., concurring in part and dissenting in part).

        The references in Bumgarner to the effect of a plaintiff's release of a wrongdoing agent upon a principal were sheer dicta and were confined to a unique set of facts not present in the case sub judice. In its exuberance to provide the victim with recompense and ensure that the moral obligation of the State was satisfied, the majority was overly generous in its comments on the effect of the statute. It is therefore inaccurate to presume that Bumgarner provides the answer to this certified question.

                

II. Analysis of Reasoning of Other Jurisdictions

        Based upon the absence of any definitive authority in West Virginia on this issue, the reasoning and conclusions of other jurisdictions are particularly instructive. As recently recognized by the South Dakota Supreme Court, a split of authority exists in other jurisdictions addressing similar issues regarding a plaintiff's release or covenant not to sue an agent tortfeasor and the effect of such release upon the principal. Estate of Williams ex rel. Williams v. Vandeberg, 620 N.W.2d 187 (S.D. 2000). “The majority of jurisdictions have held that a principal/employer is released from liability when the agent/employee is released via a settlement agreement.” Id. at 189.   (See footnote 13) 

        Where the UCAJTA is implicated, some jurisdictions have applied the language of their unique UCAJTA-based statutes to conclude that a principal is indeed deemed a “joint tortfeasor” with the agent actually engaging in the wrongdoing. In Saranillio v. Silva, 889 P.2d 685 (Haw. 1995), for instance, the precise language of the UCAJTA statute and the statutory definition of joint tortfeasor provided the reviewing court little choice . The statute provided as follows:

        A release by the injured person of one joint tortfeasor, whether before or after judgment, does not discharge the other tortfeasors unless the release so provides; but reduces the claim against the other tortfeasors in the amount of the consideration paid for the release, or in any amount or proportion by which the release provides that the total claim shall be reduced, if greater than the consideration paid.

889 P.2d at 693-94. The Saranillio court observed that the statute was designed to abrogate the common law rule that a release of one joint tortfeasor acted as a release to other joint tortfeasors and recognized the dilemma of application in matters involving principal and agent relationships. Id. at 694. The court reasoned that “[i]t clearly applies to joint tortfeasors as that term traditionally has been used; that is, it applies to 'wrongdoers' who act in concert or concurrently. Traditionally, however, an employer was not considered a joint tortfeasor with his/her employee when the employer's liability was based on respondeat superior.” Id. The Saranillio court struggled with the proper application of the statute, explaining, “On its face, therefore, Section 4 would seem not to apply to vicariously liable parties because they are not ordinarily considered joint tortfeasors.” Id. Upon application of the 1939 version of the UCAJTA, however, the Saranillio court observed that the UCAJTA “does not define joint tortfeasors in the traditional sense.” Id. The statute provides that for purposes of this statute, the “term 'joint tortfeasors' means two or more persons jointly or severally liable in tort for the same injury to person or property, whether or not judgment has been recovered against all or some of them.” Id. (quoting 1939 UCATA § 1, HRS § 663-11 (1985) (emphasis supplied)).

        The UCAJTA definition of joint tortfeasor is based upon liability rather than negligence and is “exceedingly broad and goes beyond the traditional meaning of the term.” Holve v. Draper, 505 P.2d 1265, 1267 (Idaho 1973). Thus, where a reviewing court is limited to the definition of joint tortfeasor provided in the act, as in Saranillio, many courts have held that the definition encompasses the principal liable only vicariously based upon the following reasoning:

        The basis of liability is not relevant, nor is the relationship among those liable for the tort. In short, it makes no difference whether the . . . [employer's] liability is based on the doctrine of respondeat superior or any other legal concept. The point is that both it and the [employee] are (at least) “severally” liable for the same injury to the plaintiff. Therefore, the Uniform Contribution Among Tort-feasors Act applies.

Blackshear v. Clark, 391 A.2d 747, 748 (Del. 1978).   (See footnote 14) 

        Similar reasoning was employed in applying a statute containing the “liable in tort” definition in Wrenn v. Maria Parham Hospital, 522 S.E.2d 789 (N.C. Ct. App. 1999). The statute, codifying the UCAJTA as North Carolina General Statute §§ 1B-1 to 1B- 6, provided that:             When a release or a covenant not to sue ... is given in good faith to one of two or more persons liable in tort for the same injury or the same wrongful death: (1) It does not discharge any of the other tort-feasors from liability for the injury or wrongful death unless its terms so provide[.]

522 S.E.2d at 793 (quoting N.C. Gen.Stat. § 1B-4 (1983) (emphasis supplied)). The Wrenn Court observed:

            Initially, it did not appear that the Uniform Act made any change in the established law of master and servant since the two were not considered to be joint tort-feasors. However, in Yates v. New South Pizza, Ltd., 330 N.C. 790, 412 S.E.2d 666, reh'g denied, 331 N.C. 292, 417 S.E.2d 73 (1992), our Supreme Court held that the term “tort-feasors” as used in the Uniform Act included vicariously liable masters. Thus, the release of a servant did not release a vicariously liable master, unless the terms of the release provided for release of the master. In Yates, the plaintiff was injured in an accident with a pizza deliveryman who was working for New South Pizza, Ltd., d/b/a Domino's Pizza. The plaintiff settled with the driver for $25,000.00, the amount of his insurance coverage, and executed a covenant not to sue the driver or the driver's insurer, but “expressly reserved all rights to proceed against defendant ... employer.” Id. at 791, 412 S.E.2d at 667. In a divided opinion, our Supreme Court held that “for purposes of this Act, a 'tort-feasor' is one who is liable in tort.” Id. at 794, 412 S.E.2d at 669 (emphasis in original).

Id. at 793.

B. The Weight of Well-Reasoned Authority: A Vicariously Liable Entity is Not a Tortfeasor

        Some jurisdictions, as examined above, have confined their evaluation to the strict language of their UCAJTA-based statutes. In the absence of deliberation of the fundamental tort principles and the differences between a joint tortfeasor and a vicariously liable entity, however, a decision regarding this subject is more a matter of linguistics than logic. Where jurisdictions have thoroughly examined the jurisprudential development of the law of tort, specifically the inclusion of the principal/master/employer as a party to be implicated where the impropriety was committed by the agent and the principal is completely free of wrongdoing, the well-reasoned decisions have released the vicariously liable principal upon the release of the agent.

        An excellent illustration of such reasoning is found in Biddle v. Sartori Memorial Hospital, 518 N.W.2d 795 (Iowa 1994), wherein an emergency patient was negligently discharged from the hospital by her physician. Her legal representative released the physician from liability after a settlement had been reached with the physician. The hospital was thereafter sued for the physician's negligence on a vicarious liability claim. The Iowa Supreme Court acknowledged the “fundamental distinction between the full recovery permitted under the doctrine of joint and several liability, and the limitations inherent in a claim that rests on the doctrine of vicarious liability.” Id. at 798. The Biddle court emphasized the need to address “head-on this important distinction.” Id.

         The Biddle court discussed opinions of other jurisdictions holding that settlement with the tortfeasor removes the basis for any additional recovery from the principal upon the same acts of negligence. The Biddle court adopted the theory that the agent and the principal should be treated as one, applying the single share theory for liability to permit a “settlement with an agent [to] effectively adjudicate[] and satisf[y] the vicarious claim.” Id. (citing Glover v. Tacoma Hosp., 658 P.2d 1230, 1238 (Wash. 1983)). An opposite outcome, the Biddle court reasoned, would generate multiplicity of civil actions and a circuity of claims. The court noted that the doctor's settlement would not be protected if a cause of action were permitted to go forward against the hospital. 518 N.W.2d at 799. “        This is because of the well settled rule that a principal found vicariously liable for the negligent acts of an agent retains a right of full indemnity against the actual tortfeasor. . . . The potential for enforcement of such a right. . . would clearly have discouraged rather than encouraged settlement. . . .” Id. at 799 (citations omitted).

        The Biddle court utilized the reasoning of the North Dakota Supreme Court in Horejsi v. Anderson, 353 N.W.2d 316 (N.D. 1984). Horejsi expressed the theory of vicarious liaiblity as follows:

        The “percentage of negligence” attributable to the conduct of the servant constitutes the entire “single share” of liability attributable jointly to the master and servant. . . . Because this percentage of negligence represents the “single share” of liability covered by the common liability of the master and servant, the master is necessarily released from vicarious liability for the released servant's misconduct.

Id. at 318.

        Utilizing this “single share” theory,   (See footnote 15)  the South Dakota Supreme Court in Williams clearly and conclusively stated that “the plaintiff cannot recover against the principal once recovery against the agent has been completed.” 620 N.W.2d at 190. The Williams court reasoned: “The rationales of preventing circuity of action and encouragement of settlement complement one another. The complementary aspects of both rationales serve another important goal: finality. Public policy favors finality, thus avoiding circuity of action that is merely derivative . . . .” Id. “Our holding today fosters the principle of finality while attempting to limit circuity of action and multiplicity of lawsuits, which in this Court's wisdom, is the fairer result.” Id. at 191.

        Similarly, in Copeland v. Humana of Kentucky, Inc., 769 S.W.2d 67 (Ky. Ct. App. 1989), the Kentucky tribunal found that because vicarious liability derives solely from the principal's legal relation to the wrongdoer, settlement, through a covenant not to sue, with the tortfeasor removes the basis for any additional recovery from the principal upon the same acts of negligence. Id. at 70.

            As far as the vicarious liability issue, we find that other courts have spoken to this issue with persuasive reasoning which we paraphrse and adopt. The covenant not to sue not only operated to discharge the anesthesiologists, Schafer and Nash, P.S.C. (the servants/employees) as the parties primarily responsible, it affected a complete discharge of the hospital (the master/employer) who is only secondarily liable, despite the attempted reservation by the Copelands in the covenant of all their rights against the hospital.

Id. The Kentucky court found the the plaintiffs

        had but one cause of action which the law gave to compensate them for their daughter's injuries. This cause of action for the allegedly tortious conduct of Schafer and Nash was assertable against the hospital only because Schafer and Nash were allegedly acting in their function as employees or ostensible agents of the hospital. . . .

Id.. Settlement with the wrongdoers “repaired the wrong.” Id.

        This acquittance inured to the benefit of the hospital, for the discharge of the primary tortfeasor (Schaefer and Nash) must be held to discharge the secondary tortfeasor (the hospital) also from further responsibility, as the hospital's liability for the tortious act was vicarious in nature and derived solely from its legal relation to the wrongdoer, Schafer and Nash.

Id.
    
        A plurality of the Michigan Supreme Court, in an extensive and discerning opinion in Theophelis v. Lansing General Hospital, 424 N.W.2d 478 (Mich. 1988), held that release of the agent is release of the principal even where a plaintiff indicates an express reservation of the right to sue the principal. In Theophelis, representatives of the decedent signed a release absolving the doctor and nurse of liability, expressly reserving the right to sue other entities. The Michigan court evaluated the effect of the following Michigan statute:

            “When a release or a covenant not to sue or not to enforce judgment is given in good faith to 1 of 2 or more persons liable in tort for the same injury or the same wrongful death:
            (a) It does not discharge any of the other tortfeasors from liability for the injury or wrongful death unless its terms so provide.”

Id. at 481 (emphasis supplied in original). The court reasoned that “[t]he principal, having committed no tortious act, is not a 'tortfeasor' as the term is commonly defined.” 424 N.W.2d at 483 (citing Black's Law Dictionary (5^th ed.) (defining a tortfeasor as “a wrong-doer; one who commits or is guilty of a tort.”)). The Michigan court concluded that release of the culpable agent discharges any liability of the principal. Id. at 486. The “single share” theory was adopted, citing the Horejsi rationale, as discussed above. The Michigan court elaborated as follows:

            Put in another context, if A, B and C are sued because each is guilty of negligence which resulted in injury to a plaintiff, their pro-rata shares of the common liability are to be determined under the contribution statute without regard to whether A 's principal, not a wrongdoer, is also joined as a fourth defendant. As between A, an agent, and his principal, there is only one tortfeasor, and they represent only one share of the common liability.

424 N.W.2d at 491.

        In Andrade v. Johnson, 546 S.E.2d 665 (S.C. Ct. App. 2001), the Court of Appeals of South Carolina evaluated the claims of a consumer against a heating contractor and an electric and gas utility company. The consumer had executed a covenant not to sue the contractor, and had specifically reserved the right to proceed against the utility. The issue before the court was whether a covenant not to sue the agent also released the principal, the utility company. The court held that the UCATJA   (See footnote 16)  governs only those situations involving joint tortfeasors and does not apply to an employer who is only derivatively liable. The consumer had urged the court to “expand the definition of tortfeasor under the UCATA to include vicariously liable parties.” 546 S.E.2d at 669. The court explained that the key inquiry is “whether the liability arises only vicariously because of the negligence of another party or whether the parties are true joint tortfeasors, both being independently negligent toward the third party.” Id. The court concluded that the covenant not to sue   (See footnote 17)  terminated both the consumer's claims against the contractor and the utility company's derivative liability.

        Were we to find the covenant released . . . [the agent] but not . . . [the principal], it would necessarily follow that . . . [the principal] could seek indemnification from . . . [the agent] and recover the entire amount of any verdict against it from him. This would effectively strip the covenant not to sue of any real meaning and result in what the court in Nelson v. Gillette described as a “corrosive circle of indemnity.” 571 N.W.2d 332, 339 (N.D. 1997).

Id. at 670. Based upon this reasoning, the South Carolina court concluded that even if it were to expand the definition of tortfeasor as North Carolina did in Yates, “we find the UCATA simply is not applicable to cases involving indemnity.” Id.

        In Anne Arundel Medical Center, Inc. v. Condon, 649 A.2d 1189 (Md. Ct. Spec. App. 1994), a plaintiff released an allegedly negligent pathologist and attempted to sue the medical center on the theory of vicarious liability. The court held that where liability of the medical center was based exclusively upon the negligent conduct of the purported agent, the center and the pathologist were not joint tortfeasors for purposes of the effect of the release under the UCAJTA.   (See footnote 18)  The patient's release of the pathologist was therefore found to function as a release of the medical center as a matter of law. Id. at 1191. Recognizing the split in authority regarding whether a principal and an agent are considered joint tortfeasors, the Maryland court was “persuaded that the better reasoned approach is to hold that Maryland's version of the UCATA does not include vicariously liable defendants and, therefore, that an agent and his principal are not joint-tortfeasors under . . . [the act].” Id. at 1193.

        The Maryland court considered the argument that the statute should be applied to all entities jointly or severally liable in tort, “regardless of the modality of liability.” Id. The court astutely recognized, however, that such holding would require it “to ignore the basic and significant distinctions between vicarious and joint liability.” 649 A.2d at 1193. Reviewing those basis distinctions, the court held that “[i]t is because of their independent wrongdoing [in the case of true joint tortfeasors] that . . . a plaintiff is permitted to bring an action against one joint tortfeasor after having released another joint tortfeasor from liability.” Id. “Each tortfeasor faces liability for his or her own wrongdoing.” Id.

Annotation, Release of (or Covenant not to Sue) Master or Principal as Affecting Liability of Servant or Agent for Tort or Vice Versa, 92 A.L.R.2d 533 (1963). Some jurisdictions have held that release of the agent does not release the principal. See, e.g., Yates v. New South Pizza, Ltd., 412 S.E.2d 666 (N.C. 1992) ; Van Cleave v. Gamboni Constr. Co., 706 P.2d 845 (Nev.1985); Alaska Airlines, Inc. v. Sweat, 568 P.2d 916 (Alaska 1977) . Other courts have determined that the release of the wrongdoer must act as a release of the vicariously liable entity. See, e.g., Mamalis v. Atlas Van Lines, Inc., 560 A.2d 1380 (Pa. 1989); Bristow v. Griffitts Constr. Co., 488 N.E.2d 332 (Ill. App. 1986); Horejsi v. Anderson, 353 N.W.2d 316 (N.D.1984); Craven v. Lawson, 534 S.W.2d 653 (Tenn.1976);

            While some of the cases on this subject deal with covenants not to sue and others with releases, this distinction should not be the determining factor in the end result. The most important factor is the type of liability and the relationship inter se of the various allegedly liable parties rather than the type of document used to discharge liability. It must be determined whether the liability arises only vicariously because of the negligence of another party or whether the parties are true joint tortfeasors, both being independently negligent toward the third party.

546 S.E.2d at 669. Similarly, in McCurry v. School District, 496 N.W.2d 433 (Neb. 1993), the Nebraska court noted the typical distinction between a covenant not to sue and a release but concluded that the distinction was meaningless within the principal/agent context because of the vicarious liability of the principal. The McCurry court sustained the distinction between settlements involving joint tortfeasors and settlements involving vicarious liability and concluded that “it matters not how the settlement was reached; whether by release or covenant not to sue, settlement with the agent constitutes a settlement with the principal, no matter what the parties may have intended.” Id. at 444.

            If a plaintiff, under such a hypothetical legal scheme, were able to find an agent willing to settle, to allow the plaintiff then to proceed additionally against a vicariously liable principal would, in essence, permit the plaintiff “two bites out of the apple.” If the principal could then seek indemnity from the agent, the agent's earlier settlement would be of little solace to him. Such a double exposure would act as a disincentive for agents ever to agree to a settlement.     

649 A.2d at 1196.