with whom The Chief Justice and Justice Rehnquist join, dissenting.
My disagreement with the plurality is narrow. I accept with few reservations Part II of the plurality opinion, which sets forth the basic principles that guide us in reviewing state choice-of-law decisions under the Constitution. The Court should invalidate a forum State’s decision to apply its own law only when there are no significant contacts between the State and the litigation. This modest check on state power is mandated by the Due Process Clause of the Fourteenth Amendment and the Full Faith and Credit Clause of Art. IV, § 1. I do not believe, however, that the plurality adequately analyzes the policies such review must serve. In consequence, it has found significant what appear to me to be trivial contacts between the forum State and the litigation.
*333I
At least since Carroll v. Lanza, 349 U. S. 408 (1955), the Court has recognized that both the Due Process and the Full Faith and Credit Clauses are satisfied if the forum has such significant contacts with the litigation that it has a legitimate state interest in applying its own law. The significance of asserted contacts must be evaluated in light of the constitutional policies that oversight by this Court should serve. Two enduring policies emerge from our cases.
First, the contacts between the forum State and the litigation should not be so “slight and casual” that it would be fundamentally unfair to a litigant for the forum to apply its own State’s law. Clay v. Sun Ins. Office, Ltd., 377 U. S. 179, 182 (1964). The touchstone here is the reasonable expectation of the parties. See Weintraub, Due Process and Full Faith and Credit Limitations on a State’s Choice of Law, 44 Iowa L. Rev. 449, 445-457 (1959) (Weintraub). Thus, in Clay, the insurer sold a policy to Clay “ ‘with knowledge that he could take his property anywhere in the world he saw fit without losing the protection of his insurance.’ ” 377 U. S., at 182, quoting Clay v. Sun Ins. Office Ltd., 363 U. S. 207, 221 (1960) (Black, J., dissenting). When the insured moved to Florida with the knowledge of the insurer, and a loss occurred in that State, this Court found no unfairness in Florida’s applying its own rule of decision to permit recovery on the policy. The insurer “must have known it might be sued there.” Ibid. See also Watson v. Employers Liability Assurance Corp., 348 U. S. 66 (1954).1
*334Second, the forum State must have a legitimate interest in the outcome of the litigation before it. Pacific Ins. Co. v. Industrial Accident Comm’n, 306 U. S. 493 (1939). The Full Faith and Credit Clause addresses the accommodation of sovereign power among the various States. Under limited circumstances, it requires one State to give effect to the statutory law of another State. Nevada v. Hall, 440 U. S. 410, 423 (1979). To be sure, a forum State need not give effect to another State’s law if that law is in “violation of its own legitimate public policy.” Id., at 422. Nonetheless, for a forum State to further its legitimate public policy by applying its own law to a controversy, there must be some connection between the facts giving rise to the litigation and the scope of the State’s lawmaking jurisdiction.
Both the Due Process and Full Faith and Credit Clauses ensure that the States do not “reach out beyond the limits imposed on them by their status as coequal sovereigns in a federal system.” World-Wide Volkswagen Corp. v. Woodson, 444 U. S. 286, 292 (1980) (addressing Fourteenth Amendment limitation on state-court jurisdiction). As the Court stated in Pacific Ins. Co., supra: “[T]he full faith and credit clause does not require one state to substitute for its own statute, applicable to persons and events within it, the conflicting statute of another state.” Id., at 502 (emphasis added). The State has a legitimate interest in applying a rule of decision to the litigation only if the facts to which the rule will be applied have created effects within the State, toward which the State’s public policy is directed. To assess the sufficiency of asserted contacts between the forum and the litigation, the court must determine if the contacts form a reasonable link between the litigation and a state policy. In short, examination of contacts addresses whether “the state *335has an interest in the application of its policy in this instance.” Currie, The Constitution and the Choice of Law: Governmental Interests and the Judicial Function, in B. Currie, Selected Essays on the Conflict of Laws 188, 189 (1963) (Currie). If it does, the Constitution is satisfied.
John Hancock Mut. Life Ins. Co. v. Yates, 299 U. S. 178 (1936), illustrates this principle. A life insurance policy was executed in New York, on a New York insured with a New York beneficiary. The insured died in New York; his beneficiary moved to Georgia and sued to recover on the policy. The insurance company defended on the ground that the insured, in the application for the policy, had made materially false statements that rendered it void under New York law. This Court reversed the Georgia court’s application of its contrary rule that all questions of the policy’s validity must be determined by the jury. The Court found a violation of the Full Faith and Credit Clause, because “[i]n respect to the accrual of the right asserted under the contract . . . there was no occurrence, nothing done, to which the law of Georgia could apply.” Id., at 182. In other words, the Court determined that Georgia had no legitimate interest in applying its own law to the legal issue of liability. Georgia’s contacts with the contract of insurance were nonexistent.2 See Home Ins. Co. v. Dick, 281 U. S. 397, 408 (1930).
In summary, the significance of the contacts between a forum State and the litigation must be assessed in light of *336these two important constitutional policies.3 A contact, or a pattern of contacts, satisfies the Constitution when it protects the litigants from being unfairly surprised if the forum State applies its own law, and when the application of the forum’s law reasonably can be understood to further a legitimate public policy of the forum State.
II
Recognition of the complexity of the constitutional inquiry requires that this Court apply these principles with restraint. Applying these principles to the facts of this case, I do not believe, however, that Minnesota had sufficient contacts with the “persons and events” in this litigation to apply its rule permitting stacking. I would agree that no reasonable expectations of the parties were frustrated. The risk insured by petitioner was not geographically limited. See Clay v. Sun Ins. Office, Ltd., 377 U. S., at 182. The close proximity of Hager City, Wis., to Minnesota, and the fact that Hague commuted daily to Red Wing, Minn., for many years should have led the insurer to realize that there was a reasonable probability that the risk would materialize in Minnesota. Under our precedents, it is plain that Minnesota could have applied its own law to an accident occurring within its borders. See ante, at 318, n. 24. The fact that the accident did not, in fact, occur in Minnesota is not controlling because the expectations of the litigants before the cause of *337action accrues provide the pertinent perspective. See Weintraub 455; n. 1, supra.
The more doubtful question in this case is whether application of Minnesota’s substantive law reasonably furthers a legitimate state interest. The plurality attempts to give substance to the tenuous contacts between Minnesota and this litigation. Upon examination, however, these contacts are either trivial or irrelevant to the furthering of any public policy of Minnesota.
First, the postaccident residence of the plaintiff-beneficiary is constitutionally irrelevant to the choice-of-law question. John Hancock Mut. Life Ins. Co. v. Yates, supra. The plurality today insists that Yates only held that a postoccurrence move to the forum State could not “in and of itself” confer power on the forum to apply its own law, but did not establish that such a change of residence was irrelevant. Ante, at 319. What the Yates Court held, however, was that “there was no occurrence, nothing done, to which the law of Georgia could apply.” 299 U. S., at 182 (emphasis added). Any possible ambiguity in the Court’s view of the significance of a postoccurrence change of residence is dispelled by Home Ins. Co. v. Dick, supra, cited by the Yates Court, where it was held squarely that Dick’s postaccident move to the forum State was “without significance.” 281 U. S., at 408.
This rule is sound. If a plaintiff could choose the substantive rules to be applied to an action by moving to a hospitable forum, the invitation to forum shopping would be irresistible. Moreover, it would permit the defendant’s reasonable expectations at the time the cause of action accrues to be frustrated, because it would permit the choice-of-law question to turn on a postaccrual circumstance. Finally, postaccrual residence has nothing to do with facts to which the forum State proposes to apply its rule; it is unrelated to the substantive legal issues presented by the litigation.
Second, the plurality finds it significant that the insurer does business in the forum State. Ante, at 317-318. The State *338does have a legitimate interest in regulating the practices of such an insurer. But this argument proves too much. The insurer here does business in all 50 States. The forum State has no interest in regulating that conduct of the insurer unrelated to property, persons, or contracts executed within the forum State.4 See Hoopeston Canning Co. v. Cullen, 318 U. S. 313, 319 (1943). The plurality recognizes this flaw and attempts to bolster the significance of the local presence of the insurer by combining it with the other factors deemed significant: the presence of the plaintiff and the fact that the deceased worked in the forum State. This merely restates the basic question in the case.
Third, the plurality emphasizes particularly that the insured worked in the forum State.5 Ante, at 313-317. The fact that the insured was a nonresident employee in the forum *339State provides a significant contact for the furtherance of some local policies. See, e. g., Pacific Ins. Co. v. Industrial Accident Comm’n, 306 U. S. 493 (1939) (forum State’s interest in compensating workers for employment-related injuries occurring within the State); Alaska Packers Assn. v. Industrial Accident Comm’n, 294 U. S. 532, 549 (1935) (forum State’s interest in compensating the employment-related injuries of a worker hired in the State). The insured’s place of employment is not, however, significant in this case. Neither the nature of the insurance policy, the events related to the accident, nor the immediate question of stacking coverage is in any way affected or implicated by the insured’s employment status. The plurality’s opinion is understandably vague in explaining how trebling the benefits to be paid to the estate of a nonresident employee furthers any substantial state interest relating to employment. Minnesota does not wish its workers to die in automobile accidents, but permitting stacking will not further this interest. The substantive issue here is solely one of compensation, and whether the compensation provided by this policy is increased or not will have no relation to the State’s employment policies or police power. See n. 5, supra.
Neither taken separately nor in the aggregate do the contacts asserted by the plurality today indicate that Minnesota’s application of its substantive rule in this case will further any legitimate state interest.6 The plurality focuses *340only on physical contacts vel non, and in doing so pays scant attention to the more fundamental reasons why our precedents require reasonable policy-related contacts in choice-of-law cases. Therefore, I dissent.
Home Ins. Co. v. Dick, 281 U. S. 397 (1930), is a case where the reasonable expectations of a litigant were frustrated. The insurance contract confined the risk to Mexico, where the loss occurred and where both the insurer and the insured resided until the claim accrued. This Court found a violation of the Due Process Clause when Texas, the forum State, applied a local rule to allow the insured to gain a recovery unavailable under Mexican law. Because of the geographic limitation on the risk, and *334because there were no contacts with the forum State until the claim accrued, the insurer could have had no reasonable expectation that Texas law would be applied to interpret its obligations under the contract. See Weintraub 455.
“It is manifest that Georgia had no interest in the application to this case of any policy to be found in its laws. When the contract was entered into, and at all times until the insured died, the parties and the transaction were beyond the legitimate reach of whatever policy Georgia may have had. Any interest asserted by Georgia must relate to the circumstance that the action is tried there, and must arise not from any policy directed to the business of life insurance but from some policy having to do with the business of the courts. This was apparently recognized even by the Georgia court; hence the disingenuous characterization of the matter as one of ‘procedure’ rather than of ‘substance.’” Currie 236. See also id., at 232-233.
The plurality today apparently recognizes that the significance of the contacte must be evaluated in light of the policies our review serves. It acknowledges that the sufficiency of the same contacts sometimes will differ in jurisdiction and choice-of-law questions. Ante, at 317, n. 23. The plurality, however, pursues the rationale for the requirement of sufficient contacts in choice-of-law cases no further than to observe that the forum’s application of its own law must be “neither arbitrary nor fundamentally unfair.” Ante, at 313. But this general prohibition does not distinguish questions of choice of law from those of jurisdiction, or from much of the jurisprudence of the Fourteenth Amendment.
The petitioner in John Hancock Mut. Life Ins. Co. v. Yates, 299 U. S. 178 (1936), did business in Georgia, the forum State, at the time of that case. See The Insurance Almanac 715 (1935). Also, Georgia extensively regulated insurance practices within the State at that time. See Ga. Code §56-101 et seq. (1933). This Court did not hint in Yates that this fact was of the slightest significance to the choice-of-law question, although it would have been crucial for the exercise of in personam jurisdiction.
The plurality exacts double service from this fact, by finding a separate contact in that the insured commuted daily to his job. Ante, at 314-315. This is merely a repetition of the facts that the insured lived in Wisconsin and worked in Minnesota. The State does have an interest in the safety of motorists who use its roads. This interest is not' limited to employees, but extends to all nonresident motorists on its highways. This safety interest, however, cannot encompass, either in logic or in any practical sense, the determination whether a nonresident's estate can stack benefit coverage in a policy written in another State regarding an accident that occurred on another State’s roads.
Cardillo v. Liberty Mutual Ins. Co., 330 U. S. 469 (1947), hardly establishes commutation as an independent contact; the case merely approved the application of a forum State's law to an industrial accident occurring in a neighboring State when the employer and the employee both resided in the forum State.
The opinion of Justice SteveNs concurring in the judgment supports my view that the forum State’s application of its own law to this case cannot be justified by the existence of relevant minimum contacts. As Justice SteveNs observes, the principal factors relied on by the plurality are “either irrelevant to or possibly even tend to undermine the [plurality’s] conclusion.” Ante, at 331. The interesting analysis he proposes to uphold the State’s judgment is, however, difficult to reconcile with our prior decisions and may create more problems than it solves. For example, it seems questionable to measure the interest of a State in a controversy by the degree of conscious reliance on that State’s law by private *340parties to a contract. Ante, at 324. Moreover, scrutinizing the strength of the interests of a nonforum State may draw this Court back into the discredited practice of weighing the relative interests of various States in a particular controversy. See ante, at 308, n. 10 (plurality opinion).