OPINION OF THE COURT
Chief Judge Kaye.The issue on this appeal is whether a Missouri statute barring contribution claims against an employer — which conflicts with New York law permitting such claims — should be given effect in a third-party action pending here. Applying relevant choice of law principles, we conclude that the Missouri workers’ compensation statute should be given effect, and therefore affirm the dismissal of the third-party complaint seeking contribution against a Missouri employer.
I.
The facts relevant to this appeal are essentially undisputed. In 1957 or 1958, Kling Brothers, Inc. (succeeded in interest by third-party defendant Hill Acme Co.) manufactured a 16-foot wide "Pyramid Form Bending Roll,” a machine to shape large pieces of metal. The device was sold in 1958 to a Buffalo company, American Standard Inc., through a New York sales agent, defendant Osgood Machinery, Inc., which assisted American in the setup and initial operation of the machine. American closed its Buffalo plant around 1961, and the history of the bending roll is obscured until 1969, when Crouse Company — which obtained the equipment in some unknown *70manner — sold the machine to Paul Mueller Co., a Missouri domiciliary.
Mueller installed the bending roll in its Springfield, Missouri, plant and subsequently modified it by adding a foot switch. In October 1978, plaintiff Dennis J. Cooney, a Missouri resident working at the Missouri plant, was injured while cleaning the machine. The machine was running at the time —a piece of wood having been wedged in the foot switch — and Cooney was unable to reach the switch to stop the machine and avoid injury.
In Missouri, Cooney filed for and received workers’ compensation benefits. Because under Missouri law an employer providing such benefits "shall be released from all other liability * * * whatsoever, whether to the employee or any other person” (Mo Rev Stat § 287.120 [1]), he could not additionally sue his employer, Mueller, in tort. Cooney did, however, bring a products liability action against Osgood — the machine’s initial sales agent — in Supreme Court, Erie County. (Missouri apparently would not have had personal jurisdiction over Osgood.)
Seeking contribution from parties it deems more culpable in the event it is found liable to Cooney, Osgood brought a third-party action against Mueller, American Standard, and Hill Acme. Mueller invoked the Missouri statute shielding employers from both direct claims by employees and contribution claims by others, and moved for summary judgment dismissing Osgood’s third-party complaint. In light of the conflict between the Missouri statute and New York law permitting contribution claims against employers, Supreme Court undertook a choice of law analysis and concluded that New York law should apply. The Appellate Division unanimously reversed and dismissed the third-party complaint as well as all cross claims against Mueller. We now affirm.
II.
An inevitable consequence of a mobile society, where people and goods routinely cross State and national borders, is that disputes may implicate the interests of several jurisdictions having conflicting laws. Choice of law principles become relevant, however, only when a State can, consistent with the Full Faith and Credit and Due Process Clauses of the Constitution (US Const, art IV, § 1; 14th Amend, § 1), choose between the conflicting laws. A State may lack sufficient nexus with a case *71so that choice of its law is arbitrary or fundamentally unfair (Phillips Petroleum Co. v Shultz, 472 US 797, 818; Allstate Ins. Co. v Hague, 449 US 302, 312-313; Schultz v Boy Scouts, 65 NY2d 189, 202-203, n 4). Mueller argues that New York’s connection with the case is so tenuous that a decision to apply New York contribution law would be unconstitutional.
In Hague, the Supreme Court upheld the Minnesota high court’s decision to interpret an automobile insurance policy under Minnesota law instead of under contrary Wisconsin precedent. The policy was issued to a Wisconsin domiciliary, who was a passenger on a motorcycle operated by a Wisconsin resident when he was struck and killed — in Wisconsin — by an automobile driven by another Wisconsin resident. Nevertheless, the Supreme Court plurality found that three contacts with Minnesota, in the aggregate, were sufficient to generate an adequate Minnesota interest in the case: the decedent was employed in Minnesota; his wife, the appointed representative of the estate, subsequently moved to Minnesota; and the insurance company was at all times present and doing business in Minnesota (449 US, at 313-319).
Similarly, New York’s contacts with the present case are, in the aggregate, sufficient to satisfy the constitutional threshold. Osgood has alleged that Mueller has a substantial presence in this State, and there is indication in the record that Mueller does business in New York.1 Additionally, Osgood, which seeks contribution under New York law, is a domiciliary of this State. Finally, Osgood’s alleged tortious conduct with respect to the machine arose in New York, where the machine was ordered, operated for several years, and eventually shipped out of State.
We conclude, therefore, that this State has sufficient interest in the litigation so that if we chose to apply New York law on the contribution issue, that decision would not run afoul of the Federal Constitution. Accordingly, we turn to a choice of law analysis.
III.
The traditional approach to choice of law problems arising in tort was simply to apply lex loci delicti, the law of the place *72of the tort, to all substantive issues in the case (see, e.g., Poplar v Bourjois, 298 NY 62, 66; Restatement of Conflict of Laws §§ 377-390). The theoretical underpinning of that rule was the vested rights doctrine: the right to recover in tort is created by, and exists solely to the extent of, the law of the jurisdiction where the injury occurred (Babcock v Jackson, 12 NY2d 473, 477-478).
Although the vested rights doctrine did have the salutary characteristics of predictability and ease of application, it failed to accord any significance to the policies underlying the conflicting laws of other jurisdictions (see, Babcock, 12 NY2d, at 478; Miller v Miller, 22 NY2d 12, 15). Thus in Babcock the Court adopted a more flexible approach intended to give "controlling effect to the law of the jurisdiction which, because of its relationship or contact with the occurrence or the parties, has the greatest concern with the specific issue raised in the litigation.” (Babcock, 12 NY2d, at 481.)
Of the various, sometimes competing, schools of thought on choice of law, the one that emerged as most satisfactory was "interest analysis,” which sought to effect the law of the jurisdiction having the greatest interest in resolving the particular issue (see, Schultz v Boys Scouts, 65 NY2d 189, 197, supra; Miller v Miller, 22 NY2d 12, 15-16, supra). An immediate distinction was drawn between laws that regulate primary conduct (such as standards of care) and those that allocate losses after the tort occurs (such as vicarious liability rules). If conflicting conduct-regulating laws are at issue, the law of the jurisdiction where the tort occurred will generally apply because that jurisdiction has the greatest interest in regulating behavior within its borders. But if competing "postevent remedial rules” are at stake other factors are taken into consideration, chiefly the parties’ domiciles (see, Schultz v Boy Scouts, 65 NY2d, at 197-199; Babcock, 12 NY2d, at 483).
Babcock itself involved a loss distribution rule — an Ontario guest statute limiting an automobile passenger’s ability to recover from the driver — as did subsequent cases (Dym v Gordon, 16 NY2d 120 [Colorado guest statute]; Long v Pan Am. World Airways, 16 NY2d 337 [conflicting Maryland and Pennsylvania laws governing wrongful death recoveries]; Macey v Rozbicki, 18 NY2d 289 [Ontario guest statute]; Farber v Smolack, 20 NY2d 198 [vicarious liability]; Miller v Miller, 22 NY2d 12 [Maine statute limiting wrongful death damages to $20,000]; Looker v Lopez, 24 NY2d 569 [Michigan guest statute]).
*73In Neumeier v Kuehner (31 NY2d 121), yet another guest statute case, the Court in seeking to return greater predictability and uniformity to the law, adopted a series of three rules that had been proposed by Chief Judge Fuld (see, Tooker v Lopez, 24 NY2d, at 585 [Fuld, Ch. J., concurring]). Although drafted in terms of guest statutes — drivers and passengers— these rules could, in appropriate cases, apply as well to other loss allocation conflicts (see, Schultz, 65 NY2d, at 200-201 [applying first and third Neumeier rules to conflicting charitable immunity laws]).
The Neumeier Rules
Under the first Neumeier rule, when the driver-host and passenger-guest share a common domicile, that law should control. Indeed, when both parties are from the same jurisdiction, there is often little reason to apply another jurisdiction’s loss allocation rules. The domiciliary jurisdiction, which has weighed the competing considerations underlying the loss allocation rule at issue, has the greater "interest in enforcing the decisions of both parties to accept both the benefits and the burdens of identifying with that jurisdiction and to submit themselves to its authority” (Schultz, 65 NY2d, at 198). Moreover, this rule reduces opportunities for forum shopping because the same law will apply whether the suit is brought in the locus jurisdiction or in the common domicile, the two most likely forums (Schultz, 65 NY2d, at 199-201).
The second Neumeier rule addresses "true” conflicts, where the parties are domiciled in different States and the local law favors the respective domiciliary. When plaintiff’s State, for example, would allocate the loss to defendant but defendant’s State would force plaintiff to bear the loss, a true conflict arises. The rule provides that when the driver’s (defendant’s) conduct occurred in the State of domicile and that State would not impose liability, the driver should not be exposed to liability under the law of the victim’s domicile. Conversely, when the plaintiff-passenger is injured in the place of domicile and would be entitled to recover, the out-of-State driver should generally be unable to interpose the law of his or her domicile to defeat recovery (31 NY2d, at 128). In essence, then, the second Neumeier rule adopts a "place of injury” test for true conflict guest statute cases.
Finally, the third Neumeier rule, applicable to other split-*74domicile cases, provides that the usually governing law will be that of the place where the accident occurred, unless " 'displacing that normally applicable rule will advance the relevant substantive law purposes without impairing the smooth working of the multistate system or producing great uncertainty for litigants’ ” (31 NY2d, at 128). This rule, too, generally uses the place of injury, or locus, as the determining factor.
Assuming that the interest of each State in enforcement of its law is roughly equal — a judgment that, insofar as guest statutes are concerned, is implicit in the second and third Neumeier rules — the situs of the tort is appropriate as a "tie breaker” because that is the only State with which both parties have purposefully associated themselves in a significant way (see, Korn, The Choice-of-Law Revolution: A Critique, 83 Colum L Rev 772, 801 [1983]). Moreover, locus is a neutral factor, rebutting an inference that the forum State is merely protecting its own domiciliary or favoring its own law (see, Schultz, 65 NY2d, at 201). Additionally, the place of injury was the traditional choice of law crucible.
Schultz applied both the first and third Neumeier rules. The plaintiffs in that case, residents of New Jersey, sued New Jersey and Ohio domiciliaries in New York for injuries occurring here. Defendants urged that New Jersey charitable immunity law was controlling; plaintiffs argued for New York law, which does not recognize charitable immunity. As between plaintiffs and the New Jersey defendant — codomiciliaries — the Court looked to the first Neumeier rule and, after analysis, applied the law of the shared domicile, New Jersey. The third Neumeier rule, however, applied to the plaintiffs’ claim against the Ohio domiciliary, because the parties had different domiciles and the injuries occurred in a separate jurisdiction. Although the third rule generally invokes the locus jurisdiction’s law, the Court applied the proviso permitting resort to other law when it would advance substantive law purposes without impairing the multistate system or producing great uncertainty. We thus applied New Jersey law to that claim as well (Schultz, 65 NY2d, at 200-202).
Contribution rules — as involved in the present case — are loss allocating, not conduct regulating. Had conduct regulating been at issue here, our analysis would be greatly simplified, for the traditional rule of lex loci delicti almost invariably obtains. Similarly, if the parties shared the same domi*75cile, we would generally apply that jurisdiction’s loss distribution law. Instead, our analysis is necessarily more complicated, calling upon us to evaluate the relative interests of jurisdictions with conflicting laws and, if neither can be accommodated without substantially impairing the other, finding some other sound basis for resolving the impasse.
Interest Analysis
The general scheme of workers’ compensation acts is that an employer regardless of culpability is required to make specified payments to an injured employee and in exchange, the law immunizes the employer from further liability. Immunity "is part of the quid pro quo in which the sacrifices and gains of employees and employers are to some extent put in balance, for, while the employer assumes a new liability without fault, [it] is relieved of the prospect of large damage verdicts” (2A Larsen, Workmen’s Compensation Law § 65.11 [1993]).
Some States immunize employers only from direct actions by injured workers; others extend protection from third-party contribution actions as well. The Missouri Supreme Court, in rejecting State and Federal constitutional challenges to the Missouri statute at issue here, noted that immunity " 'is the heart and soul of this legislation which has, over the years been of highly significant social and economic benefit to the working [person], the employer and the State.’ ” (State ex rel. Maryland Hgts. Concrete Contrs. v Ferriss, 588 SW2d 489, 491 [Mo], quoting Seabord Coast Line R. R. Co. v Smith, 359 So 2d 427, 429 [Fla].) The court, quoting further from the Florida case, also observed that " ' "the right to contribution is not a vested right on which legislation may not impinge” ’ ” (588 SW2d, at 491).
Missouri’s decision to shield employers from contribution claims is thus a policy choice implicating significant State interests: "to deny a person the immunity granted * * * by a work[er]’s compensation statute of a given state would frustrate the efforts of that state to restrict the cost of industrial accidents and to afford a fair basis for predicting what these costs will be.” (Restatement [Second] of Conflict of Laws § 184, comment b, at 547.) Indeed, as the Restatement concluded in a related context, for another State "to subject a person who has been held liable in work[er]’s compensation to further unlimited liability in tort or wrongful death would frustrate *76the work[er]’s compensation policy of the State in which the award was rendered.” (Restatement [Second] of Conflict of Laws § 183, comment c, at 544.)
Arrayed against Missouri’s interest in maintaining the integrity of its workers’ compensation scheme is New York’s interest in basic fairness to litigants. Under traditional joint and several liability rules, when more than one tortfeasor was responsible for plaintiffs injury, each was potentially liable for the entire judgment, irrespective of relative culpability. Indeed, plaintiff was not even required to sue all the wrongdoers, but could recover the entire judgment from the "deep pocket,” who then had no recourse (Sommer v Federal Signal Corp., 79 NY2d 540, 556).
In Dole v Dow Chem. Co. (30 NY2d 143, 148-149 [1972]), this Court mitigated the inequity by allowing a defendant that pays more than its fair share of a judgment, as apportioned by the fact finder in terms of relative fault, to recover the difference from a codefendant. The Legislature, also recognizing the desirability of contribution, subsequently codified the Dole principles in CPLR article 14 (L 1974, ch 742). Stated simply, the "goal of contribution, as announced in Dole and applied since, is fairness to tortfeasors who are jointly liable.” (Sommer v Federal Signal Corp., 79 NY2d, at 556-557.)
Manifestly, the interests of Missouri and New York are irreconcilable in this case. To the extent we allow contribution against Mueller, the policy underlying the Missouri workers’ compensation scheme will be offended. Conversely, to the extent Osgood is required to pay more than its equitable share of a judgment, the policy underlying New York’s contribution law is affronted. It is evident that one State’s interest cannot be accommodated without sacrificing the other’s, and thus an appropriate method for choosing between the two must be found.
This is a true conflict in the mold of Neumeier’s second rule, where the local law of each litigant’s domicile favors that party, and the action is pending in one of those jurisdictions. Under that rule, the place of injury governs, which in this case means that contribution is barred. This holding is consistent with the result reached historically, and reflects application of a neutral factor that favors neither the forum’s law nor its domiciliaries. Moreover, forum shopping by defendants —who might attempt to invoke CPLR 1403 and bring a separate action for contribution in New York if sued else*77where (compare, Grant Co. v Uneeda Doll Co., 19 AD2d 361, affd 15 NY2d 571) — is eliminated.2
A primary reason that locus tips the balance, of course, is that ordinarily it is the place with which both parties have voluntarily associated themselves. In this case, there is some validity to Osgood’s argument that it did nothing to affiliate itself with Missouri. Indeed, a decade after Osgood’s last contact with the bending roll, the machine wound up in Missouri through no effort, or even knowledge, of Osgood. Moreover, the record establishes that Osgood was not in the business of distributing goods nationwide, but limited its activities to New York and parts of Pennsylvania, and thus Osgood may not have reasonably anticipated becoming embroiled in litigation with a Missouri employer.
For this reason, our decision to apply Missouri law rests as well on another factor that should, at times, play a role in choice of law: the protection of reasonable expectations (see, Restatement [Second] of Conflict of Laws § 6 [2] [d]; Allstate Ins. Co. v Hague, 449 US 302, 327 [Stevens, J., concurring]; Schultz, 65 NY2d, at 198 ["protecting the reasonable expectations of the parties” is one reason locus law is generally preferred when there are conflicting conduct-regulating rules]).3 In view of the unambiguous statutory language barring third-party liability and the Missouri Supreme Court’s holding in Ferriss, Mueller could hardly have expected to be haled before a New York court to respond in damages for an accident to a Missouri employee at the Missouri plant. By contrast, in ordering its business affairs Osgood could have had no reasonable expectation that contribution would be available in a products liability action arising out of the sale of industrial equipment. Indeed, Osgood’s activity in connection with the bending roll occurred in 1958, some 14 years before Dole was decided and the principles of full contribution were introduced into our law. Moreover, even under present *78law, contribution is not foolproof. A defendant, for example, may be unable to obtain jurisdiction over a joint tortfeasor; the joint tortfeasor may be insolvent or defunct (like Kling Bros. here); or defendant’s own assets may be insufficient to pay its share of the judgment (see, Klinger v Dudley, 41 NY2d 362, 369).
In sum, we conclude that Missouri law should apply because, although the interests of the respective jurisdictions are irreconcilable, the accident occurred in Missouri, and unavailability of contribution would more closely comport with the reasonable expectations of both parties in conducting their business affairs.
IV.
Finally, we turn to Osgood’s contention that New York’s public policy precludes application of the Missouri statute in this case. Under the public policy exception, when otherwise applicable foreign law would "violate some fundamental principle of justice, some prevalent conception of good morals, some deep-rooted tradition of the common weal” (Loucks v Standard Oil Co., 224 NY 99, 111 [Cardozo, J.]), the court may refuse to enforce it.
As Judge Simons noted in Schultz, the public policy exception should be considered only after the court has first determined, under choice of law principles, that the applicable substantive law is not the forum’s law (Schultz, 65 NY2d, at 202). We have already cleared that hurdle in the present case. Moreover, the exception could apply only when New York’s nexus with the case is substantial enough to threaten our public policy (id., at 202-203). While we have yet to explore whether there is a difference between the minimum contacts needed to satisfy the constitutional choice-of-law threshold and those required to implicate the public policy exception, it is unnecessary to do so today; the facts of this case satisfy both standards. Unlike Schultz, where New York’s nexus was insufficient to implicate our public policy, this case involves a New York domiciliary who may be cast in liability by the effect of foreign law which is contrary to New York law. Accordingly, New York’s interest in this case is sufficient to warrant scrutiny under the public policy exception.
Although we have noted that public policy may be found in the State Constitution, statutes and judicial decisions (Shannon v Irving Trust Co., 275 NY 95, 102-103; Schultz, 65 NY2d, *79at 202), plainly not every difference between foreign and New York law threatens our public policy. Indeed, if New York statutes or court opinions were routinely read to express fundamental policy, choice of law principles would be meaningless. Courts invariably would be forced to prefer New York law over conflicting foreign law on public policy grounds.
The refusal of courts to enforce foreign law as repugnant to public policy reached its zenith prior to the advent of modern choice of law doctrine. In fact, commentators have opined that in earlier times the public policy rationale really substituted as a choice of law mechanism when the prevailing rigid choice of law rules permitted no flexibility (see, Paulsen and Sovern, "Public Policy ” in the Conflict of Laws, 56 Colum L Rev 969, 981 [1956]; see also, Feldman v Acapulco Princess Hotel, 137 Misc 2d 878, 888, n 14). That theory is supported by some of our cases.
In Mertz v Mertz (271 NY 466), for instance, a wife sued her husband in New York for injuries arising out of his alleged negligent operation of a car in Connecticut. Both parties were New York domiciliaries, and under New York law at that time, a wife could not sue her husband for negligence. Connecticut law, however, did not recognize spousal immunity and under lex loci delicti, Connecticut law would ordinarily have applied. Based in part on public policy considerations this Court affirmed dismissal of the complaint. The same result would likely obtain today, however, under choice of law principles. In view of modern choice of law doctrine, resort to the public policy exception should be reserved for those foreign laws that are truly obnoxious.
The thrust of Osgood’s argument is that New York’s law permitting contribution is so strong that any encroachment upon the right violates fundamental public policy. In our choice of law analysis, of course, we explicitly considered New York’s interest in allowing contribution and concluded that it is significant. Osgood’s view, however, is that no abrogation of the right may be tolerated. We disagree.
Certainly, contribution is not a deeply rooted tradition of the common weal (Loucks, 224 NY, at 111), having been introduced into our law only relatively recently. Moreover, as noted,. availability of contribution is not invariably guaranteed. And while Osgood claims that being forced to pay more than its equitable share of plaintiff’s damages is unfair, "public policy is not measured by individual notions of expediency *80and fairness or by a showing that the foreign law is unreasonable or unwise” (Schultz, 65 NY2d, at 202). In the considered judgment of the Missouri Legislature, employers providing workers’ compensation benefits are not amenable to claims for contribution. New York law is to the contrary. But as Judge Cardozo observed: "Our own scheme of legislation may be different. * * * That is not enough to show that public policy forbids us to enforce the foreign right. * * * We are not so provincial as to say that every solution of a problem is wrong because we deal with it otherwise at home.” (Loucks v Standard Oil Co., 224 NY, at 110-111.) Osgood has not sustained its "heavy burden” of proving that the Missouri statute is offensive to our public policy (Schultz, 65 NY2d, at 202).
Accordingly, the order of the Appellate Division should be affirmed, with costs.
Judges Simons, Titone, Hancock, Jr., Bellacosa and Smith concur.
Order affirmed, with costs.
. Were Mueller not subject to jurisdiction in New York, it could of course successfully move to dismiss for lack of personal jurisdiction.
. New York law permitting contribution against an employer is clearly a minority view (see generally, Annotation, Modern Status of Effect of State Workmen’s Compensation Act on Right of Third-Person Tortfeasor to Contribution or Indemnity From Employer of Injured or Killed Workman, 100 ALRSd 350). A result that might impose New York law on the carefully structured workers’ compensation schemes of other States — especially when the accident occurred there — is undesirable.
. We have eschewed reliance on the fictional expectation of the parties based on mere contact with the locus of an accident (Miller v Miller, 22 NY2d, at 20), but reasonable, justifiable expectations are another matter.