Involved in this appeal is a difficult choice-of-law problem arising from a death action brought in this State in connection with an out-of-State automobile accident. The issue arises on the granting of plaintiff’s motion to dismiss the third affirmative partial defense interposing a statutory limit on recoveries in death actions of $20,000, plus medical and funeral expenses, applicable at the time of the death in the State of Maine where both the accident and death occurred. Defendants have appealed.
Because it is concluded that the Maine rule limiting recoveries in death actions should be applied, the motion ought to be denied. Accepting as more or less valid the modern but only emerging analyses for choosing the applicable or proper rules of law to achieve a just result in this case and cases of like kind, the significant contacts, the interests of the concerned jurisdictions, the minimal requirements for uniformity of result, the relevance of what a neutral forum would select as the applicable rules of law, and the expectations of the parties to the accident suggest that the application of the Maine rule is the most reasonable choice. This is not to say that the issue is not rife with difficulties, because it is. The unsettled and often conflicting theories with regard to choice of law come to a focus in this case, exemplifying the anomalies that may be produced, on the one hand, by adhesion to one theory, or, on the other hand, by a determination of particular cases on an ad hoc basis.
The Miller brothers were engaged in business together with their father in New York, and in Maine since 1958. Defendant husband with his family left New York and settled on a permanent basis in Maine in the spring of 1960, in connection with the operation and management of a ladies’ ready-to-wear department (in a larger store) in Brunswick, Maine, an enterprise in which the Millers and their father were interested. Decedent and his family remained residents of New York State throughout the time involved, and his survivors are still such residents.
*24From time to time, decedent visited Maine and stayed at Ms brother’s home in Maine, for days at a time, for the business purposes of the Brunswick store, working in it and consulting with his brother concerning its operation. In August, 1961, while on such a visit and riding as a passenger in a station wagon driven by defendant husband, decedent sustained the injuries resulting in his death on the same day. The accident occurred when the automobile was driven into a bridge railing. The station wagon was registered in Maine in the name of codefendant wife, with whose permission the car was being used. It was garaged and insured in Maine. At the time of the accident defendant husband was still using a New York State operator’s license. The first leg of the fatal trip started in Maine at defendants’ home in Brunswick and was to terminate in Calais, Maine, where the brothers were to look into the possibilities for a new family enterprise in that city, and then return to Brunswick.
After the accident, defendants, in November, 1961, moved back to New York and resumed their residence in Westchester County where they still reside. This action was begun in June, 1962. Three years later, the Maine Legislature repealed its limitations on recoveries in death actions (L. 1965, ch. 255).
At the outset it should be observed that if the significant contacts are considered under the analysis in the leading case of Babcock v. Jackson (12 N Y 2d 473) it is the Maine rule that should be applied (Restatement, Conflicts, 2d, Proposed Official Draft, Part I, May 2, 1967, § 6; Part II, May 1, 1968, § 145). The accident happened in Maine. The automobile was registered, garaged, and insured in Maine. Defendants were residents of Maine. The projected trip was planned and was to take place wholly in Maine. The trip was in connection with Maine business. Decedent’s stay in Maine was not transient but was one of several recurring sojourns in connection with a business in Maine. In cases like Babcock v. Jackson (supra), Macey v. Rozbicki (18 N Y 2d 289), and Farber v. Smolack (20 N Y 2d 198), the first two involving an out-of-State statute governing guest status and the third that of owner liability for a driver’s conduct, the transient aspect of the accident, the location being little more than a happenstance, was denigrated as a significant contact. Thus, in the Farber case, the court said:
*25“ The transitory use of the car does not necessarily impose the law of State of transit on this relationship and, in this respect, the local law of the State of transit as to permissive use of the motor vehicle by its owner is not essentially different in its New York consequence from the law of the State of transit relating to liability to a gratuitous guest considered in Babcock.
“In addressing ourselves to the policy of treating this sort 6f transitory tort arising entirely from New York relationships as governed by New York law, there is no logical basis to distinguish the application to out-of-State accidents of the New York law of liability to gratuitous guests and the New York law of liability arising from permissive use of a vehicle ” (pp. 203-204).
In Dym v. Gordon (16 N Y 2d 120) this court, applying a grouping of contacts analysis, held that a Colorado planned and executed trip by New York residents in a New York car owned by one of them required the application of the Colorado guest statute. Of course, in the Macey case (supra), the court, again looking to contacts analysis, but with somewhat greater emphasis on the significance of selected contacts, held that New York law, rather than an Ontario guest statute, would apply to New York residents, a New York car, and a wholly Canadian trip, ‘ ‘ especially since the arrangements for that visit had undoubtedly been made in New York State ” (p. 292). The Dym case (supra) was distinguished on the ground that the New York residents had gone separately to Colorado and there adventitiously planned the Colorado trip.
There is a competing, but not wholly inconsistent, emerging theory in conflicts of laws. It goes variously by the name of interest analysis, that is, by the analysis of interests of concerned jurisdictions, or the search for the jurisdiction with the greatest concern in the transaction (see, e.g., Cavers, The Choice-of-Law Process; Currie, Selected Essays on the Conflict of Laws; cf. von Mehren and Trautman, The Law of Multistate Problems).
Even on an interest analysis, leaving aside for a moment the post-accident events, it may be argued that the Maine limitation should be applied. The case presents a ‘ ‘ true conflict ’ ’ in that both Maine and New York have an interest in applying their rules regarding damage limitations to the transaction in question. Hence there is room for application of a principle of preference *26to select between the Maine and New York rules (see Cavers, op. cit., pp. 114-138). Thus, in such analysis it has been suggested that the ‘ ‘ lower standard of conduct or of financial protection of the State where the defendant acted and injury resulted should be applied in the absence of a previously existing relationship between the parties (id., p. 146). Here .unlike the guest statute cases there is no need to look to the seat of the relationship, or domicile of guest and host, in order to assure that the legal incidents of a transitory relationship, persisting through travel in several States, not vary with the passage from State to State (Babcock v. Jackson, 12 N Y 2d 473, 483, supra; Cavers, op. cit., pp. 146, 166-180; Rosenberg, An Opinion for the New York Court of Appeals, 67 Col. L. Rev. 459). Consequently, under the above-mentioned principle of preference, Maine law should be applied.
It is true that under another choice-of-law technique based on interest analysis, the rule of New York, an interested forum, should be applied (see, e.g., Currie, op. cit., pp. 183-184). This approach leads inevitably to the conclusion that a person carries most of the rights and defenses of his domiciliary law about with him, at least insofar as suits brought in his domiciliary State are concerned. Indeed, even more sweeping inroads upon the territorial system are suggested. It is said that States have an interest in applying their law to protect legitimate objects of their legislative concern (Currie, op. cit., pp. 85-86; Currie, Survival of Actions: Adjudication versus Automation In the Conflict of Laws, 10 Stan. L. Rev. 205, 221-222). Since States undoubtedly have an ‘ ‘ interest ’ ’ in protecting persons other than true domiciliaries (e.g., “residents,” and sojourners of extended stay), under this approach many persons would carry a number of different personal laws with them which they could invoke in the respective forums after a transaction leading to their injury. Under this approach, a rigid personal law would be substituted for the one-time, just as rigid, territorial law which invariably applied the law of the place of the tort, leading to anomalous results or results which belied the theory. Such a conclusion is contrary to the facts of a society territorially organized as this one is.
Governmental interest analysis assumes that a State’s policy concerning a particular issue of private law is a vector of fixed *27quantity and direction, and that a court’s duty in a conflicts case is to apply its forum’s “ policy ” whenever such application is reasonable. Rather, in civil cases courts attempt to do justice between the parties. Of course, in making these decisions, they draw upon the moral, economic and political assumptions of the society which they represent. This is not to say, however, that the society has an “ interest ” in seeing that its rule of civil law is applied whenever it has the power to apply it.1 Moreover, a court is in no sense bound to prefer its domestic rule; for, although the court must presume that rule to be most just in an all-domestic case, the very existence of out-of-State elements in the transaction may make it unjust to apply that rule. (See Ehrenzweig, A Counter-Revolution in Conflicts Law? From Beale to Cavers, 80 Harv. L. Rev. 377, 392; Kegel, The Crisis of Conflict of Laws, 112 Recueil des Cours, 91,182-186, 202; Rheinstein, Book Review, 32 U. Chi. L. Rev. 369, 371; Yntema, The Objectives of Private International Law, 35 Can. B. Rev. 721, 740.)
To be sure, interest analysis, and more particularly governmental interest analysis, provides a valuable method, as did the simpler grouping of contacts in the earlier stage of the development of modern theories of conflicts of laws. But the value of the method should not be impaired by rigid application to the exclusion of everything else that has been learned about the difficult problems in this field. Overall, the greatest value of interest analysis is in resolving false conflicts. When true conflicts are involved, as in this case, excluding, of course, the post-accident events, the problem does not lend itself to such automatic solution or the facile evaluation of predominant interests.
Infusing the old territorial rules as well as the newer theories of grouping of contacts or interest analysis is a desire to satisfy the reasonable expectations of persons participating in transactions. This is perhaps the dominant motif in the adjudication of multistate transactions, and therefore generally leads to the *28“ justice ” of the determination (see Rheinstein, Book Review, 32 U. Chi. L. Rev. 369).
Justice favors the fulfillment of expectations for two reasons. First, parties may have acted in reliance upon their assumption that courts would apply a certain rule of decision, and application of a different rule to their detriment would then be unjust. Of course, this pragmatic significance of expectations varies with the type of legal rule involved. It is undoubtedly strongest in contract cases (see, e.g., Auten v. Auten, 308 N. Y. 155; Ehrenzweig, Conflict of Laws, §§ 175-184). But it exists even in tort cases (see Cavers, op. cit., pp. 23-26, 36, 45, relating the views of Professor Rheinstein).
Justified expectations are also relevant in a second, more intangible, way: it is jurisprudentially significant that parties’ rights be determined by the law or system of rules which they most probably believed would control their relationship. In this respect, the application of the proper law of the tort exercises an influence in “ promoting an unconscious acceptance of legality and legal order ” (Kegel, op. cit., pp. 91, 184). Thus, in guest-statute cases, courts have applied a rule of guest-host liability which, in effect, reflects the parties’ unexpressed but undoubted assumption that a single system of rules will control guest-host liability no matter where the accident happens to take place. For this reason there is much to be said for applying the rule of the seat of the relationship, or, alternatively, that of the common domicile of guest and"host (Cavers, op. cit., pp. 166-180; Rosenberg, op. cit., pp. 459, 463). In the present case, however, there exists no touring relationship, amounting to a virtual status relation, of the sort presented in a guest-statute case. In view of all the Maine elements in this case, it is hard to deny that these parties would have felt that Maine law controlled every aspect of their liability inter se regarding a Maine accident during a trip occasioned by and incidental to their mutual business in Maine. To this extent, the law of the place of the wrong assumes significance, albeit not that which it had under older territorial thinking involving a mechanical, rigid and unacceptable approach, but to the degree that the place of the wrong may reflect the conscious understanding and choice of persistent localized activity of the parties.
*29It is often said that the old conflicts thinking was parochial, and indeed it was. But there could develop a new and even more perverse parochialism if the forum should apply its own local law to every or to too many foreign transactions, solely on the basis of its concern with its own domiciliaries. This would be truly parochial in a jurisprudence that otherwise rejects the doctrine of personal or national law as inherent in the factor of domicile. The territorial conceptualism of another day rested on a vested right arising from the creation of the cause of action in its presumed locus. It would be easy to substitute a domiciliary conceptualism that rested on a vested right accruing from the fact of domicile. Neither extreme reflects the understanding and basis on which human beings conduct their affairs, especially in multistate transactions, and especially in a society territorially organized (see, e.g., Cavers, op. cit., p. 135).
If this action had been brought in a neutral forum, the court would be likely, however it might present its reasoning, to look to the reasonable expectations of the parties as influenced or based upon territorial or personal factors which bore relevantly on the transaction (e.g., conduct), or the status of persons involved (e.g., intra-family' liability, guest-host relations, and the like). Interestingly, this court in Long v. Pan Amer. World Airways (16 N Y 2d 337), finding no New York interest (p. 342), used an interest analysis, but reached a result wholly consistent with one based on the reasonable expectations of the parties, which is as it ought to be. In Tramontana v. S. A. Empresa De Viacao Aerea Rio Grandense (350 F. 2d 468) the court also used variously a grouping of contacts and an interest analysis and, as a neutral forum, applied a Brazilian death recovery limitation contrary to the rule of decedent’s domicile.
At this point it is essential to consider the rule in Kilberg v. Northeast Airlines (9 N Y 2d 34). That rule sustained plaintiff’s right to recover under a Massachusetts death statute but excised the then limits in that statute as procedural and contrary to the strong policy of New York. The procedural basis for the holding was qualified in a subsequent holding by this court (Davenport v. Webb, 11 N Y 2d 392). But the case presents no problem, once the procedural basis is disregarded, because the facts are so distinguishable. In the Kilterg case, decedent was a New Yorker, he purchased his airline ticket in New York, and the *30fatal trip originated in New York. Massachusetts was but the terminus of the trip, the site of the fatal crash, and the home office of the airline. It was not too difficult, therefore, on either contacts or interest analysis, to apply New York law, which the court, in effect, did. Again it must be emphasized that there is a difference between an accident during a transient passage through a State and one where the parties have a more settled or persisting local relationship. Moreover, there are, arguably, differences between airplanes and motorcars or other ground-oriented means of transportation. The particular place in which an airplane happens to be when it crashes jiist is not a significant tie to the particular ground over which it flies. It is significant too that several authorities have distinguished airplane accidents and other transactions on precisely this ground (see Wilcox v. Wilcox, 26 Wis. 2d 617, 629-630; cf. Casey v. Manson Constr. & Eng. Co., 428 P. 2d 898 [Ore., 1967]; Cavers, op. cit., pp. 148 [n. 14], 149). And two cases have held the Kilberg rule inapplicable to airplane flights made entirely within one jurisdiction (Tramontana v. S. A. Empresa De Viacao Aerea Rio Grandense, 350 F. 2d 468, supra; Ciprari v. Servicos Aereos Cruzeiro, 245 F. Supp. 819, 824 [S. D. N. Y., 1965] [applying New York conflicts rules]).
Finally, one comes to the post-accident events in this case. If these are considered the case can be disposed of as involving only a false conflict. Important policies, however, require that these events be disregarded.
The first post-accident event is the repeal by Maine of its monetary limitations in death actions. The accident happened in 1961. This action was brought the next year. Four years after the accident and three years after the action was brought the statute was repealed. It satisfies no acceptable analysis to give the repealer retroactive effect, which concededly it would not have in Maine.2 A monetary limitation is hardly a pro*31cedural matter, but one of substance and, apart from the option a choice of law might give, such retrospective application of the repealer would suggest grave constitutional problems. The repealer is significant, undoubtedly, in suggesting that Maine’s policy in limiting recoveries in death actions was a waning one and therefore its waning policy ought to yield to the strong, and even constitutional, policy of New York in allowing unlimited recoveries. This reasoning assumes that New York’s interest should apply in any event, but if so then it ought to be immaterial whether Maine’s statute has been repealed or not. Of course, on the views taken here such governmental interest analysis, if applied exclusively, is largely inappropriate to any particular private civil litigation. Moreover, to give recognition to the repealer involves a new and rather undersirable consequence of determining liabilities for wrongful conduct on the basis of post-accident events, and, even worse, of post-litigation events.
The other post-accident event, of course, is the return to New York residence of the defendants. This is an unsound basis for determining the choice of law, even if used merely as one of several factors. Again, it would mean that liabilities would be determined by an irrelevant (to the accident) post-accident event. When would the change of residence become an acceptable factor: before or after action begun; before or after the Statute of Limitations had run in Maine; before or after the normal Statute of Limitations had run in New York? Would the change be just as effective if defendants had never before resided in New York? What weight would a neutral forum be likely to give to such a factor? All of these are questions that suggest anomalies for the future and prevent a rational and, therefore, a just development of the law in this difficult field. Thus far too, ho court, even with the new freedom provided by modem choice-of-law rules, has given any effect to post-transaction events not directly relevant to the transaction, and the reasons are rather obvious.
It is quite helpful to look to the handling of a case, very similar in many, but not all, respects, to this one. In California, the Supreme Court of that State recently decided Reich v. Purcell (63 Cal. Rptr. 31). There, the accident occurred in Missouri which has a statutory monetary limitation in death *32actions. The decedents were Ohioans on their way to California where they were thinking of settling. Their survivors, the plaintiffs in the death action, now reside in California. Defendants in the action were Californians on their way to Illinois for a vacation. The decedents’ automobile and that of defendants collided in Missouri, resulting in the deaths upon which the California action was brought. Using an interests analysis, the California court, in an opinion by Mr. Chief Justice Traynor, held that the Ohio law should be applied. The law of Missouri was rejected as having only transient significance. The law of California which had no monetary limitations was also rejected as inapplicable, the post-accident residence of the survivors being held to be irrelevant. Thus wrote Mr. Chief Justice Traynor: ‘ ‘ As the forum we must consider all of the foreign and domestic elements and interests involved in this case to determine the rule applicable. Three states are involved. Ohio is where plaintiffs and their decedents resided before the accident and where the decedents’ estates are being administered. Missouri is the place of the wrong. California is the place where defendant resides and is the forum. Although plaintiffs now reside in California, their residence and domicile at the time of the accident are the relevant residence and domicile. At the time of the accident the plans to change the family domicile were not definite and fixed, and if the choice of law were made to turn on events happening after the accident, forum shopping would be encouraged. (See Cavers, op. cit., supra, p. 151, fn. 16.) Accordingly, plaintiffs’ present domicile in California does not give this state any interest in applying its law, and since California has no limitation of damages, it also has no interest in applying its law on behalf of defendant. As a forum that is therefore disinterested in the only issue in dispute, we must decide whether to adopt the Ohio or the Missouri rule as the rule of decision for this case.”
The limitation upon the use of post-transaction events does not rest exclusively upon a desire to avoid forum shopping by prospective plaintiffs. Thus, in Gore v. Northeast Airlines (373 F. 2d 717, 723), the Second Circuit refused to consider the removal of decedent’s widow and two children from New York to another State after the transaction, even though the removal could not have been based upon forum shopping since the change, if con*33sidered determinative, "would have limited plaintiff’s recovery. Other policies supporting the rule against considering post-transaction events can be easily postulated. First, permitting reference to defendant’s post-accident domicile in New York so as to permit application of New York’s unlimited measure of damages would improperly discourage a defendant from becoming domiciled in this State. Second, where, as here, the case involves litigation between family members, the possibilities of collusive change of domicile to fix broader liability upon the insurer are obvious.
In the foregoing effort at analysis there is no complete allegiance to any one of the competing theories current in the conflicts of laws. Nor has there been a complete rejection of any one theory. Each, in particular areas of law, and each, with respect to particular rules of law, offers valuable insights into what courts in fact do in deciding multistate cases. To some extent, they also offer valuable insights into what courts ought to do. Anomalies, however, are inevitable, and, therefore, injustice, if any one of the theories is applied rigorously to each of the many situations confronting the courts. In this very difficult and still inchoate field of law a case-by-case development of rules is necessary in order to elicit or establish sound principles. Such a rational approach must be attempted if adjudication is not to be limited to ad hoc determinations. The objective is to achieve justice in a particular case and cases of like kind, avoiding ideology, on the one hand, and particularistic result-oriented determinations, on the other (see Cavers, op. cit., pp. 121-128). As for the precedents, in none of the recent landmark cases in this court did the court purport to enact any code of conflicts rules but to decide the cases before it in the light of modern creative reasoning in this troubled area. The instant case presents another facet of the problem and an effort to accommodate the developments thus far established to its resolution. On the views expressed here, in this case and similar cases that may arise, involving predominantly localized activity, the situs rule with regard to limitations of recovery, extant at the time of the accident, ought to be applied.
Accordingly, I dissent and vote to reverse the order dismissing the affirmative defense.
*34Chief Judge Fuld and Judges Burke and Bergan concur with Judge Keating ; Judge Breitel dissents and votes to reverse in an opinion in which Judge Jasen concurs; Judge Scileppi dissents and votes to reverse on the authority of Dym v. Gordon (16 NY 2d 120).
Order affirmed, with costs. Question certified answered in the affirmative.
. This is true even where, as here, the domestic civil law rule is incorporated in the State Constitution. Like statutes, constitutional provisions are almost always intended to apply only to acts within the territory. Absent an express choice-of-law provision, they are deemed intended as plenary within a territory and nonfunctioning outside the territory (e.g., McKinney’s Cons. Laws of N. Y., Book 1, Statutes, § 149, and cases cited).
. Dalton v. McLean (137 Me. 4) is of interest in this regard. In the Dalton case, the Supreme Judicial Court of Maine accepted the rule that whether a claim for damages for tort survives the death of the tort-feasor is determined by the place of the wrong. Nevertheless, it refused to apply a statute of New Brunswick (the situs of the accident) which retroactively provided for the survival of actions, on the ground that the retroactive feature violated the public policy of Maine (citing Loucks v. Standard Oil Co., 224 N. Y. 99). The court noted the undesirability of subjecting administrations of estates to liability under statutes enacted after the death of the tort-feasor.