These two appeals are from orders denying motions to dismiss complaints of New York residents in automobile accident actions against nonresident defendants wherein jurisdiction was predicated on attachment of the defendants’ interests in liability insurance policies issued by companies doing business in New York, CPLR §§ 5201, 6202. They require us to consider the constitutionality of the procedure sanctioned by the New York Court of Appeals in the much discussed case of Seider v. Roth, 17 N.Y. 2d 111, 269 N.Y.S.2d 99, 216 N.E.2d 312 (1966) , as elaborated in Simpson v. Loeh-mann, 21 N.Y.2d 305, 287 N.Y.S.2d 633, 234 N.E.2d 669 (1967), motion for re-argument denied, 21 N.Y.2d 990, 290 N.Y.S.2d 914, 238 N.E.2d 319 (1968). See also Victor v. Lyon Associates, Inc., Hanover Ins. Co., Appellant, 21 N.Y.2d 695, 287 N.Y.S.2d 424, 234 N.E.2d 459 (1967) , appeal dismissed for want of a substantial federal question, 392 U.S. 8, 89 S.Ct. 44, 21 L.Ed.2d 8 (1968).

Marie Minichiello, a resident of New York, brought suit in the Supreme Court of New York for Schuyler County, in her own right and as executrix of her husband Thomas, to recover $205,050 for injuries to Thomas, for his death, for injuries to herself, and for damage to their car in an accident near Harrisburg, Pa., allegedly caused by the negligence of the defendant Rosenberg, a resident of Pennsylvania. An order of attachment was served at an office in New York on Allstate Insurance Co., which had issued to Rosenberg in Pennsylvania a policy insuring against liability in an amount less than the recovery sought. Rosenberg removed the action to the District Court for the Western District of New York and there sought an order dismissing the complaint on the ground that the procedure sanctioned by Seider v. Roth violated the Federal Constitution. Judge Burke denied the motion1 but *108later made the certificate required by 28 U.S.C. § 1292(b). Since the problem was of obvious importance and the decision ran counter to Judge Croake’s in Podolsky v. Devinney, 281 F.Supp. 488 (S.D.N.Y. 1968), we granted leave to appeal.

Elwin W. Stevens, a resident of New York, brought suit in the Supreme Court for Rensselaer County on his own behalf and as guardian for his son Dennis against Stephen Tyng, Sr., and Stephen Tyng, Jr., residents.of Massachusetts, to recover $200,350 for injury to Dennis, loss of Dennis’ services, and damage to a motor scooter sustained at Orleans, Massachusetts, as a result of the alleged negligence of Tyng, Jr., in driving a car owned by his father. Plaintiff attached a policy insuring against liability in the amount of $100,000 for any person and $300,000 for any accident issued outside New York by American Motorists Insurance Co., which does business in New York. Defendants removed the action to the District Court for the Northern District of New York and moved, inter alia, to dismiss the complaint for want of jurisdiction. Judge Ryan, sitting in the Northern District by designation, denied the motion2 but, recognizing the contrary decision in Podolsky, made the certificate specified in 28 U.S.C. § 1292(b). Since the issue was identical with that already argued in Minichiello v. Rosenberg, we granted leave to appeal, with the case to be submitted on briefs.

Seider v. Roth was an action by New York residents against Lemiux, a Canadian, for injuries suffered in an automobile accident in Vermont.3 Plaintiffs attached in New York a liability policy issued to Lemiux in Canada by Hartford Accident and Indemnity Co., which does business in New York. The Court of Appeals, speaking through Chief Judge Desmond, treated the problem solely as one of statutory construction, namely, whether “as soon as the accident occurred there was imposed on Hartford a contractual obligation which should be considered a ‘debt’ within the meaning of CPLR 5201 and 6202.” The court answered this in the affirmative. Neither the majority opinion nor the dissent by Judge Burke, in which Judges Bergan and Scileppi concurred, adverted to constitutional doubts.

Seider v. Roth received a poor press from the commentators. See, e. g., Reese, The Expanding Scope of Jurisdiction over Non-Residents — New York Goes Wild, 35 Ins. Counsel J. 118 (1968); Comment, Attachment of “Obligations” —A New Chapter in Long-Arm Jurisdiction, 16 Buffalo L.Rev. 769 (1967); Comment, Garnishment of Intangibles: Contingent Obligations and the Interstate Corporation, 67 Colum.L.Rev. 550 (1967); Note, Seider v. Roth: The Constitutional Phase, 43 St. John’s L. Rev. 58 (1968); Comment, Quasi in Rem Jurisdiction Based on Insurer's Obligations, 19 Stan.L.Rev. 654 (1967); and a number of comments by Professor David Siegel of St. John’s University School of Law, one antedating the Court of Appeals’ decision, which have appeared in the annual supplements to CPLR § 5201 in McKinney’s Consolidated Laws of New York. In Simpson v. Loehmann, supra, 21 N.Y.2d 305, 287 N.Y.S.2d 633, 234 N.E.2d 669, the court was asked to reconsider Seider v. Roth hot only as going beyond the New York attachment statutes but as offending the due process clause of the Fourteenth Amendment, imposing an undue burden on interstate commerce in insurance, and impairing the obligations of the policy. The court reaffirmed its earlier ruling. While only Judges Burke and Scileppi dissented *109from this, Judges Breitel4 and Bergan concurred solely “on constraint of Seider v. Roth,” saying that “only a major reappraisal by the court, rather than the accident of a change in its composition, would justify the overruling of that precedent,” which, however, they severely criticized “if only, perhaps, to hasten the day of its overruling or its annulment by legislation.”

On the constitutional claims, Chief Judge Fuld, writing for himself and Judge Van Voorhis, summarily disposed of the interstate commerce and impairment of contract obligation arguments, 21 N.Y.2d at 309 n. 2, 287 N.Y.S.2d at 635 n. 2, 234 N.E.2d at 670 n. 2. He then pointed out that under modern views the sufficiency of a basis for jurisdiction depends on “a practical appraisal of the /situation of the various parties rather / than an emphasis upon somewhat magical and medieval concepts of presence and power.” 21 N.Y.2d at 311, 287 N.Y.S.2d at 637, 234 N.E.2d at 672. He stressed the insurer’s control of the litigation and that “where the plaintiff is a resident of >/ the forum state and the insurer is present in and regulated by it, the State has a substantial and continuing relation with l^the controversy.” Judge Keating, concurring, reasoned that New York could validly enact a direct action statute in favor of its residents for out-of-state accidents and that there was “no policy reason for not holding that service of process on the real party defendant — the insurer — is sufficient to compel it to defend in this State, provided it transacts business here and is thus subject to the jurisdiction of our courts.” 21 N.Y. 2d at 312-314, 287 N.Y.S.2d at 638, 234 N.E.2d at 673. Judges Breitel and Bergan found it unnecessary to deal with the constitutional issues since, on their view, “if the court was right in the Seider case, then there is no constitutional question,” 21 N.Y.2d at 316, 287 N.Y.S.2d at 642, 234 N.E.2d at 675. Judges Burke and Scileppi thought Seider was of doubtful constitutionality on any of the bases asserted in its support.

Although Judge. Keating’s opinion in Simpson v. Loehmann is more explicit than Chief Judge Fuld’s on this score, it is reasonably clear that the Court of Appeals regards Seider v. Roth as in effect a judicially created direct action statute. The insurer doing business in New York is considered the real party in interest and the nonresident insured is viewed simply as a conduit, who has to be named as a defendant in order to provide a conceptual basis for getting at the insurer. See 1968 Commentary to CPLR § 5201, paras. 88, 89. It is thus useful to address ourselves in the first instance to Judge Keating’s premise that New York could constitutionally provide^ for a direct action against an insurer doing business in New York by a New < York resident with respect to an injury ¡ suffered elsewhere. If that premise should fail, Seider would almost certainly fall with it. If it stands we must then examine whether Seider may still be vulnerable because, unlike a true direct action statute, it involves the nonresident insured as a party defendant.

In Watson v. Employers Liability Assurance Corp., 348 U.S. 66, 75 S.Ct. 166, 99 L.Ed. 74 (1954), the Supreme Court sustained Louisiana’s direct action statute, which overrode contrary provisions in liability policies issued by insurers doing business in that state whether or not the policy had been written or delivered within it. The statute, however, applied only to accidents or injuries occurring in Louisiana. The first set of considerations marshaled by Mr. Justice Black in favor of its validity, 348 U.S. at 72-73, 75 S.Ct. 166, 170, would relate with equal force to one on behalf of residents. He stressed that “persons injured or killed in Louisiana are most likely to be Louisiana residents, and even if not, Louisiana may have to care for *110them. Serious injuries may require treatment in Louisiana homes or hospitals by Louisiana doctors. The injured may be destitute. They may be compelled to call upon friends, relatives or the public for help.” On the other hand, his next argument, that “Louisiana courts in most instances provide the most convenient forum for trial of these cases,” by its very terms could not apply. While the place of the plaintiff’s residence may be a convenient forum for the trial of an action arising from an out-of-state accident in that not only he but much of the evidence of damages may be there, it will rarely be “the most convenient” one since the other witnesses to the accident are elsewhere — possibly far away.5 But the Justice’s final consider"ation — the plaintiff’s difficulty in bringing the defendant before the forum— applies with even greater force to the state of plaintiff’s residence than to that of injury in light of the development of long-arm statutes that will generally allow the state of injury to obtain personal jurisdiction of the insured and so avoid the need for a direct action against the insurer.

We thus believe that, all things considered, the Supreme Court would sustain the validity of a state statute permitting direct actions against insurers doing business in the state in favor of residents as well as on behalf of persons injured within it. See Note, Direct-Action Statutes: Their Operational and Conflict-of-Law Problems, 74 Harv.L. Rev. 357, 361 (1960). The state’s interest in protecting its residents is as great as in the case of nonresidents injured within the state. While the burden on the insurer in trying a case in a state other than the locus of the accident is heavier, there has been, as we have recently noted, “ ‘a movement away from the bias favoring the defendant,’ in matters of personal jurisdiction ‘toward permitting the plaintiff to insist that the defendant come to him’ when there is a sufficient basis for doing so,” Buckley v. New York Post Corp., 373 F.2d 175, 181 (2 Cir. 1967), quoting Von Mehren & Trautman, Jurisdiction to Adjudicate: A Suggested Analysis, 79 Harv.L.Rev. 1121, 1128 (1966). Indeed, the Supreme Court may have already decided this point by dismissing Hanover Ins. Co. v. Victor, 1968, “for want of a substantial federal question,” see 393 U.S. 7, 89 S.Ct. 44 (1968), although the order there sought to be reviewed would not seem to have been appealable under 28 U.S.C. § 1257. Since, on the other hand, we doubt that the Court would sanction a direct action statute where the state was neither the place of injury nor the plaintiff’s residence, we equally doubt whether Seider could validly be applied in such a case.-6

With so much established, we turn to the question whether the position of the insured is as nominal as the New York Court of Appeals conceives. Here we proceed from the basis, apparently accepted by everyone, that the mere fact of a nonresident’s having taken out liability insurance with a company doing business in New York would not be a sufficient contact to enable New York to render a judgment enforceable against *111other assets even at the suit of a resident.

In Podolsky v. Devinney, supra, 281 F.Supp. at 495-497, Judge Croake cogently argued that Seider v. Roth came close enough to this to constitute a denial of due process. He thought that, under the seemingly clear language of CPLR § 320(c), if the insured defended on the merits, he would subject himself to in personam jurisdiction with consequent personal liability for any judgment in excess of the policy limits. On the other hand, if the insured did not defend and judgment was entered by default, the insurer should be able to resist a levy on the basis of the insured’s non-corporation. In cases where the plaintiff’s claim exceeded the policy limits, Seider thus produced either a constitutionally unacceptable result or a nullity.

These difficulties were largely swept away by what has been termed “a miraculous per curiam opinion,” see Commentary on CPLR § 5201, at 15 (McKinney Supp. 1968), denying reargument in Simpson v. Loehmann, 21 N.Y.2d 990, 290 N.Y.S.2d 914, 238 N.E.2d 319 (1968). The Court of Appeals began by criticizing the argument just summarized, which had not been earlier advanced before it, as failing to take account of its statement, 21 N.Y.2d at 310, 287 N.Y.S. 2d at 636, 234 N.E.2d at 671, that the Seider doctrine did not “expand the basis for in persona jurisdiction in view of the fact that the recovery is necessarily limited to the value of the asset attached, that is, the liability insurance policy.” The “miracle” lay in the next sentences:

“This, it is hardly necessary to add, means that there may not be any recovery against the defendant in this sort of case in an amount greater than the face value of such insurance policy even though he proceeds with the defense on the merits. Consideration of CPLR 320 (subd. [c]) and its effect in other • types of action begun by attachment must, of course, await future cases.” (Emphasis supplied.)7

Appellants respond that while this unexpected construction of CPLR § 320(c) protects the defendant in the New York Seider action, it does not fully meet the problem where as here the demands exceed the policy limits. Their fear is that if the plaintiff sues later for the unmet balance in a state where he has acquired jurisdiction in personam, a *112plaintiff’s judgment in the New York Seider action will be taken to have adjudicated the defendant’s liability. It is plain that the New York judgment could not be given effect as such beyond the amount of the insurance; the troubling issue is whether a state could deem it effective as a collateral estoppel. Commentators have indicated that the due process clause prevents a state from giving a quasi in rem judgment such effect. Carrington, Collateral Estoppel and Foreign Judgments, 24 Ohio State L.J. 381, 384 (1963); Cramton & D. Cur-rie, Conflict of Laws, 645-646 (1968); Developments in the Law — Res judicata, 65 Harv.L.Rev. 818, 835 (1952). Accord, Combs v. Combs, 249 Ky. 155, 60 S.W.2d 368, 89 A.L.R. 1095 (1933); contra, Harnischfeger Sales Corp. v. Sternberg Dredging Co., 189 Miss. 73, 191 So. 94, 195 So. 322 (1940). Other authorities have concluded that a quasi in rem judgment may work an estoppel on issues of fact that the defendant has actually litigated, at least where he has had a fair opportunity to do this. See Restatement of Judgments § 76(2) (1942); Taintor, supra, 8 U.Pittsburgh L.Rev. at 235; cf. Sherman v. Jacobson, 247 F.Supp. 261, 266-267 (S.D.N.Y. 1963); Sherman v. Kirshman, 369 F.2d 886, 890 (2 Cir. 1966). Whatever the right rule may be as to quasi in rem judgments generally, we think it clear that neither New York nor any other state could constitutionally give collateral estoppel effect to a Seider judgment when the whole theory behind this procedure is that it is in effect a direct action against the insurer and that the latter rather than the insured will conduct the defense. To be sure it may be cold comfort to a nonresident defendant to have our assurance that if some state should be so misguided as to consider a New York Seider judgment as concluding him, he will be able to have this ruling overturned by the Supreme Court of the United States. But we cannot fairly hold that New York has denied due process merely because of the possibility that some other state may do so.

We can foresee that application of Seider may produce other problems of due process. One may arise from CPLR 3110, subd. 1, which provides that a party must appear in New York for the taking of his deposition. This has been held to apply to a nonresident defendant over whom jurisdiction has been obtained only by the Seider route, Gazerwitz v. Adrian, 28 A.D.2d 556, 280 N.Y.S.2d 233 (2d Dept. 1967). Nevertheless, if such a defendant declined so to appear despite the immunity from service of process that New York has long recognized,8 we doubt that a default judgment could validly be entered. But we can hardly assume either that Seider plaintiffs will recklessly imperil their own recoveries by insisting on defendants’ coming into the state or that, if they did, the Court of Appeals would not rescue them from their folly by reading into CPLR 3110, subd. 1 a conclusive presumption of hardship in Seider cases sufficient to excuse the defendants’ appearing. Moreover, such matters are of no concern to the defendants here since service of subpoenas upon them is governed by F.R. Civ.P. 45. Problems may also arise about counterclaims if the pronouncement in the opinion denying reargument in Simpson as to the limited effect of an appearance in a Seider action does not cover their assertion. What would otherwise be the most serious difficulties in this area are removed by the absence of a compulsory counterclaim rule in New York, Frank M. Herbert, Inc. v. M & P Scrap Iron & Metal Corp., 41 Misc.2d 1020, 1024, 247 N.Y.S.2d 193, 197 (2d Dept. 1964), and, so far as concerns these actions by the careful exclusion in F.R. Civ.P. 13(a) of a case where “the opposing party brought suit upon his claim by attachment or other process by which the court did not acquire jurisdiction to render a personal judgment on that claim, and the pleader is not stating any coun*113terclaim * * The argument thus has to become that it is unfair to put a nonresident with whom New York has no real contact in a position where he can assert a counterclaim only at the price of exposing himself to liability above the policy limits. But the insured would be in no position to assert a counterclaim under a direct action statute, and we do not now see how New York would deny him due process if it allowed him to assert one in a Seider action only at the risk of personal liability. In any event no such problem is presented here; we leave such questions and others that may be hypothesized for further consideration when they arise. It suffices for decision that the appellants have not demonstrated that refusal to dismiss the complaints will deny them the due process of law which the Fourteenth Amendment prescribes.

Appellants’ argument based on the clause of the Constitution, Art. I, § 10, clause 1, forbidding impairment of the obligation of contracts does not require extended answer. The argument is that a Seider attachment compels the insured to breach his pledge of cooperation by making the assertion of a defense too hazardous and thus jeopardizes his insurance coverage. But this contention has been drained of force by the per curiam opinion denying reargument in Simpson, 21 N.Y.2d 990, 290 N.Y.S.2d 914, 238 N.E.2d 319, which we have previously discussed.

Affirmed.

(dissenting) :

I dissent. As the analysis of Seider v. Roth, 17 N.Y.2d 111, 269 N.Y.S.2d 99, 216 N.E.2d 312 (1966), and related cases in the majority opinion states, the Seider procedure is in effect a judicially created direct action statute that provides a New York plaintiff with a New York forum with respect to an injury suffered elsewhere, provided that in per-sonam jurisdiction is obtained over the insurance company of the defendant. So regarded, the constitutional question presents itself as follows:

May a state, which has in personam jurisdiction over an insurer and is the residence of the injured plaintiff, exercise jurisdiction in a tort case involving an accident which took place in a foreign jurisdiction and which was allegedly caused by a resident of that jurisdiction, to secure the proceeds of a liability policy issued in that foreign jurisdiction?1

The case which must provide the cue for the resolution of the issue is Watson v. Employers Liability Insurance Corp., 348 U.S. 66, 75 S.Ct. 166, 99 L.Ed. 74 (1954). Considerations which governed the decision in that case appear in the following portion of the opinion:

“Louisiana’s direct action statute is not a mere intermeddling in affairs beyond her boundaries which are no concern of hers. Persons injured or killed in Louisiana are most likely to be Louisiana residents, and even if not, Louisiana may have to care for them. Serious injuries may require treatment in Louisiana homes or hospitals by Louisiana doctors. The injured may be destitute. They may be compelled to call upon friends, relatives, or the public for help. Louisiana has manifested its natural interest in the injured by providing remedies for recovery of damages. It has a similar interest in policies of insurance which are designed to assure ultimate payment of such damages. Moreover, Louisiana courts in most instances provide the most convenient forum for the trial of these cases. But modern transportation and business methods have made it more difficult to serve process on wrongdoers who live or do business in other states. In this case efforts to serve the Gillette Company were answered by a motion to dismiss on the ground- that Gillette had no *114Louisiana agent on whom process could be served. If this motion is granted, Mrs. Watson, but for the direct action law, could not get her case tried without going to Massachusetts or Illinois although she lives in Louisiana and her claim is for injuries from a product bought and used there. What has been said is enough to show Louisiana’s legitimate interest in safeguarding the rights of persons injured there. In view of that interest, the direct action provisions here challenged do not violate due process.”2 3

The majority compares the above quoted language with the Seider procedure and concludes that Louisiana’s interest in protecting its medical creditors and in preventing the expenditure of public funds in the case of the injury of one of its residents in an out-of-state accident is as great as in the case of an accident occurring in Louisiana. This might be true if the plaintiff is hardy enough to reach Louisiana before requiring medical aid or public assistance; otherwise the state where the accident occurred would be required to pick up the tab. But I read this portion of Mr. Justice Black’s opinion as stating certain pertinent examples of the interests which the state of the accident has, rather than as an exhaustive catalogue of those interests. It seems to me that there are other substantial interests of the accident state which do not coincide with those of the residence state. For example, one such traditional and significant interest, which the accident state may have in enacting a direct action statute, is to deter tortious conduct within the state.3 Another is the fact that the accident state is often in a very real sense the center of the controversy with at least a greater statistical probability that its residents will be more directly affected by any resolution of disputes arising from an accident within its territorial limits than the residents of any other state.

Moreover, as commentators have noted,4 the second major argument of *115Mr. Justice Black also emphasizes the occurrence of the accident in Louisiana: “Louisiana courts in most instances provide the most convenient forum for trial of these cases.” The opinion places great stress upon considerations of convenience in litigation, which point to the accident state, e. g., the availability of witnesses, the availability of police and medical records, the possibility of affording a jury an opportunity to view the scene of the accident, the ease of ascertaining the governing law, etc. By comparison, the convenience in carrying on the litigation which the majority ascribes to the residence state, is to say the least, a very one-sided accommodation. It would only be in the very small minority of these tort actions, transferred to the United States District Courts, as most of them will be, that the defendant could subpoena witnesses who were within one hundred miles of the New York courthouse, and he would, in most cases be forced to prove his defenses through depositions.

Recognizing inconvenience of trial in New York courts as a vulnerable point in its argument, the majority seeks to offset it by the following a fortiori argument:

“But the Justice’s final consideration —the plaintiff’s difficulty in bringing the defendant before the forum — applies with even greater force to the state of plaintiff’s residence than to that of injury in light of the development of long-arm statutes that will generally allow the state of injury to obtain personal jurisdiction of the insured and so avoid the need for a direct action against the insurer.”

Every plaintiff, however, is not entitled to a legal remedy in the form of a device to bring the defendant into the forum most convenient to the plaintiff and commensurate with the degree of difficulty which that effort entails, regardless of all other considerations. Certainly it must first be determined whether the plaintiff has the right to require defendant’s presence in plaintiff’s forum and the resolution of this question depends upon the fairness and reasonableness of the device as it affects all of the parties involved, the functioning of the courts, and the administration of justice in such an action. Watson and the long-arm statutes are plainly grounded in fairness because the nonresident defendant has of his own volition brought himself within the bounds and jurisdiction of the state where the accident occurred. But, for example, for a resident of Alaska, as the accident state, who had never resided elsewhere and who had never entered the State of New York or had any business there, suddenly to discover that he is under the jurisdiction of the State of New York and must go there to defend himself is quite a different matter. It is apparent from the Supreme Court’s opinion in Watson, that “plaintiff’s difficulty in bringing the defendant before the forum” is manifestly unjust only because the forum in question is “the most convenient forum.” To repeat the passage under discussion:

“Moreover, Louisiana courts in most instances provide the most convenient forum for the trial of these cases. But modern transportation and business methods have made it moire difficult to serve process on wrongdoers who live or do business in other states.”

348 U.S. at 72, 75 S.Ct. at 170.

The legislatures which have enacted long-arm statutes and courts which have interpreted them have been motivated and restrained by considerations of fairness and reasonableness. In deciding *116the due process issue,5 as it relates to the Seider procedure, “traditional notions of fair play and substantial justice,” International Shoe Co. v. Washington, 326 U.S. 310, 316, 66 S.Ct. 154, 158, 90 L.Ed. 95 (1945),6 must be used to determine whether or not it falls within permissible bounds. Weighed in this light, it seems to me that only the long-arm statutes asserting jurisdiction of the state where the accident occurred7 qualify as due process, whereas the assertion of jurisdiction by the state of the plaintiff’s residence does not.

The Seider procedure potentially subjects to liability to out-of-state process, a defendant whose activities, as noted above, may have been entirely local. It, therefore, may come as something of a shock to such a defendant, untutored in New York law, to discover that he is exposed to the traditionally high verdicts returned by New York juries. In Watson, on the other hand, the defendant was engaged in activity across state lines, and it was foreseeable to him that he should insure himself against potential liability in Louisiana. True, recovery in New York could not exceed the insurance proceeds, but if the actual verdict or judgment were in excess of the policy limits, would the defendant be equitably estopped from asserting that the damages were less than those found by the New York court, where the plaintiff sued on the judgment in a court outside of the State of New York?

In terms of fairness to the plaintiff, Seider involves by definition a plaintiff who is able to foresee that he may be injured in a foreign jurisdiction and compelled to litigate there. In Watson, however, the plaintiff’s activity was local, and it was certainly much less foreseeable to her that she might be forced to go to Massachusetts or Illinois in order to recover for her local injuries.

A balancing of considerations of fairness to the parties, of litigational convenience, and of the interests of the *117state (which is to say the interest of its citizens) (1) in compensating the injured party, both for his own sake and to prevent him from becoming a public charge, (2) in compensating its medical and other creditors, and (3) in deterring tortious conduct, leads, on the one hand to a finding that a direct action statute in the state where an accident occurs is fair and reasonable in the due process sense. On the other hand, a similar weighing of interests leads to the conclusion that a direct action statute purporting to provide a New York forum for a New York plaintiff, regardless of other local contacts or considerations, is unreasonable and parochial in the due process sense. It is one thing to require a resident of Alaska to make trips to New York to respond to an action arising out of an accident with a resident of New York in New York state to which the Alaskan had gone, and quite another to make him do so for an accident which occurred in Alaska when the New Yorker elected to go to Alaska and use its highways.

The present case in its application deals only with a New York plaintiff, following the Seider procedure and its impact on a defendant in any of the other forty-nine states which happened to be the accident state. If its constitutionality is upheld, it is quite likely that all fifty states will be compelled to enact like measures. As a result each party will commence suit as quickly as possible in an endeavor to have the litigation in the state where he resides. It would also appear to be a natural and logical consequence of the approval of the Seider procedure that retaliatory laws will be adopted by other states to impose certain conditions upon residents from Seider-procedure states to protect the states’ own residents from being subject to the jurisdiction of the foreign state. This is likely to impose a serious burden on interstate travel and commerce. It is also reasonably certain that the adoption of this procedure will work a substantial increase in the cost of insurance, which is not in the public interest.

It is my opinion that the Seider procedure violates due process and is, therefore, unconstitutional.

On rehearing in banc

Before LUMBARD, Chief Judge, and WATERMAN, MOORE, FRIENDLY, SMITH, KAUFMAN, HAYS, ANDERSON and FEINBERG, Circuit Judges.

FRIENDLY, Circuit Judge, with whom WATERMAN, SMITH, KAUFMAN, HAYS and FEINBERG, Circuit Judges, concur:

After reconsideration in banc on the original and supplemental briefs, the court adheres to the affirmance of these orders.

The burden of the petitions for rehearing is that the previous majority opinion was overly preoccupied with a conception of Seider v. Roth, 17 N.Y.2d 111, 269 N.Y.S.2d 99, 216 N.E.2d 312 (1966), as elaborated in Simpson v. Loehmann, 21 N.Y.2d 305, 287 N.Y.S.2d 633, 234 N.E.2d 669 (1967), motion for reargument denied, 21 N.Y.2d 990, 290 N.Y.S.2d 914, 238 N.E.2d 319 (1968), as “in effect a judicially created direct action statute,”1 at 109. Appellants complain *118we did not adequately focus on the burden to a nonresident of having to defend a personal-injury action in another state, conceivably far from his domicile and the place of the accident, simply because his liability insurer has chosen to do business there.

We find this argument unpersuasive so long as Harris v. Balk, 198 U.S. 215, 25 S.Ct. 625, 49 L.Ed. 1023 (1905), stands. Balk, a North Carolinian, had no more control over the peregrinations of his fellow-citizen, Harris, which caused the latter to be garnisheed in Maryland by Balk’s alleged creditor, Epstein, than these appellants have over their insurers’ decisions where to do business.2 In contrast, appellants’ problem is significantly less serious than was Balk’s in several respects. Balk had to decide whether to hire a Maryland lawyer to protect his interest in the $180 debt Harris owed him; the appellants are entitled to have lawyers in New York furnished by their insurers without expense. The Maryland judgment deprived Balk of money he could have used for whatever purpose he willed; a Seider judgment would mean simply that liability policies, on which appellants could not have realized for any purpose other than to protect themselves against losses to others, will be applied to the very objective for which they were procured. Whatever Balk’s situation may have been, appellants are assured that they may defend the New York actions without exposing themselves to personal liability and that any recovery by the plaintiffs “is necessarily limited to the value of the asset attached, that is, the liability insurance policy.” Simpson v. Loehmann, 21 N.Y.2d 990, 290 N.Y.S.2d 914, 238 N.E.2d 319 (1968), discussed at 111.

Appellants claim they are worse off than Balk in one respect, namely, that the cooperation condition of the policies forces them to come to New York and participate in the defense of the personal-injury actions, and that this is a serious inconvenience even though they are reimbursed for their expenses and, as we have concluded, will suffer no untoward consequences in the form of liability. We have previously noted that so far as judicially compelled attendance in New York is concerned, appellants are amply protected against undue inconvenience by F.R.Civ.P. 45(d) and (e). We are not required to determine whether refusal by the insured to attend a trial in a foreign state, when such attendance is not required by the applicable procedural law and he is willing to be deposed elsewhere, would constitute a material breach of the duty of cooperation even if a policy literally required him to appear in person. See Appleman, Insurance Law and Practice §§ 4773, 4784 (1962). For, as we understand Simpson v. Loehmann, supra, 21 N.Y.2d 309 n. 2, 287 N.Y.S.2d 635 n. 2, 234 N.E.2d 670 n. 2, the only consequence of such a breach would be to relieve the insurer of liability under the attachment.

We note a further point which at least mitigates some of the many undesirable practical consequences of Seider.3 By *119their very definition actions of the Seider type will generally be removable to the federal courts,4 or indeed can be brought there. Once in federal court, they become subject to the salutary provision of 28U.S.C. § 1404(a) that “for the convenience of parties and witnesses, in the interest of justice, a district court may transfer any civil action to any other district or division where it might have been brought.” Either the district of the defendant’s residence or that of the accident, or both, will qualify for transfer. Whenever a trial in New York would cause true hardship to the defendant, a remedy thus will almost always be at hand. See Farrell v. Wyatt, 408 F.2d 662 (2 Cir. 1968). While an influx of cases of this type into the federal courts is most unwelcome, we have here an unforeseen demonstration of the utility of diversity jurisdiction, by way of removal at the instance of a nonresident, to lessen the impact of an overly enthusiastic but not unconstitutional assertion of jurisdiction by another state. To be sure, if the Constitution demanded that a personal-injury claim be tried only at the place of the accident or the defendant’s residence, the availability of a discretionary transfer by a court to which nearly all such actions can be removed would be an inadequate answer. But the availability of a procedure assuring that in almost all such cases the trial can take place in what a federal district judge deems an appropriate forum bears strongly on the issue whether the Constitution does demand such a general rule. The very fact that the discretion will not always be exercised in favor of transfer — as here in the Stevens case where the judge found that the preponderant issue was the extent of plaintiff’s injury — sufficiently indicates that the places of accident or of defendant’s residence are not always the only convenient fora. Cf. Lykes Bros. S. S. Co. v. Sugarman, 272 F.2d 679 (2 Cir. 1959).

We add in conclusion that in these cases, where the accidents gave rise to what from a practical standpoint were single claims, we are not required to decide whether Seider might produce constitutionally impermissible results if applied so as to prefer one claimant over another in cases of multiple claims where the damages exceed the policy limits and the insured is without funds to pay the excess. We would hope that realization of the complexities of such problems and the considerations urged in our brother ANDERSON’S dissents might stimulate the reexamination of Seider intimated in Simpson v. Loehmann, supra, 21 N.Y.2d at 312, 287 N.Y.S.2d at 638, 234 N.E.2d at 672 (opinion of Chief Judge Fuld), 21 N.Y.2d at 814, 287 N.Y.S.2d at 640, 234 N.E.2d at 674 (concurring opinion of Judge Breitel, before the decision is taken up, perhaps only defensively, by other states.

Affirmed.

(concurring) :

I concur in the result and in Judge FRIENDLY’S opinion.

I wish to add a word simply to emphasize a point which is taken for granted in both Judge FRIENDLY’S and Judge ANDERSON’S opinions.

There is nothing in the character of attachment or of resulting in rem jurisdiction which would require that the procedure be available only to plaintiffs who are residents in the forum state. While in both Seider and Simpson the plaintiffs were residents of New York, nothing was decided in either of those cases which *120would deny to residents of states other than New York the procedure approved in the cases. In Judge Breitel’s concurrence in Simpson, 21 N.Y.2d 314 at p. 316,* 287 N.Y.S.2d 640 at p. 641, 234 N.E.2d 674 at p. 675 and in Judge Burke’s dissent in the same case ** a question is raised as to. whether non-residents of New York may resort to the Seider procedure.

Although I read Judge FRIENDLY’S opinion as finding the Seider procedure acceptable only when plaintiffs are residents of New York, I believe that it is worthwhile to put additional stress on this point and to make it completely clear that our opinion refers only to the situation where plaintiffs are New York residents.

with whom LUMBARD, Chief Judge, and MOORE, Circuit Judge, concur (dissenting):

In view of the majority’s strong reliance on Harris v. Balk, 198 U.S. 215, 25 S.Ct. 625, 49 L.Ed. 1023 (1905), as a justification for brushing aside the argument of the appellants, on their motion for rehearing, that the majority holding places an unfair burden on a nonresident (thereby compelled to defend an action in New York, regardless of other contacts, provided only that in per-sonam jurisdiction can be obtained over his insurance company) it is necessary to comment on the Harris case.

Assuming the continued constitutional viability of the Harris rule,1 that the ob*121ligation of a debtor to pay clings to and accompanies him wherever he goes and that the debt is subject to garnishment in any state in which the debtor is subject to in personam jurisdiction,2 nevertheless that case is clearly distinguishable from the procedure sanctioned by Seider, Simpson, et al. In the first place, the Supreme Court in Harris was considering “ordinary debts,” 3 whereas the complex package of rights and duties which derive from an automobile liability insurance policy is something quite different. The debt in Harris was fixed in amount, and concededly currently owed, but the Seider “debt,” consisting of the obligations to defend and indemnify, is indefinite, not readily ascertainable in amount, and contingent, i. e. with respect to the obligation to defend, upon the obtaining of jurisdiction over and the commencement of suit against the insured, and also, with respect to the obligation to indemnify, upon the security by the plaintiff of a judgment against the defendant insured. Moreover, as Judge Croake noted in Podolsky v. Devinney, 281 F.Supp. 488, 494 (S.D.N.Y. 1968), “the obligation to defend is an important right of the insurer, essential in protecting its financial interest in the outcome of the litigation.” Seider is the first case ever to hold that an insurer’s contractual obligations under an automobile liability policy are subject to foreign attachment and thereby furnish quasi in rem jurisdiction over a non-resident defendant. See the excellent article entitled “Jurisdiction by Attachment of Liability Insurance” by Malcolm L. Stein, 43 N.Y.U. L.Rev. 1075, Dec. 1968.

There is another distinction between Harris and Seider,4 In Harris the debt- or-garnishee, as a natural person, could be physically present in only one state at a time; therefore, although the situs of the debt might change, there could never exist more than one quasi in rem jurisdictional situs at any given time. In Seider the garnishee is an interstate corporation subject to jurisdiction in the state of incorporation as well as in the states where it is doing business. If Seider stands, other states will be compelled to follow New York’s example by engrafting a Seider procedure onto their attachment statutes. The inevitable result will be a rash of contrary and inconsistent adjudications, particularly in multi-party cases, well described by Professor Siegel as “jurisdictional chaos.” He says:

“Balk at least enabled a single state to be isolated as the situs of the intangible. Under Seider, that is not possible; all of the states in which the insurer does business could seize the policy; and -if they were to do so (i. e. adopt Seider) there would be a kind of jurisdictional chaos awaiting in major accident cases.

* * -X- * -X* *

Suppose the other states do adopt it? If all of them can then lay claim to being the ‘situs’ of the policy to the same extent that New York makes that claim under Seider, it is not difficult to contemplate a race-to-attach (or seize or sequester or secure ‘trustee process’ or whatever else a state may call the quasi in rem process that goes under the *122name of ‘attachment’ in New York) in multi-party accidents.” 5

The rule of Harris, which in effect sanctions the mere presence of the garnishee as a sufficient jurisdictional base, “cannot be justified in terms of fairness.” Developments in the Law-State Court Jurisdiction, 73 Harv.L.Rev. 909, 960 (1960). In other words, the relevant contact with the forum state which establishes a jurisdictional base is not attuned to the modern jurisdictional search to attain substantial justice by balancing interests.

The majority opinion is quite correct in observing that the Harris rule imposed a substantial burden on Mr. Balk in forcing him to litigate, if he should elect to contest, wherever his debtor’s pereginations might take him. It does not follow that if Balk was treated unfairly, then appellants can be treated with a much greater degree of unfairness. The unstated assumption that Harris controls the case at bar should be rejected. In view of the age of that case, the revolution in jurisdictional thinking since 1905 and the criticism of its potential unfairness in modern terms,6 the rule of that case should be confined to its facts, and applied only in cases where there is a fixed obligation, when the amount may be determined with fair certainty and any contingency concerns nothing more than time of maturity.

Exception must also be taken to the majority’s view that the availability of removal under § 1404(a) “bears strongly” on the decision of the constitutional issue. The constitutionality of this procedure should not rest on the dubious assumption that discretion will always be exercised as the Constitution demands. In addition, 28 U.S.C. § 1441(b) provides that where diversity is the claimed basis of federal jurisdiction, a civil action

“ * * * shall be removable only if none of the parties in interest properly joined and served as defendants is a citizen of the State in which such action is brought.”

In other words, if any of the defendants in a multi-party action is a New York resident, the non-resident co-defendants would be barred from the benefit of § 1404(a). (The majority opinion conveniently notes other “exceptions” in footnote 4.) Nor should the constitutionality of this procedure turn on whether the matter in controversy exceeds $10,000.

For these reasons and for those set forth in the prior dissenting opinion I reassert my disagreement with the majority decision.