JUSTICE MARSHALL delivered the opinion of the Court.
We granted certiorari in this case to determine whether the Parental Kidnaping Prevention Act of 1980, 28 U. S. C. § 1738A, furnishes an implied cause of action in federal court to determine which of two conflicting state custody decisions is valid.
The Parental Kidnaping Prevention Act (PKPA or Act) imposes a duty on the States to enforce a child custody determination entered by a court of a sister State if the determination
is consistent with the provisions of the Act.1 In order for a state court's custody decree to be consistent with the provisions of the Act, the State must have jurisdiction under its own local law and one of five conditions set out in
§ 1738A(c)(2) must be met. Briefly put, these conditions authorize the state court to enter a custody decree if the child's home is or recently has been in the State, if the child has no home State and it would be in the child's best interest for the State to assume jurisdiction, or if the child is present in the State and has been abandoned or abused. Once a State exercises jurisdiction consistently with the provisions of the Act, no other State may exercise concurrent jurisdiction over the custody dispute, § 1738A(g), even if it would have been empowered to take jurisdiction in the first instance,2 and all States must accord full faith and credit to the first State's ensuing custody decree.
As the legislative scheme suggests, and as Congress explicitly specified, one of the chief purposes of the PKPA is to "avoid jurisdictional competition and conflict between State courts." Pub. L. 96-611, 94 Stat. 3569, § 7(c)(5), note following 28 U. S. C. § 1738A. This case arises out of a jurisdictional stalemate that came to pass notwithstanding the strictures of the Act. In July 1978, respondent Susan Clay (then Susan Thompson) filed a petition in Los Angeles Superior Court asking the court to dissolve her marriage to petitioner David Thompson and seeking custody of the couple's infant
son, Matthew. The court initially awarded the parents joint custody of Matthew, but that arrangement became infeasible when respondent decided to move from California to Louisiana to take a job. The court then entered an order providing that respondent would have sole custody of Matthew once she left for Louisiana. This state of affairs was to remain in effect until the court investigator submitted a report on custody, after which the court intended to make a more studied custody determination. See App. 6.
Respondent and Matthew moved to Louisiana in December 1980. Three months later, respondent filed a petition in Louisiana state court for enforcement of the California custody decree, judgment of custody, and modification of petitioner's visitation privileges. By order dated April 7, 1981, the Louisiana court granted the petition and awarded sole custody of Matthew to respondent. Two months later, however, the California court, having received and reviewed its investigator's report, entered an order awarding sole custody of Matthew to petitioner. Thus arose the current impasse.
In August 1983, petitioner brought this action in the District Court for the Central District of California. Petitioner requested an order declaring the Louisiana decree invalid and the California decree valid, and enjoining the enforcement of the Louisiana decree. Petitioner did not attempt to enforce the California decree in a Louisiana state court before he filed suit in federal court. The District Court granted respondent's motion to dismiss the complaint for lack of subject-matter and personal jurisdiction. Civ. Action No. 83-5221 (Apr. 10, 1984). The Court of Appeals for the Ninth Circuit affirmed. Although it disagreed with the District Court's jurisdictional analyses, the Court of Appeals affirmed the dismissal of the complaint on the ground that petitioner had failed to state a claim upon which relief could be granted. 798 F.2d 1547 (1986). Canvassing the background, language, and legislative history of the PKPA, the Court of Appeals held that the Act does not create
a private right of action in federal court to determine the validity of two conflicting custody decrees. Id., at 1552-1559. We granted certiorari, 479 U.S. 1063 (1987), and we now affirm.
In determining whether to infer a private cause of action from a federal statute, our focal point is Congress' intent in enacting the statute. As guides to discerning that intent, we have relied on the four factors set out in Cort v. Ash, 422 U.S. 66, 78 (1975), along with other tools of statutory construction. See Daily Income Fund, Inc. v. Fox, 464 U.S. 523, 535-536 (1984); California v. Sierra Club, 451 U.S. 287, 293 (1981); Touche Ross & Co. v. Redington, 442 U.S. 560, 575-576 (1979). Our focus on congressional intent does not mean that we require evidence that Members of Congress, in enacting the statute, actually had in mind the creation of a private cause of action. The implied cause of action doctrine would be a virtual dead letter were it limited to correcting drafting errors when Congress simply forgot to codify its evident intention to provide a cause of action. Rather, as an implied cause of action doctrine suggests, "the legislative history of a statute that does not expressly create or deny a private remedy will typically be equally silent or ambiguous on the question." Cannon v. University of Chicago, 441 U.S. 677, 694 (1979). We therefore have recognized that Congress' "intent may appear implicitly in the language or structure of the statute, or in the circumstances of its enactment." Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U.S. 11, 18 (1979). The intent of Congress remains the ultimate issue, however, and "unless this congressional intent can be inferred from the language of the statute, the statutory structure, or some other source, the essential predicate for implication of a private remedy simply does not exist." Northwest Airlines, Inc. v. Transport Workers, 451 U.S. 77, 94 (1981). In this case, the essential predicate for implication of a private remedy plainly does not exist. None of the factors
that have guided our inquiry in this difficult area points in favor of inferring a private cause of action. Indeed, the context, language, and legislative history of the PKPA all point sharply away from the remedy petitioner urges us to infer.
We examine initially the context of the PKPA with an eye toward determining Congress' perception of the law that it was shaping or reshaping. See Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U.S. 353, 378 (1982); Cort v. Ash, supra, at 69. At the time Congress passed the PKPA, custody orders held a peculiar status under the full faith and credit doctrine, which requires each State to give effect to the judicial proceedings of other States, see U.S. Const., Art. IV, § 1; 28 U. S. C. § 1738. The anomaly traces to the fact that custody orders characteristically are subject to modification as required by the best interests of the child. As a consequence, some courts doubted whether custody orders were sufficiently "final" to trigger full faith and credit requirements, see, e. g., Hooks v. Hooks, 771 F.2d 935, 948 (CA6 1985); McDougald v. Jenson, 596 F. Supp. 680, 684-685 (ND Fla. 1984), aff'd, 786 F.2d 1465 (CA11), cert. denied, 479 U.S. 860 (1986), and this Court had declined expressly to settle the question. See Ford v. Ford, 371 U.S. 187, 192 (1962). Even if custody orders were subject to full faith and credit requirements, the Full Faith and Credit Clause obliges States only to accord the same force to judgments as would be accorded by the courts of the State in which the judgment was entered. Because courts entering custody orders generally retain the power to modify them, courts in other States were no less entitled to change the terms of custody according to their own views of the child's best interest. See New York ex rel. Halvey v. Halvey, 330 U.S. 610, 614-615 (1947). For these reasons, a parent who lost a custody battle in one State had an incentive to kidnap the child and move to another State to relitigate the issue. This circumstance contributed to widespread jurisdictional deadlocks like this one, and more importantly, to a national epidemic of parental kidnaping.
At the time the PKPA was enacted, sponsors of the Act estimated that between 25,000 and 100,000 children were kidnaped by parents who had been unable to obtain custody in a legal forum. See Parental Kidnaping Prevention Act of 1979: Joint Hearing on S. 105 before the Subcommittee on Criminal Justice of the Judiciary Committee and the Subcommittee on Child and Human Development of the Committee on Labor and Human Resources, 96th Cong., 2d Sess., 10 (1980) (hereinafter PKPA Joint Hearing) (statement of Sen. Malcolm Wallop).
A number of States joined in an effort to avoid these jurisdictional conflicts by adopting the Uniform Child Custody Jurisdiction Act (UCCJA), 9 U. L. A. §§ 1-28 (1979). The UCCJA prescribed uniform standards for deciding which State could make a custody determination and obligated enacting States to enforce the determination made by the State with proper jurisdiction. The project foundered, however, because a number of States refused to enact the UCCJA while others enacted it with modifications. In the absence of uniform national standards for allocating and enforcing custody determinations, non-custodial parents still had reason to snatch their children and petition the courts of any of a number of haven States for sole custody.
The context of the PKPA therefore suggests that the principal problem Congress was seeking to remedy was the inapplicability of full faith and credit requirements to custody determinations. Statements made when the Act was introduced in Congress forcefully confirm that suggestion. The sponsors and supporters of the Act continually indicated that the purpose of the PKPA was to provide for nationwide enforcement of custody orders made in accordance with the terms of the UCCJA. As Acting Deputy Attorney General Michel testified:
"[C]urrent law in many States encourages a parent who does not have custody to snatch the child from the parent who does and take the child to another State to relitigate
the custody issue in a new forum. This kind of 'forum shopping' is possible because child custody orders are subject to modification to conform with changes in circumstances. Consequently, a court deciding a custody case is not, as a Federal constitutional requirement of the full faith and credit clause, bound by a decree by a court of another State even where the action involves the same parties.
"In essence [the PKPA] would impose on States a Federal duty, under enumerated standards derived from the UCCJA, to give full faith and credit to the custody decrees of other States. Such legislation would, in effect, amount to Federal adoption of key provisions of the UCCJA for all States and would eliminate the incentive for one parent to remove a minor child to another jurisdiction." PKPA Joint Hearing 48.3
The significance of Congress' full faith and credit approach to the problem of child snatching is that the Full Faith and Credit Clause, in either its constitutional or statutory incarnations, does not give rise to an implied federal cause of action. Minnesota v. Northern Securities Co., 194 U.S. 48, 72 (1904); see 13B C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 3563, p. 50 (1984). Rather, the Clause "only prescribes a rule by which courts, Federal and state, are to be guided when a question arises in the
progress of a pending suit as to the faith and credit to be given by the court to the public acts, records, and judicial proceedings of a State other than that in which the court is sitting." Northern Securities, supra, at 72. Because Congress' chief aim in enacting the PKPA was to extend the requirements of the Full Faith and Credit Clause to custody determinations, the Act is most naturally construed to furnish a rule of decision for courts to use in adjudicating custody disputes and not to create an entirely new cause of action. It thus is not compatible with the purpose and context of the legislative scheme to infer a private cause of action. See Cort v. Ash, 422 U.S., at 78.
The language and placement of the statute reinforce this conclusion. The PKPA, 28 U. S. C. § 1738A, is an addendum to the full faith and credit statute, 28 U. S. C. § 1738. This fact alone is strong proof that the Act is intended to have the same operative effect as the full faith and credit statute. Similarly instructive is the heading to the PKPA: "Full faith and credit given to child custody determinations." As for the language of the Act, it is addressed entirely to States and state courts. Unlike statutes that explicitly confer a right on a specified class of persons, the PKPA is a mandate directed to state courts to respect the custody decrees of sister States. See Cannon v. University of Chicago, 441 U.S., at 690, n. 13; Cort v. Ash, supra, at 81-82. We agree with the Court of Appeals that "[i]t seems highly unlikely Congress would follow the pattern of the Full Faith and Credit Clause and section 1738 by structuring section 1738A as a command to state courts to give full faith and credit to the child custody decrees of other states, and yet, without comment, depart from the enforcement practice followed under the Clause and section 1738." 798 F.2d, at 1556.
Finally, the legislative history of the PKPA provides unusually clear indication that Congress did not intend the federal
courts to play the enforcement role that petitioner urges. Two passages are particularly revealing. The first of these is a colloquy between Congressmen Conyers and Fish. Congressman Fish had been the sponsor of a competing legislative proposal -- ultimately rejected by Congress -- that would have extended the district courts' diversity jurisdiction to encompass actions for enforcement of state custody orders. In the following exchange, Congressman Conyers questioned Congressman Fish about the differences between his proposal and "the Bennett proposal," which was a precursor to the PKPA.
"Mr. Conyers: Could I just interject, the difference between the Bennett proposal and yours: You would have, enforcing the full faith and credit provision, the parties removed to a Federal court. Under the Bennett provision, his bill would impose the full faith and credit enforcement on the State court.
"It seems to me that that is a very important difference. The Federal jurisdiction, could it not, Mr. Fish, result in the Federal court litigating between two State court decrees; whereas, in an alternate method previously suggested, we would be imposing the responsibility of the enforcement upon the State court, and thereby reducing, it seems to me, the amount of litigation.
"Do you see any possible merit in leaving the enforcement at the State level, rather than introducing the Federal judiciary?
"Mr. Fish: Well, I really think that it is easier on the parent that has custody of the child to go to the nearest Federal district court . . . .
"Mr. Conyers: Of course you know that the Federal courts have no experience in these kinds of matters, and they would be moving into this other area. I am just thinking of the fact that they have [many areas of federal concern and] on the average of a 21-month docket, you
would now be imposing custody matters which it seems might be handled in the courts that normally handle that . . . ." Parental Kidnaping: Hearing on H. R. 1290 before the Subcommittee on Crime of the House Committee on the Judiciary, 96th Cong., 2d Sess., 14 (1980).
This exchange suggests that Congress considered and rejected an approach to the problem that would have resulted in a "Federal court litigating between two State court decrees." Ibid.
The second noteworthy entry in the legislative history is a letter from then Assistant Attorney General Patricia Wald to the Chairman of the House Judiciary Committee, which was referred to extensively during the debate on the PKPA. The letter outlined a variety of solutions to the child-snatching problem. It specifically compared proposals that would "grant jurisdiction to the federal courts to enforce state custody decrees" with an approach, such as was proposed in the PKPA, that would "impose on states a federal duty, under enumerated standards derived generally from the UCCJA, to give full faith and credit to the custody decrees of other states." Addendum to Joint Hearing 103. The letter endorsed the full faith and credit approach that eventually was codified in the PKPA. More importantly, it "strongly oppose[d] . . . the creation of a federal forum for resolving custody disputes." Id., at 108. Like Congressman Conyers, the Justice Department reasoned that federal enforcement of state custody decrees would increase the workload of the federal courts and entangle the federal judiciary in domestic relations disputes with which they have little experience and which traditionally have been the province of the States. That the views of the Justice Department and Congressman Conyers prevailed, and that Congress explicitly opted for a full faith and credit approach over reliance on enforcement by the federal courts, provide strong evidence against inferring a federal cause of action. Cf. Cort v. Ash,
U.S., at 82 (congressional determination not to create a private cause of action is dispositive).
Petitioner discounts these portions of the legislative history. He argues that the cause of action that he asks us to infer arises only in cases of an actual conflict between two state custody decrees, and thus is substantially narrower than the cause of action proposed by Congressman Fish and rejected by Congress. The Fish bill would have extended federal diversity jurisdiction to permit federal courts to enforce custody orders in the first instance, before a second State had created a conflict by refusing to do so. This cause of action admittedly is farther reaching than that which we reject today. But the considerations that prompted Congress to reject the Fish bill also militate against the more circumscribed role for the federal courts that petitioner proposes. See Rogers v. Platt, 259 U.S. App. D.C. 154, 164, 814 F.2d 683, 693 (1987). Instructing the federal courts to play Solomon where two state courts have issued conflicting custody orders would entangle them in traditional state-law questions that they have little expertise to resolve.4 This is
a cost that Congress made clear it did not want the PKPA to carry.5
In sum, the context, language, and history of the PKPA together make out a conclusive case against inferring a cause of action in federal court to determine which of two conflicting state custody decrees is valid. Against this impressive evidence, petitioner relies primarily on the argument that failure to infer a cause of action would render the PKPA nugatory. We note, as a preliminary response, that ultimate review remains available in this Court for truly intractable jurisdictional deadlocks. In addition, the unspoken presumption in petitioner's argument is that the States are either unable or unwilling to enforce the provisions of the Act. This is a presumption we are not prepared, and more importantly, Congress was not prepared, to indulge. State courts faithfully administer the Full Faith and Credit Clause every day; now that Congress has extended full faith and credit requirements to child custody orders, we can think of no reason why the courts' administration of federal law in custody disputes will be any less vigilant. Should state courts prove as obstinate as petitioner predicts, Congress may choose to revisit the issue. But any more radical approach to the problem will have to await further legislative action; we "will not engraft a remedy on a statute, no matter how salutary, that Congress did not intend to provide." California v. Sierra Club, 451 U.S. 287, 297 (1981). The judgment of the Court of Appeals is affirmed.
It is so ordered.
798 F.2d 1547, affirmed.
JUSTICE O'CONNOR, concurring in part and concurring in the judgment.
For the reasons expressed by JUSTICE SCALIA in Part I of his opinion in this case, I join all but the first full paragraph of Part II of the Court's opinion and judgment.
JUSTICE SCALIA, concurring in the judgment.
I write separately because in my view the Court is not being faithful to current doctrine in its dicta denying the necessity of an actual congressional intent to create a private right of action, and in referring to Cort v. Ash, 422 U.S. 66 (1975), as though its analysis had not been effectively overruled by our later opinions. I take the opportunity to suggest, at the same time, why in my view the law revision that the Court's dicta would undertake moves in precisely the wrong direction.
I agree that the Parental Kidnaping Prevention Act, 28 U. S. C. § 1738A, does not create a private right of action in federal court to determine which of two conflicting child custody decrees is valid. I disagree, however, with the portion of the Court's analysis that flows from the following statement:
"Our focus on congressional intent does not mean that we require evidence that Members of Congress, in enacting the statute, actually had in mind the creation of a private cause of action." Ante, at 179.
I am at a loss to imagine what congressional intent to create a private right of action might mean, if it does not mean that Congress had in mind the creation of a private right of action. Our precedents, moreover, give no indication of a secret meaning, but to the contrary seem to use "intent" to mean "intent." For example:
"[T]he focus of the inquiry is on whether Congress intended to create a remedy. Universities Research Assn., Inc. v. Coutu, 450 U.S., at 771-772; Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U.S. at 23-24; Touche Ross & Co. v. Redington, [442 U.S.], at 575-576. The federal judiciary will not engraft a remedy on a statute, no matter how salutary, that Congress did not intend to provide." California v. Sierra Club, 451 U.S. 287, 297 (1981) (WHITE, J.).
We have said, to be sure, that the existence of intent may be inferred from various indicia; but that is worlds apart from today's Delphic pronouncement that intent is required but need not really exist.
I also find misleading the Court's statement that, in determining the existence of a private right of action, "we have relied on the four factors set out in Cort v. Ash, . . . along with other tools of statutory construction." Ante, at 179. That is not an accurate description of what we have done. It could not be plainer that we effectively overruled the Cort v. Ash analysis in Touche Ross & Co. v. Redington, 442 U.S. 560, 575-576 (1979), and Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U.S. 11, 18 (1979), converting one of its four factors (congressional intent) into the determinative factor, with the other three merely indicative of its presence or absence. Compare Cort v. Ash, supra, at 78, with Transamerica, supra, at 23-24.
Finally, the Court's opinion conveys a misleading impression of current law when it proceeds to examine the "context" of the legislation for indication of intent to create a private right of action, after having found no such indication in either text or legislative history. In my view that examination is entirely superfluous, since context alone cannot suffice. We have held context to be relevant to our determination in only two cases -- both of which involved statutory language that, in the judicial interpretation of related legislation prior to the subject statute's enactment, or of the same legislation prior
to its reenactment, had been held to create private rights of action. See Cannon v. University of Chicago, 441 U.S. 677 (1979); Merrill Lynch, Pierce, Fenner & Smith, Inc. v. Curran, 456 U.S. 353 (1982). Since this is not a case where such textual support exists, or even where there is any support in legislative history, the "context" of the enactment is immaterial.
Contrary to what the language of today's opinion suggests, this Court has long since abandoned its hospitable attitude towards implied rights of action. In the 23 years since Justice Clark's opinion for the court in J. I. Case Co. v. Borak, 377 U.S. 426 (1964), we have twice narrowed the test for implying a private right, first in Cort v. Ash, supra, itself, and then again in Touche Ross & Co. v. Redington, supra, and Transamerica Mortgage Advisers, Inc. v. Lewis, supra. See also Cannon v. University of Chicago, supra, at 730 (Powell, J., dissenting), and California v. Sierra Club, supra, at 301 (REHNQUIST, J., joined by Burger, C. J., and Stewart and Powell, JJ., concurring). The recent history of our holdings is one of repeated rejection of claims of an implied right. This has been true in 9 of 11 recent private right of action cases heard by this Court, including the instant case. See Touche Ross, supra; Transamerica, supra; Universities Research Assn., Inc. v. Coutu, 450 U.S. 754 (1981); Northwest Airlines, Inc. v. Transport Workers, 451 U.S. 77, 91-94 (1981); California v. Sierra Club, supra; Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 639-640 (1981); Middlesex County Sewerage Authority v. National Sea Clammers Assn., 453 U.S. 1, 13-18 (1981); Daily Income Fund, Inc. v. Fox, 464 U.S. 523, 535-536 (1984); and Massachusetts Mut. Life Ins. Co. v. Russell, 473 U.S. 134, 145-148 (1985). But see Merrill Lynch, supra, and Cannon, supra. The Court's opinion exaggerates the difficulty of establishing an implied right when it surmises that "[t]he implied cause of action doctrine would be a virtual dead letter were it limited to correcting drafting errors when Congress simply forgot to codify its evident intention to provide a
cause of action." Ante, at 179. That statement rests upon the erroneous premise that one never implies anything except when he forgets to say it expressly. It is true, however, that the congressional intent test for implying private rights of action as it has evolved since the repudiation of Cort v. Ash is much more stringent than the Court's dicta in the present case suggest.
I have found the Court's dicta in the present case particularly provocative of response because it is my view that, if the current state of the law were to be changed, it should be moved in precisely the opposite direction -- away from our current congressional intent test to the categorical position that federal private rights of action will not be implied.
As Justice Powell observed in his dissent in Cannon, supra, at 730-731:
"Under Art. III, Congress alone has the responsibility for determining the jurisdiction of the lower federal courts. As the Legislative Branch, Congress also should determine when private parties are to be given causes of action under legislation it adopts. As countless statutes demonstrate, including Titles of the Civil Rights Act of 1964, Congress recognizes that the creation of private actions is a legislative function and frequently exercises it. When Congress chooses not to provide a private civil remedy, federal courts should not assume the legislative role of creating such a remedy and thereby enlarge their jurisdiction." (Footnote omitted.)
It is, to be sure, not beyond imagination that in a particular case Congress may intend to create a private right of action, but chooses to do so by implication. One must wonder, however, whether the good produced by a judicial rule that accommodates this remote possibility is outweighed by its adverse effects. An enactment by implication cannot realistically be regarded as the product of the difficult lawmaking process our Constitution has prescribed. Committee reports,
floor speeches, and even colloquies between Congressmen, ante, at 184-185, are frail substitutes for bicameral vote upon the text of a law and its presentment to the President. See generally INS v. Chadha, 462 U.S. 919 (1983). It is at best dangerous to assume that all the necessary participants in the law-enactment process are acting upon the same unexpressed assumptions. And likewise dangerous to assume that, even with the utmost self-discipline, judges can prevent the implications they see from mirroring the policies they favor.
I suppose all this could be said, to a greater or lesser degree, of all implications that courts derive from statutory language, which are assuredly numerous as the stars. But as the likelihood that Congress would leave the matter to implication decreases, so does the justification for bearing the risk of distorting the constitutional process. A legislative act so significant, and so separable from the remainder of the statute, as the creation of a private right of action seems to me so implausibly left to implication that the risk should not be endured.
If we were to announce a flat rule that private rights of action will not be implied in statutes hereafter enacted, the risk that that course would occasionally frustrate genuine legislative intent would decrease from its current level of minimal to virtually zero. It would then be true that the opportunity for frustration of intent "would be a virtual dead letter[,] . . . limited to . . . drafting errors when Congress simply forgot to codify its . . . intention to provide a cause of action." Ante, at 179. I believe, moreover, that Congress would welcome the certainty that such a rule would produce. Surely conscientious legislators cannot relish the current situation, in which the existence or nonexistence of a private right of action depends upon which of the opposing legislative forces may have guessed right as to the implications the statute will be found to contain.
If a change is to be made, we should get out of the business of implied private rights of action altogether.
* Briefs of amici curiae urging reversal were filed for Sacramento County et al. by L. B. Elam, Steven M. Basha, Mark A. Wasser, Ann M. Haralambie, and DeWitt W. Clinton; and for the Women's Legal Defense Fund et al. by Edward R. Leahy and Steven A. Fennell.
1. Section 1738A reads in relevant part: "(a) The appropriate authorities of every State shall enforce according to its terms, and shall not modify except as provided in subsection (f) of this section, any child custody determination made consistently with the provisions of this section by a court of another State. . . . . . "(c) A child custody determination made by a court of a State is consistent with the provisions of this section only if -- "(1) such court has jurisdiction under the law of such state; and "(2) one of the following conditions is met: "(A) such State (i) is the home State of the child on the date of the commencement of the proceeding, or (ii) had been the child's home State within six months before the date of the commencement of the proceeding and the child is absent from such State because of his removal or retention by a contestant or for other reasons, and a contestant continues to live in such State; "(B)(i) it appears that no other State would have jurisdiction under subparagraph (A), and (ii) it is in the best interest of the child that a court of such State assume jurisdiction because (I) the child and his parents, or the child and at least one contestant, have a significant connection with such State other than mere physical presence in such State, and (II) there is available in such State substantial evidence concerning the child's present or future care, protection, training, and personal relationships; "(C) the child is physically present in such State and (i) the child has been abandoned, or (ii) it is necessary in an emergency to protect the child because he has been subjected to or threatened with mistreatment or abuse; "(D)(i) it appears that no other State would have jurisdiction under subparagraph (A), (B), (C), or (E), or another State has declined to exercise jurisdiction on the ground that the State whose jurisdiction is in issue is the more appropriate forum to determine the custody of the child, and (ii) it is in the best interest of the child that such court assume jurisdiction; or "(E) the court has continuing jurisdiction pursuant to subsection (d) of this section. "(d) The jurisdiction of a court of a State which has made a child custody determination consistently with the provisions of this section continues as long as the requirement of subsection (c)(1) of this section continues to be met and such State remains the residence of the child or of any contestant. . . . . . "(f) A court of a State may modify a determination of the custody of the same child made by a court of another State, if -- "(1) it has jurisdiction to make such a child custody determination; and "(2) the court of the other State no longer has jurisdiction, or it has declined to exercise such jurisdiction to modify such determination. "(g) A court of a State shall not exercise jurisdiction in any proceeding for a custody determination commenced during the pendency of a proceeding in a court of another State where such court of that other State is exercising jurisdiction consistently with the provisions of this section to make a custody determination."
2. The sole exception to this constraint occurs where the first State either has lost jurisdiction or has declined to exercise continuing jurisdiction. See § 1738A(f).
3. Mr. Michel's remarks are echoed in numerous other portions of the legislative history. See, e. g., PKPA Joint Hearing 40-41 (statement of Sen. Durenberger); id., at 12 (statement of Sen. Wallop); PKPA: Addendum to Joint Hearing on S. 105 before the Subcommittee on Criminal Justice of the Committee on the Judiciary and the Subcommittee on Child and Human Development of the Committee on Labor and Human Resources, 96th Cong., 2d Sess., 104-105 (1980) (hereinafter Addendum to Joint Hearing) (letter from Assistant Attorney General Patricia Wald to Rep. Peter Rodino).
4. Petitioner argues that determining which of two conflicting custody decrees should be given effect under the PKPA would not require the federal courts to resolve the merits of custody disputes and thus would not offend the longstanding tradition of reserving domestic relations matters to the States. Petitioner contends that the cause of action he champions would require federal courts only to analyze which of two States is given exclusive jurisdiction under a federal statute, a task for which the federal courts are well qualified. We cannot agree with petitioner that making a jurisdictional determination under the PKPA would not involve the federal courts in substantive domestic relations determinations. Under the Act, jurisdiction can turn on the child's "best interest" or on proof that the child has been abandoned or abused. See §§ 1738A (c)(2)(B), (C), and (D). In fact, it would seem that the jurisdictional disputes that are sufficiently complicated as to have provoked conflicting state-court holdings are the most likely to require resolution of these traditional domestic relations inquiries. See Rogers v. Platt, 259 U.S. App. D.C. 154, 162, 814 F.2d 683, 691 (1987). Cf. Cort v. Ash, 422 U.S. 66, 84 (1975) (possibility that implied federal cause of action may in certain instances turn on state-law issues counsels against inferring such an action.)
5. Moreover, petitioner's argument serves to underscore the extraordinary nature of the cause of action he urges us to infer. Petitioner essentially asks that federal district courts exercise appellate review of state-court judgments. This is an unusual cause of action for Congress to grant, either expressly or by implication. Petitioner's proposal is all the more remarkable in the present case, in which he seeks to have a Federal District Court in California enjoin enforcement of a Louisiana state-court judgment before the intermediate and highest appellate courts of Louisiana even have had an opportunity to review that judgment.