Victims of the Hungarian Holoc v. Magyar Allamvasutak Zrt

In the United States Court of Appeals For the Seventh Circuit ____________________ No. 13-3073 PAUL CHAIM SHLOMO FISCHER, et al., Plaintiffs-Appellants, v. MAGYAR ÁLLAMVASUTAK ZRT., Defendant-Appellee. ____________________ Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:10-cv-00868—Samuel Der-Yeghiayan, Judge. ____________________ No. 14-1319 MICHAEL J. ALBERT, et al., Plaintiffs-Appellants, v. MAGYAR NEMZETI BANK and ERSTE GROUP BANK, Defendants-Appellees. ____________________ Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:10-cv-01884—Samuel Der-Yeghiayan, Judge. ____________________ 2 Nos. 13-3073 & 14-1319 ARGUED SEPTEMBER 30, 2014 — DECIDED JANUARY 23, 2015 ____________________ Before KANNE, WILLIAMS, and HAMILTON, Circuit Judges. HAMILTON, Circuit Judge. These appeals arise from suits brought by Holocaust survivors and the heirs of other Holo- caust victims against the Hungarian national railway, the Hungarian national bank, and several private banks for the roles they or their predecessors played in carrying out geno- cide against Hungarian Jews during World War II. These claims for takings of property arise from events in Hungary 70 years ago. They are asserted against both foreign sover- eign entities and private banks with relatively few ties to the United States. The cases bring to the United States courts as- pects of the horrific crimes of the Holocaust, but the cases have also posed difficult questions about whether they might be heard in a United States court. In the earlier 2012 appeals in these cases, we held that the national railway and national bank—both instrumentalities of the Hungarian government—could be sued on these claims in a United States federal court, but only if the plain- tiffs could demonstrate on remand that they had exhausted any avail-able Hungarian remedies or had a legally compel- ling reason for their failure to do so. Abelesz v. Magyar Nemzeti Bank, 692 F.3d 661 (7th Cir. 2012). In addition, while we mandated dismissal of claims against two private banks for lack of personal jurisdiction, Abelesz v. OTP Bank, 692 F.3d 638 (7th Cir. 2012), we denied interlocutory requests by Erste Group Bank AG (“Erste Bank”), a private Austrian bank that had acquired a Hungarian bank that plaintiffs alleged had participated in the Holocaust, to review the denial of its mo- Nos. 13-3073 & 14-1319 3 tion to dismiss on several grounds. Abelesz v. Erste Group Bank AG, 695 F.3d 655 (7th Cir. 2012). On remand, the national bank, national railway, and Erste Bank all continued to seek dismissal. As to the national bank and railway, the district court held that the plaintiffs had not exhausted Hungarian remedies and had not provid- ed a legally compelling reason for not doing so. For that rea- son, the district court concluded that it could no longer en- tertain plaintiffs’ international law claims and dismissed the claims against the national defendants. Following that dismissal, the district court also dismissed Erste Bank from the suit on forum non conveniens grounds. Al-though the district court had previously denied Erste Bank’s motion to dismiss based on forum non conveniens, it took a fresh look at the issue once the national defendants were not subject to suit in the United States. The court con- cluded that dismissal on these grounds was appropriate. Plaintiffs have appealed. We affirm all the dismissals. First, as we held in 2012, in- ternational law does not require exhaustion of domestic remedies before plaintiffs can say that international law was violated. But principles of international comity make clear that these plaintiffs must attempt to exhaust domestic reme- dies before foreign courts can provide remedies for those vi- olations. These plaintiffs have not exhausted available Hun- garian remedies, and the district court did not abuse its dis- cretion when it found that plaintiffs should not be excused from doing so. In addition, because the national bank was properly dismissed from the case against the banks, the dis- trict court properly granted Erste Bank’s motion to reconsid- er dismissal for forum non conveniens. Nevertheless, while the 4 Nos. 13-3073 & 14-1319 doors of United States courts are closed to these claims for now, they are not locked forever. All dismissals are without prejudice. If plaintiffs find that future attempts to pursue remedies in Hungary are frustrated unreasonably or arbi- trarily, a United States court could once again hear these claims. I. Factual and Procedural Background A. Overview of Claims We summarize briefly the more detailed account of the facts from our 2012 Abelesz opinions. Plaintiffs’ complaints describe the seizure, transport, and murder of hundreds of thousands of Hungarian Jews during the Holocaust, particu- larly during 1944 and 1945. The Foreign Sovereign Immuni- ties Act bars jurisdiction in United States federal courts against foreign sovereigns for claims for death or personal injury or damage to or loss of property that does not occur in the United States. See Abelesz v. Magyar Nemzeti Bank, 692 F.3d at 677, citing 28 U.S.C. § 1605(a)(5). Nevertheless, the FSIA permits jurisdiction over foreign sovereigns for claims for takings of property in violation of international law. 28 U.S.C. § 1605(a)(3). As a result, plaintiffs’ claims focus on the role that the Hungarian national railway and Hungarian banks played in expropriating money and other property from Jews. Plaintiffs allege that these expropriations were essential to finance the continued German war effort and even the Hungarian genocide itself. See Abelesz, 692 F.3d at 675. Plaintiffs brought two separate suits: one against a group of Hungarian banks (along with an Austrian bank) and an- other against the Hungarian national railway. Against the Nos. 13-3073 & 14-1319 5 banks, plaintiffs allege seven causes of action: genocide, aid- ing and abetting genocide, bailment, conversion, unjust en- richment, constructive trust, and accounting. Plaintiffs allege nine causes of action against the national railway: takings in violation of international law, aiding and abetting genocide, complicity in genocide, violations of customary international law, unlawful conversion, unjust enrichment, fraudulent misrepresentations, accounting, and declaratory relief pur- suant to 28 U.S.C. § 2201. Plaintiffs seek to hold the national and private banks jointly and severally responsible for damages of approxi- mately $75 billion. They seek damages of approximately $1.25 billion from the national railway. Plaintiffs rely on sev- eral bases of jurisdiction, including the Foreign Sovereign Immunities Act (FSIA), 28 U.S.C. § 1330(a), the Alien Tort Statute, 28 U.S.C. § 1350, diversity jurisdiction under the Class Action Fairness Act (CAFA), 28 U.S.C. § 1332(d), and federal question jurisdiction, 28 U.S.C. § 1331. B. The Prior Appeals In 2012 we considered several appeals and mandamus petitions seeking review of the district court’s denial of de- fendants’ motions to dismiss on multiple grounds. In Abelesz v. Magyar Nemzeti Bank, 692 F.3d 661 (7th Cir. 2012), we con- sidered whether and under what circumstances the district court could exercise subject matter jurisdiction over the two instrumentalities of Hungary—the national bank and na- tional railway. We held that the district court could exercise jurisdiction under the expropriation exception to the FSIA, but only if plaintiffs could demonstrate on remand that they either exhausted available Hungarian remedies or could show a legally compelling reason for not doing so. Id. at 684. 6 Nos. 13-3073 & 14-1319 Because exhaustion is also at the center of these appeals, we repeat our earlier conclusions. The national bank and national railway of Hungary are instrumentalities of a foreign sovereign under the FSIA. See 28 U.S.C. § 1603(b). Accordingly, the FSIA is the exclusive basis for exercising jurisdiction over those entities in United States courts. Argentine Republic v. Amerada Hess Shipping Corp., 488 U.S. 428, 434–36 (1989). Under the FSIA, foreign sovereigns and their instrumentalities are immune from suit in United States courts unless a specific statutory exception applies. 28 U.S.C. § 1604. Plaintiffs argued that two FSIA exceptions might allow jurisdiction over the national bank: the waiver exception in § 1605(a)(1) and the expropriation exception in § 1605(a)(3). We rejected the waiver exception. While the Hungarian con- stitution recognized international law norms, it did not go so far as to waive sovereign immunity for those claims. Abelesz, 692 F.3d at 670–71. The expropriation exception presented a closer and more complex question. We explained that the expropriation ex- ception defeats sovereign immunity only where “(1) rights in property are in issue; (2) the property was taken; (3) the tak- ing was in violation of international law; and (4) at least one of the two nexus requirements is satisfied.” Id. at 671, citing Zappia Middle East Constr. Co. v. Emirate of Abu Dhabi, 215 F.3d 247, 251 (2d Cir. 2000). We held that “plaintiffs have suf- ficiently alleged that rights in property are at issue, that their property was taken, and that the national bank meets the nexus requirement.” Id. at 695. It was less clear that the na- tional railway met either of the nexus requirements; accord- Nos. 13-3073 & 14-1319 7 ingly, we remanded for jurisdictional discovery on whether the railway meets the nexus requirements. Id. That left the most important and complex problem: whether plaintiffs alleged expropriations that could have vi- olated international law. Abelesz, 692 F.3d at 673. We rejected defendants’ federal preemption argument, id. at 677–78, and made clear that because plaintiffs based their claims upon violations of customary international law, they had actiona- ble rights, id. at 685–86. At the same time, because “a sover- eign could expropriate the property of its own nationals within its own territory without violating international law,” id. at 674, the national bank and railway argued that the al- leged expropriations taking place during the Holocaust did not violate international law and would not be justiciable under the FSIA. We recognized that courts should tread carefully in this field of property expropriation: “Actions that might appear to one regime or nation as unfair expropria- tions might seem to another to be a just remedy for decades or more of exploitation of the poor and downtrodden.” Id. at 675. Nevertheless, we held that the domestic takings rule did not apply where the expropriations funded the transport and murder of a country’s nationals in a campaign of geno- cide, which also sought to leave any survivors of that geno- cide impoverished. Id. The national bank and railway defendants also argued that either the FSIA itself or international law norms re- quired exhaustion of domestic remedies before plaintiffs could assert a violation in a United States court. Id. at 678. We rejected the statutory exhaustion argument, finding that nothing in the language of the FSIA expropriation exception suggests that plaintiffs must exhaust domestic remedies be- 8 Nos. 13-3073 & 14-1319 fore resorting to United States courts. Id., citing § 1605(a)(3). In so doing, we joined the Ninth and D.C. Circuits. See id., citing Cassirer v. Kingdom of Spain, 616 F.3d 1019, 1034–37 (9th Cir. 2010), and Agudas Chasidei Chabad of U.S. v. Russian Fed’n, 528 F.3d 934, 948–49 (D.C. Cir. 2008). Even though § 1605(a)(3) itself does not require exhaus- tion, we went on to conclude that the provision’s reliance on international law norms made clear that plaintiffs would need to exhaust domestic remedies before they could assert a violation of customary international law in a United States court. This exhaustion principle, based on comity, is a well- established rule of customary international law. The Su- preme Court has suggested that customary international law may require exhaustion. See Abelesz, 692 F.3d at 679, citing Sosa v. Alvarez-Machain, 542 U.S. 692, 733 n.21 (2004). This rule has also been invoked in other foreign and domestic sit- uations, including by the United States government itself when defending against takings claims. Id. at 679–80 (collect- ing cases). At bottom, international law favors giving a state accused of taking property in violation of international law an opportunity to “redress it by its own means, within the framework of its own legal system” before the same alleged taking may be aired in foreign courts. Id. at 680. For these reasons, we required plaintiffs “either to pursue and exhaust domestic remedies in Hungary or to show con- vincingly that such remedies are clearly a sham or inade- quate or that their application is unreasonably prolonged.” Id. at 681, citing Restatement (Third) of the Foreign Relations Law of the United States § 713 cmt. f. Keeping in mind that hearing these claims in a United States court “without even giving Hungarian courts an opportunity to address them” Nos. 13-3073 & 14-1319 9 would be an “extraordinary step,” we addressed and found unpersuasive some reasons offered by plaintiffs that the domestic exhaustion rule should not bar their claims. Id. at 684. And although we held that “plaintiffs [had] not present- ed a legally compelling reason for why the domestic exhaus- tion rule does not apply to their claims,” we found it pru- dent to remand the cases and to direct the district court to do “a more detailed examination of this pivotal exhaustion is- sue.” Id. In particular, we directed the defendants to specify the Hungarian remedies that might be available to individuals in plaintiffs’ position. Id. We said that plaintiffs would then have three options on remand: (1) They can voluntarily dismiss their claims against the national bank and national railway without prej- udice and pursue their claims in Hungary using the remedies identified by defendants, with a possibility that they might refile their case in a U.S. court if and when they exhaust their remedies in Hungary. (2) They can ask the district court to stay their cases against the national bank and national railway while they pursue the Hungarian remedies identified by de- fendants. (3) They can ask the district court for an op- portunity to develop further their arguments regard- ing the actual adequacy and availability of those rem- edies and the applicability of the domestic exhaustion rule. Id. In the separate but related appeals by the private banks, we issued a writ of mandamus directing the district court to 10 Nos. 13-3073 & 14-1319 dismiss the claims against two banks for lack of personal ju- risdiction. Abelesz v. OTP Bank, 692 F.3d 638. At the same time, we concluded that Erste Bank—a private Austrian bank over which personal jurisdiction was not disputed— could not immediately appeal the district court’s denial of its motion to dismiss on other grounds and was not entitled to a writ of mandamus because it had not “demonstrated a clear and indisputable right to relief on par” with those of the other two banks. Abelesz v. Erste Group, 695 F.3d at 658. Thus, Erste Bank remained in the bank case. C. Remand On remand, the national bank and railway summarized potential Hungarian remedies for individuals in plaintiffs’ position. Plaintiffs elected not to dismiss voluntarily or to ask for a stay while they pursued those remedies. They sought to show why exhaustion of Hungarian remedies should not be required at all, either because it is not legally required or because the political situation in Hungary, in- cluding rising levels of anti-Semitism, makes exhaustion fu- tile. The district court determined that the plaintiffs had failed to demonstrate that they had exhausted available Hungarian remedies or that there was a legally compelling reason for their failure to do so. Abelesz v. Magyar Nemzeti Bank, No. 10 C 1884, Dkt. 353 (N.D. Ill. Aug. 20, 2013); Fischer v. Magyar Allamvasutak Zrt., No. 10 C 868, 2013 WL 4525408, at *1 (N.D. Ill. Aug. 20, 2013) (railway case). The district court found that plaintiffs had not convincingly shown “that the Hungarian courts are clearly a sham or inadequate, or that such court proceedings will be unreasonably pro- longed.” Id. Accordingly, the district court dismissed the Nos. 13-3073 & 14-1319 11 claims against the national bank and railway without preju- dice. Erste Bank also sought dismissal on two separate grounds. First, following the dismissal of the national bank, Erste Bank sought reconsideration of the district court’s ear- lier decision to deny its motion to dismiss for forum non con- veniens. Second, Erste Bank moved to dismiss for lack of sub- ject matter jurisdiction in light of the Supreme Court’s deci- sion in Kiobel v. Royal Dutch Petroleum Co., 133 S. Ct. 1659 (2013). The district court opted to dismiss the case on forum non conveniens grounds without reaching subject matter ju- risdiction. Plaintiffs appeal both dismissals, and we address each in turn. First, we affirm the district court’s dismissal of claims against the national bank and national railway because plaintiffs have not exhausted apparently available Hungari- an remedies. Second, we affirm dismissal of Erste Bank for forum non conveniens. Like the district court, we do not reach subject matter jurisdiction regarding Erste Bank. II. Foreign Sovereign Immunity and Domestic Remedies Plaintiffs make two principal arguments that the district court erred by dismissing claims against the national bank and railway for failure to exhaust domestic remedies. First, they argue the district court should not have imposed an ex- haustion requirement in the first place. Second, they argue that they offered sufficient reasons to excuse exhaustion in these cases. A. Exhaustion Requirement for Expropriation Claims Plaintiffs argue first, in essence, that the first appeals were wrongly decided in 2012. We rarely consider argu- 12 Nos. 13-3073 & 14-1319 ments seeking to re-litigate issues already decided in earlier appeals. The proper way to argue that this court erred would have been to file a petition for rehearing, either by the panel or en banc. See Morrison v. Duckworth, 898 F.2d 1298, 1299 n.2 (7th Cir. 1990) (invoking law of the case doctrine). A party who is not satisfied with the outcome of the prior ap- peal should not try to reargue the point on remand to a dis- trict court that is required to follow our mandate, nor in a later appeal. At the same time, plaintiffs’ arguments here in- dicate that there may be some lingering ambiguity about the role that exhaustion of domestic remedies plays with respect to the expropriation exception to foreign sovereign immuni- ty. We therefore address the issue and may perhaps clarify the earlier opinion. Plaintiffs argue that we should revisit the exhaustion analysis in Abelesz v. Magyar Nemzeti Bank because we did not consider that the expropriations alleged here were “dis- criminatory.” Plaintiffs rely on § 712(1) of the Restatement (Third) of the Foreign Relations Law of the United States, which says that a state is responsible under international law for injury resulting from a taking by the state of the property of a national of another state that (a) is not for a public pur- pose, or (b) is discriminatory, or (c) is not accompanied by provision for just compensation. Plaintiffs reason that § 712 teaches that a “discriminatory” taking is always a violation of international law, whether or not a domestic remedy of- fers just compensation, so no exhaustion should be required. For non-discriminatory takings, by contrast, the violations arise only when the plaintiffs can demonstrate that the tak- ings occurred without just compensation. Plaintiffs argue that they have alleged discriminatory takings that violated international law regardless of whether Hungarian law has Nos. 13-3073 & 14-1319 13 offered any procedure to obtain compensation, so that they should not be required to exhaust domestic remedies in Hungary. This argument misunderstands the relationship between finding a violation of international law and whether exhaus- tion is required. An exhaustion requirement could serve two distinct roles. On one hand, plaintiffs might be required to exhaust domestic remedies before they can even say that in- ternational law was violated. This might be the case, for ex- ample, when the violation is based on a taking without just compensation. There is no violation until the plaintiff has sought redress and has been unfairly denied just compensa- tion. On the other hand, even if plaintiffs can allege a viola- tion of international law, customary international law may impose an exhaustion requirement that limits plaintiffs’ abil- ity to bring that claim outside the country against which they bring suit. To bring that claim in courts outside of the potentially offending nation—here, Hungary—plaintiffs would need to demonstrate that they exhausted remedies or that it could not be worthwhile to bring suit in that nation. To be clear, our prior decision invoked the second form of exhaustion. The opinion in Abelesz v. Magyar Nemzeti Bank, 692 F.3d 661, cited case law and other sources invoking both sorts of exhaustion, so we understand how someone might read the opinion as implying that international law would not be violated unless plaintiffs could show they exhausted domestic remedies. Understood correctly, however, the prior opinion imposed an exhaustion requirement that limits where plaintiffs may assert their international law claims. We did not hold that plaintiffs failed to allege violations of international law in the first instance. We made clear that the 14 Nos. 13-3073 & 14-1319 question was whether “international law require[s] plaintiffs to exhaust domestic remedies before pursuing expropriation claims elsewhere.” Abelesz, 692 F.3d at 679. We said yes, but we did not ask or answer the question whether plaintiffs needed to exhaust domestic remedies before they might say that international law was violated. Without answering that question, we found that plaintiffs had alleged violations of international law due to the genocidal nature of the expro- priations. Plaintiffs also argue that a separate basis supported find- ing violations of international law: that the takings were “discriminatory” within the meaning of § 712. But this ar- gument misunderstands both the nature of “discrimination” in international expropriation cases and the role of §§ 712 and 713 of the Restatement in our analysis. Section 712 applies, by its terms, only to a state’s takings of property of nationals of other states, not to its takings of property from its own nationals, as alleged in these cases. The discrimination that concerns § 712 is discrimination against aliens, not discrimination among a state’s own na- tionals based on race, religion, ethnicity, or similar grounds, however despicable such discrimination might be. Comment f explains that § 712 refers to discrimination against aliens generally, or against aliens of a particular nationality or par- ticular aliens. There is no doubt that the genocidal expropriations al- leged by plaintiffs would be considered discriminatory based on religion and/or ethnicity. But that is not the concern of § 712, which does not suggest that another nation’s dis- criminatory takings of property from its own nationals based on race, religion, or ethnicity violate international law and Nos. 13-3073 & 14-1319 15 can be the subject of litigation in the United States or any other nation. In our decision in 2012, we acknowledged the general “domestic takings” rule, under which international law does not address a nation’s taking of property from its own na- tionals. Abelesz, 692 F.3d at 674–75. We concluded, however, that genocide is so different and so universally condemned by international law that plaintiffs’ allegations of takings as an integral part of and a means of funding the genocidal campaign against Hungary’s Jews should not be subject to the domestic takings rule. Id. at 675–77. It was the strong links to genocide, not a broader concept of discrimination, that led us to find that plaintiffs had alleged takings in viola- tion of international law. If we had not considered the geno- cidal nature of the takings, we could not have found that violations of international law had been alleged. We would have applied the domestic takings rule to find that the Hun- garian government’s expropriations of property from its own nationals did not violate international law, consistent with the text of § 712(1) of the Restatement. Though we agreed then and now that violations were al- leged, that does not mean that international law allows those claims to be heard in any court in the world. We found that the comity at the heart of international law required plain- tiffs either to exhaust domestic remedies in Hungary or to show a powerful reason to excuse the requirement. So long as plaintiffs might get a fair shake in a domestic forum, in- ternational law expects plaintiffs at least to attempt to seek a remedy there first. The text and structure of the Third Restatement of For- eign Relations Law confirm this understanding of the role 16 Nos. 13-3073 & 14-1319 exhaustion plays with respect to any takings claim under in- ternational law. No matter what type of taking is alleged under § 712—whether discriminatory or otherwise—§ 713 explains that the same remedial scheme applies. And com- ment f of § 713 indicates that international law typically re- quires exhaustion of domestic remedies before any § 712 tak- ings claim can be heard in a foreign court. In other words, comment f’s domestic exhaustion requirement applies equal- ly to either type of taking specified in § 712, whether dis- criminatory or not. See Restatement (Third) of Foreign Rela- tions Law § 713 and cmt. f. The general requirement to exhaust domestic remedies as a matter of comity is not inconsistent with our comments in Flomo v. Firestone Nat’l Rubber Co., 643 F.3d 1013, 1025 (7th Cir. 2011), to the effect that no one could imagine requiring victims of genocide to have sued in Nazi Germany. The same point could have been made if these plaintiffs or their rela- tives had sought relief in 1945 in Hungary. There is of course no need to exhaust futile or imaginary domestic remedies. In these cases, though, plaintiffs filed suit in 2010, when the po- litical and legal landscapes were entirely different. See Abelesz, 692 F.3d at 681–82. We cannot reject out of hand the principle of comity calling for exhaustion of domestic reme- dies. 1 Plaintiffs contend that we have impermissibly implied an exhaustion requirement in the FSIA. After all, § 1605(a)(3) says nothing about a prudential exhaustion requirement 1United States law even today does not authorize suits against for- eign sovereigns for personal injury and death inflicted in genocide that occurs outside the United States. See 28 U.S.C. § 1605(a)(5), cited in Abelesz, 692 F.3d at 677. Nos. 13-3073 & 14-1319 17 based on international comity concerns. 2 As the Supreme Court recently held in Republic of Argentina v. NML Capital, Ltd., 134 S. Ct. 2250, 2256 (2014), “any sort of immunity de- fense made by a foreign sovereign in an American court must stand on the Act’s text. Or it must fall.” But here defendants need not rely on a special immunity defense found in the FSIA. Rather, they seek to invoke the customary rule itself: the well-established rule that exhaus- tion of domestic remedies is preferred in international law as a matter of comity. And there is no reason to think that this well-established rule is limited to foreign sovereigns. See Sosa, 542 U.S. at 733 n.21 (indicating that “in an appropriate case” international law may require that “the claimant must have exhausted any remedies available in the domestic legal sys- tem”); Kiobel, 133 S. Ct. at 1674 (Breyer, J., concurring in the judgment) (highlighting the limiting principles, such as ex- haustion and forum non conveniens, that “help to minimize international friction”). The Court in NML Capital was ad- dressing when a foreign sovereign can override the normal rule of post-judgment execution discovery and receive spe- cial treatment solely because it is a foreign sovereign. See NML Capital, 134 S. Ct. at 2255 (“The single, narrow question before us is whether the Foreign Sovereign Immunities Act specifies a different rule when the judgment debtor is a for- eign state.”). Because the FSIA’s text did not provide for spe- cial treatment, the foreign sovereign was subject to the nor- mal rule of post-judgment execution discovery. Id. at 2258. 2 Other circuit courts have declined to consider whether prudential exhaustion should apply to expropriation claims but have not foreclosed that possibility. See Cassirer, 616 F.3d at 1036–37; Agudas Chasidei Chabad, 528 F.3d at 949. 18 Nos. 13-3073 & 14-1319 When foreign sovereigns rely on the normal rule of interna- tional law, however, as in these cases, the FSIA’s silence on the issue does not prevent them from doing so. If the FSIA were read as preventing foreign nations from asking United States courts to apply a well-established in- ternational law principle—requiring exhaustion of domestic remedies—the result would be quite anomalous. It would become easier to sue foreign sovereigns than to sue private foreign entities in a United States court. The private entities could of course invoke the more general rule. Moreover, such a reading of the FSIA would call into question whether other settled practices such as forum non conveniens could still be invoked by foreign sovereigns. B. Reasons for Failing to Exhaust Thus, as we held in 2012, plaintiffs needed to show either that they exhausted any available Hungarian remedies or that there was a legally compelling reason to excuse such an effort. See Abelesz, 692 F.3d at 684. Plaintiffs have not argued that they have actually tried to exhaust available Hungarian remedies. They argue that circumstances in Hungary should excuse them from the domestic exhaustion rule. Before turning to the new arguments in these appeals, we note our earlier holdings that certain reasons would not es- tablish that “Hungarian courts would be so obviously inca- pable of providing a fair and impartial hearing” that a Unit- ed States court should step in. Abelesz, 692 F.3d at 684. Rely- ing on guidance from the Restatement, plaintiffs made three main arguments attempting to demonstrate that the seem- ingly available remedies were in fact not available. Id. at 682– 83 & n.12, citing § 713 cmt. f. First, they argued that the na- Nos. 13-3073 & 14-1319 19 tional defendants’ denial of the factual allegations of the complaint meant that Hungary denied responsibility. Sec- ond, plaintiffs argued that a 1993 Hungarian Constitutional Court decision finding a lack of compliance with Hungary’s obligations to pay reparations meant that the domestic rem- edies were inadequate. Finally, plaintiffs argued that any remedy would have been unreasonably prolonged given the delay between the events leading to the suit and the time when the Hungarian courts might redress the wrong. Id. at 682–83. These arguments did not amount to “a persuasive showing that Hungarian law is unresponsive.” Abelesz, 692 F.3d at 684. Remand provided an opportunity for plaintiffs to show further reasons that exhaustion should not be required. As explained below, we agree with the district court that the Hungarian remedies identified by defendants provide a fa- cially adequate mechanism for plaintiffs to seek redress. Plaintiffs have not established that procedural rules would arbitrarily or unreasonably bar their claims. Plaintiffs also have not shown that structural or political circumstances would prevent Hungarian courts from providing a fair and impartial hearing for these claims. To be sure, plaintiffs have offered explanations for their understandable doubts about the ability of Hungarian courts to treat them fairly. We be- lieve, however, that in the face of uncertainty, international comity requires that those courts be given the first oppor- tunity to hear the claims rather than have foreign courts as- sume the worst about them. 1. Existence of Hungarian Remedies In the district court the defendants identified various Hungarian remedies that might be available (or have been 20 Nos. 13-3073 & 14-1319 available) to plaintiffs. See AA 163–73 (national railway); AA 1170–87 (national bank). 3 The national defendants pointed to the First, Second, and Third Compensation Acts, as well as the Jewish Heritage Public Foundation, as remedies that may have been available to plaintiffs. The district court held that such non-judicial remedies “were not truly available to Plaintiffs due to the time and circumstances surrounding the application for such remedies and certain limitations placed on recoveries under such remedies.” Fischer, 2013 WL 4525408, at *1. Whether these compensation acts provide(d) an adequate remedy is no longer an issue on appeal. The national defendants also identified judicial remedies that may be available in a civil action in Hungary. These primarily include property-based claims and contractual claims that plaintiffs could assert against the banks. Hungar- ian courts will also entertain international law claims. AA 1179 & n.16 (national bank explaining that “Hungarian courts would be receptive” to international law based on its reading of the former and new Constitutions, Hungarian Constitutional Court decisions, and a Hungarian law ex- pert’s opinion), citing AA 928 (Act XX of 1949 Constitution [Former Constitution], art. 7(1)), AA 772 (Translation of The Fundamental Law [New Constitution] Apr. 25, 2011, art. Q), AA 892–901 (Constitutional Court Decision No. 53/1993), and AA 985–86 (Dr. Sonnevend Declaration ¶¶ 43–45). Plain- tiffs have not shown that the remedies identified by defend- ants are illusory. For this reason, the district court concluded 3The parties have provided record documents in various appendices to which we refer throughout the opinion. Citations are to the Short Ap- pendix (SA), Appellees’ Appendix (AA), and Erste Bank Group Appen- dix (EA). Nos. 13-3073 & 14-1319 21 that defendants had “shown that Plaintiffs can bring a civil action in the Hungarian courts to seek a remedy for the wrongs” alleged in their complaints. Fischer, 2013 WL 4525408, at *1. We agree that these judicial remedies are sufficiently promising that plaintiffs should be required to bring suit in Hungary before their suits may proceed in the United States. Given the sort of international law and property claims available in Hungary, the claims are congruent enough to those that plaintiffs assert here that requiring plaintiffs to sue first in Hungary will not deprive plaintiffs of the remedy they seek. As we explained in the first appeal, remedies available in Hungary “need not be perfectly congruent with those available in the United States to be deemed adequate.” See Abelesz, 692 F.3d at 685. 2. Potential Procedural Difficulties Plaintiffs suggest that several procedural obstacles might still deny them access to the remedies defendants identified. In the related context of forum non conveniens, the Supreme Court has made clear that an “unfavorable change in law may be given substantial weight” only when “the remedy provided by the alternative forum is so clearly inadequate or unsatisfactory that it is no remedy at all.” See Piper Aircraft Co. v. Reyno, 454 U.S. 235, 254 (1981). “[T]he relief need not be as comprehensive or as favorable as a plaintiff might ob- tain in an American court.” Chang v. Baxter Healthcare Corp., 599 F.3d 728, 736 (7th Cir. 2010). None of the asserted proce- dural obstacles deny relief to the extent that plaintiffs can claim that Hungary provides no remedy at all. 22 Nos. 13-3073 & 14-1319 First, plaintiffs argue that the lack of a class action device in Hungary ensures that these claims will never be brought. But the absence of a class action device does not mean as a matter of law that a nation’s courts fail to offer effective rem- edies. See Aguinda v. Texaco, Inc., 303 F.3d 470, 478 (2d Cir. 2002) (finding adequate a forum that lacked a class action device but allowed unlimited joinder of parties with similar claims arising out of the same facts); In re Union Carbide Corp. Gas Plant Disaster at Bhopal, India, 809 F.2d 195, 199, 202 (2d Cir. 1987) (finding adequate a forum that allowed repre- sentative suits in place of class actions). Until recently, rules allowing for group litigation were rare outside of the United States. Deborah R. Hensler, The Future of Mass Litigation: Global Class Actions and Third-Party Litigation Funding, 79 Geo. Wash. L. Rev. 306, 306 (2011). Even though many na- tions have joined the United States in adopting group litiga- tion procedural devices, American-style class actions remain uncommon. Id. at 307–08. A nation need not allow the rela- tively uncommon American-style class action to be consid- ered an adequate forum. See Aguinda, 303 F.3d at 478; see al- so Howe v. Goldcorp Investments, Ltd., 946 F.2d 944, 952 (1st Cir. 1991) (Breyer, C.J.). While Hungary does not have an identical class action device, it permits plaintiffs to use a joinder-like mechanism to enforce jointly claims arising out of the same legal rela- tionship or where the claims are based on similar legal and factual grounds. AA172 (national railway identifying in its submission on remedies a joinder mechanism), citing section 51 of Act III of 1952 on the Civil Procedure Code. While the lack of an identical class action device may well impose ad- ditional burdens on plaintiffs, the equivalent mechanism in Hungary does not appear to be so burdensome as to deprive Nos. 13-3073 & 14-1319 23 plaintiffs of an effective remedy. See Aguinda, 303 F.3d at 478 (“While the need for thousands of individual plaintiffs to au- thorize the action in their names is more burdensome than having them represented by a representative in a class ac- tion, it is not so burdensome as to deprive the plaintiffs of an effective alternative forum.”). Second, to the extent that plaintiffs worry that their claims may be time-barred, Hungary appears to have for- mally extended the statute of limitations for Holocaust- related claims. See AA1177 (national bank explaining in its submission on remedies that Hungary had formally extend- ed “the statute of limitations regarding Holocaust-era claims”), citing Decree No. 11 of 1960, art. 85(3); Abelesz, 692 F.3d at 682 n.11 (“Plaintiffs have advised [the Court] that Hungary has amended its constitution to declare that there are no statutes of limitations on crimes visited upon the Hungarian people during World War II.”). Moreover, coun- sel for the national defendants told us at oral argument that if plaintiffs bring these claims in Hungary, the national de- fendants would not assert any statute of limitations defens- es. The parties have not presented nor could we find any reason to think that Hungarian courts would not enforce such a waiver. With this waiver, we need not consider the effect such defenses might have on the adequacy of Hungar- ian remedies. Cf. Chang, 599 F.3d at 736 (noting that an alter- native forum might be inadequate if the plaintiff’s suit would be time-barred unless defendant agrees to waive the statute of limitations and the waiver would be enforced in the alternative forum). Finally, plaintiffs argue that changes in the Hungarian constitution might subject them to a risk of prosecution if 24 Nos. 13-3073 & 14-1319 they brought these Holocaust-era claims in Hungary. On the record before us, this fear of prosecution is too speculative to show that the Hungarian courts could not hear these claims fairly. Hungarian courts have ruled that good-faith argu- ments put forward during civil litigation cannot be used as a basis for civil and criminal defamation charges. See AA 497– 516 (collecting translations of case law explaining when liti- gants might be liable for statements made during litigation). The cases provided by the litigants indicate that liability might attach for statements made during litigation only if they are unreasonably offensive or false. 3. Structural Issues Concerning Hungarian Courts Plaintiffs also raise structural issues on the adequacy of Hungarian courts. They argue that even if remedies might be available to them in theory, limits on judicial independence would prevent those remedies from being effective. See gen- erally Sarei v. Rio Tinto, PLC, 550 F.3d 822, 832 (9th Cir. 2008) (en banc) (explaining that a “remedy must be available, ef- fective, and not futile” and that “a court must look at the cir- cumstances surrounding the access to a remedy” to measure effectiveness). For example, in Flomo v. Firestone Natural Rub- ber Co., 643 F.3d 1013, 1025 (7th Cir. 2011), we said it would “border on the ridiculous” to require litigants “to file suit in a court in Nazi Germany complaining about genocide” be- fore being allowed to bring suit in the United States. But we also made clear that as a “matter of international comity” we might also “give the courts of the nation in which the viola- tion had occurred a chance to remedy it, provided that the nation seemed willing and able to do that.” Id. The courts of Nazi Germany—or those of the regime in Hungary at that time—would have been unable to handle genocide claims Nos. 13-3073 & 14-1319 25 fairly. But we should not presume that the modern regimes replacing them many decades later are unwilling or unable to remedy the wrongs asserted by plaintiffs, absent specific evidence to that effect. Accordingly, plaintiffs offer two theories why Hungary is not willing and able to provide otherwise available reme- dies. First, they argue that the Hungarian government—and in particular, the judicial system—has been restructured so that claims like these would not receive a fair hearing in Hungary. Second, they contend that rising anti-Semitic atti- tudes in Hungary suggest that they would not receive a fair and impartial hearing and that their safety would be jeop- ardized if they sued there. In the years leading up to 2012, all parties agree, plain- tiffs could have trusted Hungary to handle these claims fair- ly. Plaintiffs’ own expert witness on Hungarian law agreed with the defendants’ expert that prior to 2012, “Hungary [was] a well-established European state, with a well func- tioning legal system that operates under established and cognizable rules of law.” AA 219–20 (plaintiffs’ expert agree- ing with this characterization before questioning whether the situation had changed and whether that system would allow for effective remedies for Holocaust victims), quoting Declaration of Dr. Sonnevend. In the 2012 appeals, the rec- ord supported our conclusion that Hungary had “a func- tional and independent judiciary” with an “apparent ability to provide an adequate remedy to plaintiffs.” Abelesz, 692 F.3d at 683. While we were sympathetic to plaintiffs’ con- cerns about anti-Semitism, there was not sufficient evidence to conclude that “Hungarian courts would be so obviously 26 Nos. 13-3073 & 14-1319 incapable of providing a fair and impartial hearing” that United States courts should step in. Id. at 684. For this reason, plaintiffs focus on how the situation has changed since 2012. Plaintiffs first point to new language in the 2011 New Constitution’s preamble as evidence that Hungary implicitly disavows responsibility for the Holo- caust. The preamble dates Hungary’s “restoration of our country’s self-determination, lost on the nineteenth day of March 1944, from the second day of May 1990.” That date is considered “the beginning of [Hungary’s] new democracy and constitutional order.” AA 767 (translation of New Con- stitution, Apr. 25, 2011). Though this was brought to our at- tention during the prior appeal, plaintiffs present additional opinion evidence that this provision should be read as the Hungarian government’s denial of responsibility for the Hungarian Holocaust. At the same time, others have read the provision as serving different ends, including a desire to emphasize the nation’s constitutional heritage. See, e.g., AA 467–68 (European Centre for Law and Justice Memorandum on the Hungarian New Constitution). Either way, absent a clear interpretation of that provision, the language of the preamble falls well short of the firm denial required by the Restatement to excuse the need for exhaustion. § 713 cmt. f (excusing domestic exhaustion for expropriation claim when the “state firmly denies responsibility”). Plaintiffs also stress that a combination of laws passed by the government and the Fourth Amendment to the New Constitution enabled the Hungarian government to under- mine the judiciary’s independence. The government lowered the retirement age for judges from 70 to 62, added six judges to the Constitutional Court, created a National Judicial Of- Nos. 13-3073 & 14-1319 27 fice with power to control dockets, and eliminated the prec- edential value of twenty years of Constitutional Court deci- sions. In response, defendants explain that the most controver- sial measures are either not to be worried about or no longer in place. The later Fifth Amendment eliminated the National Judicial Office’s power to assign cases. AA 1431, 1438 (decla- ration of defendants’ expert). The European Court of Justice rejected the proposed change in retirement age, and the gov- ernment allowed any forcibly retired judges to return to their jobs. AA1437–38. Finally, defendants clarify that the provision rendering prior decisions of the Constitutional Court no longer binding was not meant to undermine the legal effects of those decisions. Rather, defendants argue that it was designed to clarify that, going forward, the Constitu- tional Court should endeavor to interpret the new Constitu- tion. AA 1430. Though one might worry that the speed with which change has been made to the judiciary signifies a lack of re- spect for the rule of law, we are also encouraged by the Hungarian government’s willingness to revisit these provi- sions. By doing so, the government has responded to inter- national criticisms and shown some effort to respect interna- tional norms and values, including an independent judiciary. Again, we believe we understand plaintiffs’ concerns, but we believe they are too speculative to override the norm of re- quiring exhaustion of domestic remedies before resorting to foreign courts. In addition to these concerns about the independence of the Hungarian courts, plaintiffs argue that growing anti- Semitic attitudes coupled with attempts to minimize Hunga- 28 Nos. 13-3073 & 14-1319 ry’s role in the Holocaust make Hungary unable or unwilling to hear Holocaust-era claims. They also argue that their safe- ty could be jeopardized if they visited Hungary to testify in court. More specifically, plaintiffs point to troubling evidence that anti-Semitism is on the rise in Hungary and that Hunga- ry may be among the worst in Europe today on that score. AA 240 (declaration of Dr. Halmai pointing to survey evi- dence), citing Anti-Defamation League, Attitudes Toward Jews in Ten European Countries, March 2012. The plaintiffs’ expert expressed concern that some factions would take political advantage of these sentiments and push the governing party to do the same. AA 237–39 (declaration of Dr. Halmai), citing The Trajectory of Democracy—Why Hungary Matters: Hearing Before the Commission on Security & Cooperation in Europe (2013) (written and oral statements of Paul A. Shapiro, Direc- tor, Center for Advanced Holocaust Studies). Along these lines, plaintiffs highlight testimony explaining that the Hun- garian government has attempted to minimize Hungary’s role in the Holocaust and to rehabilitate the reputations of historical and cultural figures known to have been anti- Semitic or perpetrators of the Holocaust. Id. These events have led some Hungarian Jews to question whether there might now be good reason to leave Hungary. AA 1213–15 (Lisa Abend, Ancient Fear Rises Anew, Time Magazine, Apr. 1, 2013). At the same time, plaintiffs’ own expert testified clearly that the “current governing parties in Hungary are certainly not anti-Semitic.” AA 222 (Dr. Halmai declaration). The Prime Minister of Hungary has expressed a zero-tolerance policy with respect to anti-Semitism. AA 327–28 (Prime Min- Nos. 13-3073 & 14-1319 29 ister: Anti-Semitism is Unacceptable and Intolerable, Prime Min- ister’s Office—News (May 6, 2013)); AA 330–31 (Speech by Prime Minister Viktor Orbán at the 14th Plenary Assembly of the World Jewish Congress, Prime Minister’s Office—Speeches (May 6, 2013)). Unlike the United States, Hungary has made public denial of the Holocaust a criminal offense. AA 492 (translation of section 269/C of Act IV of 1978 on the Crimi- nal Code). And while there are some political groups who do not adhere to that positive vision of how Jews should be treated, the same might be said of other countries through- out Europe and the world. As plaintiffs acknowledge, anti- Semitism unfortunately has been on the rise throughout Eu- rope and is also present in the United States. But absent gov- ernmental policies or other evidence that such discrimina- tion is barring access to or punishing resort to domestic rem- edies, United States courts should not take the step of hear- ing these claims without first giving the Hungarian courts a chance to rule on them. To hold otherwise would imply that United States courts should presume that the courts of other nations cannot fairly hear claims brought by historically per- secuted groups. Altogether, the evidence in the record supports under- standable concerns about whether plaintiffs can receive a fair hearing in Hungary. But those concerns remain too specula- tive to justify taking this case from Hungarian courts. One could easily imagine that Thurgood Marshall and the NAACP Legal Defense and Educational Fund had similar concerns about many United States courts’ ability to hear claims by African Americans in 1950 and later. Yet our courts by and large rose to the challenge in the following decades. 30 Nos. 13-3073 & 14-1319 We are not persuaded that “Hungarian courts would be so obviously incapable of providing a fair and impartial hearing” that United States courts should intervene. Abelesz, 692 F.3d at 684; cf. Stroitelstvo Bulgaria Ltd. v. Bulgarian- American Enterprise Fund, 589 F.3d 417, 421 (7th Cir. 2009) (finding anecdotal complaints insufficient to allow a federal court to declare a foreign legal system inadequate). The rec- ord identifies developments that signal that the Hungarian government or people might do troubling things if confront- ed with these claims. As far as we can tell from the record here, though, plaintiffs have not offered examples where other people have attempted to bring Holocaust claims only to have the Hungarian courts close their doors to them or otherwise treat them unfairly. For this reason, we agree with the district court that plaintiffs “offer mere speculation and unsupported fears that they may not be treated fairly in the Hungarian court system.” Fischer, 2013 WL 4525408, at *1. In fact, other evidence in the record indicates that plain- tiffs’ fears are unlikely to be realized if they file suit in Hun- gary. The Hungarian government and courts appear to have ably handled discussion of the Holocaust as well as Holo- caust-era litigation. Defendants have cited cases where Hol- ocaust victims sued in Hungary for the return of works of art and antique furniture taken during the Holocaust. AA 290– 92 (return of stolen artwork); AA 337–38 (return of paint- ings). In these cases, plaintiffs successfully sued instrumen- talities of the government and won return of their property. More recently, the Hungarian government itself has estab- lished a program to return artwork that the state owns but may have obtained through unclear circumstances or wrongdoing. AA 342–43 (new office designed to review goods entrusted to the state). On top of these examples, de- Nos. 13-3073 & 14-1319 31 fendants show that the Hungarian government recently complied with a request by an attorney in the United States to depose a Hungarian war criminal. See Charles S. Fax, “A Tale of Discovery under the Hague Convention,” American Bar Association: Litigation News, Fall 2013. While these ex- amples do not guarantee, of course, that plaintiffs’ claims will be treated fairly, they tend to indicate that such claims can be heard fairly. In sum, the district court did not err in finding that plain- tiffs had not presented a legally compelling reason for their failure to exhaust remedies in Hungary. We emphasize, however, that the district court’s dismissal of claims against the national railway and bank was properly without preju- dice. If plaintiffs attempt to bring suit in Hungary and are blocked arbitrarily or unreasonably, United States courts could once again be open to these claims against the national railway and bank. III. Forum Non Conveniens We now turn to the district court’s dismissal of the claims against Erste Bank based on the doctrine of forum non conven- iens. When the district court dismissed for forum non conven- iens, Erste Bank had also moved to dismiss for lack of subject matter jurisdiction in light of the Supreme Court’s decision in Kiobel v. Royal Dutch Petroleum Co., 133 S. Ct. 1659 (2013). A district court may grant a motion to dismiss for forum non conveniens before it first determines its own subject matter jurisdiction over an action. Sinochem Int’l Co. v. Malaysia Int’l Shipping Corp., 549 U.S. 422, 429 (2007). We affirm on forum non conveniens, so we do not decide whether the district court could exercise subject matter jurisdiction over the 32 Nos. 13-3073 & 14-1319 claims against Erste Bank under the Alien Tort Statute or any other jurisdictional statute. 4 A district court may dismiss a case on forum non conven- iens grounds when it determines that there are “strong rea- sons for believing it should be litigated in the courts of an- other, normally a foreign, jurisdiction.” Abad v. Bayer Corp., 563 F.3d 663, 665 (7th Cir. 2009), citing Sinochem, 549 U.S. at 429–30, and In re Factor VIII or IX Concentrate Blood Products Litigation, 484 F.3d 951, 954–56 (7th Cir. 2007). While many considerations are part of this inquiry, the focus is “the con- venience to the parties and the practical difficulties that can attend the adjudication of a dispute in a certain locality.” Si- nochem, 549 U.S. at 429 (internal quotation marks and cita- tion omitted); see also Kamel v. Hill-Rom Co., 108 F.3d 799, 802 (7th Cir. 1997) (forum non conveniens dismissal appropriate when dismissal “best serves the convenience of the parties and the ends of justice”), citing Gulf Oil Corp. v. Gilbert, 330 U.S. 501, 507 (1947), and Koster v. Lumbermens Mutual Casual- ty Co., 330 U.S. 518, 527 (1947). Thus, when an alternative fo- rum has jurisdiction to hear a dispute, a case can be dis- missed if trial in the plaintiff’s chosen forum would be more oppressive to the defendant than it would be convenient to the plaintiff or if the forum otherwise creates administrative and legal problems that render it inappropriate. See Sino- chem, 549 U.S. at 429; Kamel, 108 F.3d at 802. 4 While we do not resolve the merits of the jurisdictional dispute Ki- obel raises, at least some Justices have indicated that forum non conveniens may limit the international frictions threatened by hearing cases such as these in non-domestic courts. Kiobel, 133 S. Ct. at 1674 (Breyer, J., concur- ring in the judgment). Nos. 13-3073 & 14-1319 33 Given the variety of relevant factors, their case-specific nature, and the absence of any formula for weighing them precisely, a district court’s decision to dismiss a suit for forum non conveniens is “committed to the sound discretion of the trial court” and “may be reversed only when there has been a clear abuse of discretion.” Piper Aircraft Co. v. Reyno, 454 U.S. 235, 257 (1981); see also Abad, 563 F.3d at 665–66. “Where a district court has contemplated all relevant public and private interest factors and where its balancing of these factors is reasonable, its forum non conveniens determination warrants substantial deference.” Kamel, 108 F.3d at 802, citing Piper, 454 U.S. at 254. Accordingly, we must take care not to “substitute[] [our] own judgment for that of the District Court.” Piper, 454 U.S. at 257. To decide an appeal of a forum non conveniens dismissal, we determine first whether there is an adequate alternative forum available to hear the case. Kamel, 108 F.3d at 802. If so, we evaluate the various private and public interest factors to see whether a finding that the balance counseled in favor of dismissal was reasonable. Id. at 803. A. The Adequacy of Hungary as an Alternative Forum A forum meets the adequate alternative forum require- ment when the forum is both available and adequate. Kamel, 108 F.3d at 802. The availability requirement is satisfied here. An alternative forum is “available if all parties are amenable to process and are within the forum’s jurisdiction.” Id. at 803. The other parties that plaintiffs sought to hold jointly and severally liable with Erste Bank—including private Hungar- ian banks and the national Hungarian bank—are all located in Hungary. All should be subject to the jurisdiction of the Hungarian courts. Also, the Austrian party (Erste Bank) con- 34 Nos. 13-3073 & 14-1319 sented to jurisdiction of the Hungarian courts. EA 171–72 (declaration of Erste Bank’s chief legal officer). With all par- ties subject to jurisdiction in Hungary, it counts as an availa- ble forum. See Stroitelstvo Bulgaria Ltd. v. Bulgarian-American Enterprise Fund, 589 F.3d 417, 421 (7th Cir. 2009). As shown by our discussion of the domestic exhaustion issue above, the adequacy of Hungary as a forum for these claims is hotly disputed. An “alternative forum is adequate when the parties will not be deprived of all remedies or treated unfairly.” Kamel, 108 F.3d at 803. Because jurisdiction would otherwise be appropriate in the foreign forum, it is not enough to say that the transfer will “lead to a change in applicable law unfavorable to the plaintiff.” In re Factor VIII or IX Concentrate, 484 F.3d at 956, citing Piper, 454 U.S. at 247, 254. Rather, only if “‘the remedy provided by the alternative forum is so clearly inadequate or unsatisfactory that it is no remedy at all’ should the unfavorable change be given sub- stantial—or even dispositive—weight.” Id., quoting Piper, 454 U.S. at 254. For the reasons stated above on the exhaustion issue, the district court correctly determined that the remedies provid- ed by Hungary would not be so clearly inadequate so as to provide no remedy at all. To be sure, the burden of proof dif- fers between the two inquiries. In the exhaustion analysis, it was up to plaintiffs to point to a legally compelling reason that the remedies might be inadequate. Here, by contrast, the burden ultimately falls on defendants to establish that the remedies are adequate. Sinochem, 549 U.S. at 430 (“A defend- ant invoking forum non conveniens ordinarily bears a heavy burden in opposing the plaintiff’s chosen forum.”); 14D Wright & Miller, Federal Practice and Procedure Jurisdiction Nos. 13-3073 & 14-1319 35 § 3828.3 (4th ed.). Thus, it would be at least theoretically pos- sible for a district court to find that neither side had met its burden of persuasion on the adequacy of remedies in the other forum. That theoretical possibility is not a problem on this rec- ord. Between the briefing in support of domestic exhaustion and the briefing in support of the forum non conveniens mo- tion, Erste Bank—along with its co-defendants—presented the evidence needed to determine the adequacy of Hunga- ry’s remedies. The district court had before it a list of the available remedies, plaintiffs’ concerns with bringing suit in Hungary, and expert testimony from both sides on whether those concerns were enough to render the forum inadequate. The court did not err by finding that “there is no evidence that the Hungarian courts do not offer an adequate alterna- tive forum for the claims brought against Erste.” Starting from that baseline, we address plaintiffs’ remain- ing contentions that the adequacy of Hungarian remedies was not established. As discussed above, plaintiffs’ argu- ments relating to the lack of a class action procedure and the safety concerns posed by anti-Semitism are not enough to defeat adequacy. Plaintiffs also suggest that a “loser pays” fee-shifting mechanism renders any possible relief too burdensome. But that approach to attorney fees is common throughout the world. In re Factor VIII or IX Concentrate, 484 F.3d at 958 (not- ing that the “United States stands almost alone in its ap- proach toward attorneys’ fees”); see also Piper, 454 U.S. at 252 n.18. That it might be more expensive for plaintiffs to lit- igate in the alternative forum is not a sufficient reason, standing alone, to find that forum inadequate. See In re Fac- 36 Nos. 13-3073 & 14-1319 tor VIII or IX Concentrate, 484 F.3d at 958; Stroitelstvo, 589 F.3d at 424 (holding that financial hardship caused by additional filing fees did not render alternative forum inadequate). We have rejected this particular difference in forum rules as a reason to find that a forum non convenience dismissal was in- appropriate. In re Factor VIII or IX Concentrate, 484 F.3d at 958. The district court properly concluded there was an ade- quate alternative forum. B. Balance of the Interests Where another adequate forum is available, to dismiss on forum non conveniens the district court must also balance the private interests of the parties and the public interests of the alternative forums and find that those balances favor a dif- ferent forum. Stroitelstvo, 589 F.3d at 424; Kamel, 108 F.3d at 803, citing Gilbert, 330 U.S. at 508. More specifically, courts evaluate the private interest by focusing on the “(1) relative ease of access to sources of proof; (2) availability of compul- sory process and costs for attendance of witnesses; (3) possi- bility of view of premises, if appropriate; and (4) other prac- tical issues, including ease of enforcement of any ultimate judgment.” In re Factor VIII or IX Concentrate, 484 F.3d at 958, citing Gilbert, 330 U.S. at 508. The court must also consider the public interest. Relevant public interest factors include the administrative difficulties stemming from court congestion; the local interest in having localized disputes decided at home; the interest in having the trial of a diversity case in a forum that is at home with the law that must govern the action; the avoidance of unnecessary problems in conflicts of laws or in the application of foreign law; and the unfairness of bur- dening citizens in an unrelated forum with jury duty. Nos. 13-3073 & 14-1319 37 Stroitelstvo, 589 F.3d at 425 (citations omitted). Plaintiffs argue that the district court erred at a basic lev- el by failing to consider all of these factors in granting Erste Bank’s motion for reconsideration. From the court’s initial forum non conveniens determination and the shorter order on reconsideration, however, we believe it is clear that the dis- trict court evaluated the public and private interests relevant to the decision. In its order granting reconsideration, the district court fo- cused on the relevant factors and explained why the “devel- opments” in the case had pushed it to change its initial rul- ing. The court’s initial determination in 2011 was that the balance tipped in favor of retaining the case. The district court found then that the defendants had “not shown that the convenience of the parties, or the interests of justice would be best served by a dismissal of the instant action.” More specifically, the district court focused on the relative convenience of the parties, the broad dispersion of the evi- dence and witnesses, and the interest the United States has in providing remedies for human rights violations. Against that backdrop, the district court limited its ex- planation on reconsideration to those factors that might have changed in light of intervening developments, in terms of both this court’s precedent and the fact that all other defend- ants had been dismissed from the bank case. With all other defendants out of the case, and this court having stressed the value of Hungarian courts having a chance to hear these claims, it is easy to see why the district court would see that the balance had shifted. In this way, the district court rea- sonably determined that what was initially a close question decided in favor of retaining the suit had tipped the other 38 Nos. 13-3073 & 14-1319 way once the relative burdens of suit and the public interest factors started to favor Hungary as a forum. Even so, plaintiffs argue there was not sufficiently de- tailed treatment of each factor. Precedent makes clear that substantial deference is warranted when all relevant factors are discussed. Kamel, 108 F.3d at 802. Though the oft-quoted list is long, the Supreme Court has also noted that it did not catalogue all of the circumstances when forum non conveniens dismissal might be appropriate. Gilbert, 330 U.S. at 508, ac- cord, Abad, 563 F.3d at 668. Perhaps that’s why the Court has made clear that the central focus is on the convenience of the parties and the practical difficulties of hearing a dispute in a certain locality. See Sinochem, 549 U.S. at 429. And in the past we have responded to the “laundry list” of factors provided by parties not by proceeding down the list but by consider- ing whether the district judge reasonably concluded that the litigation should continue in the foreign forum. Abad, 563 F.3d at 668. We apply that approach here. The district court acted well within its discretion in find- ing that the balance favored dismissal. It is hard to see how the district court might have reached any other result here given the weight of international comity concerns in this case. Because the national bank and other private banks were dismissed from the suit for lack of subject matter and personal jurisdiction, respectively, Erste Bank is the only bank defendant still subject to suit in the United States. That fact alone means that the Erste Bank litigation becomes less convenient in the United States and more so in Hungary. When dismissal might eliminate the need for similar litiga- tion in multiple places, forum non conveniens is favored. See, e.g., Sinochem, 549 U.S. at 435. While it is true that in other Nos. 13-3073 & 14-1319 39 cases the related proceedings in the foreign forum had al- ready begun, it is also true that Hungary is the only forum where all the litigation relating to these claims against all de- fendants might be heard. Requiring suit in Hungary is the best way to ensure that plaintiffs’ claims against the banks are litigated only once on the merits. Plaintiffs suggest that they can ensure there will only be one suit—the one remaining in the United States—because they simply will not bring suit in Hungary against the pre- viously dismissed parties. Even if that assurance were bind- ing, plaintiffs seek to hold Erste Bank jointly and severally liable for the entire amount of the substantial harm caused to plaintiffs as a result of these crimes. For that reason, it is like- ly that Erste Bank would sue the national bank defendant and private bank defendants for either indemnification or contribution. Such problems can sometimes be resolved by having a defendant join additional parties as third-party defendants. The fact that the United States does not have jurisdiction over the relevant parties precludes that solution here. This inability to join the other parties supports holding trial in Hungary. See Piper, 454 U.S. at 259. Also, we would be espe- cially wary of ruling on these claims absent these third par- ties because doing so might affect the interests of a foreign sovereign who previously was dismissed from suit. See Re- public of Philippines v. Pimentel, 553 U.S. 851, 868–69 (2008) (concluding that the immunity sovereigns enjoy is dimin- ished “if an important and consequential ruling affecting the sovereign’s substantial interest is determined, or at least as- sumed, by a federal court in the sovereign’s absence and over its objection”). The national bank should have the op- 40 Nos. 13-3073 & 14-1319 portunity to participate in any litigation that decides wheth- er the various Hungarian banks are holding property in vio- lation of Hungarian or international law. At the end of the day, then, at least some additional liti- gation would have to occur in Hungary even if plaintiffs won their suit against Erste Bank here. The type of evidence presented at both trials, which would look to establish the culpability of various bank defendants for the expropria- tions, would likely be similar. For these reasons, the district court correctly concluded that dismissing this case against the banks will not put additional litigation burdens on plain- tiffs. Beyond the efficiency gains of litigating all claims in Hungary, the district court was right to point out that the dismissal of Erste Bank’s co-defendants raises concerns about access to evidence. As we noted before, “Hungary is where much of the evidence and surviving witnesses are lo- cated.” Abelesz, 692 F.3d at 684. While we take plaintiffs’ point that the testimony of plaintiffs themselves might be the most essential, Erste Bank also suggests that at least some important evidence is in the control of the dismissed parties. Erste Bank casts doubt on whether treaty obligations would oblige the national bank to comply with international dis- covery requests. See EA 173–76 (declaration of Erste Bank’s lawyer and attached exhibit laying out Hungary’s reserva- tions to the Hague Evidence Convention). That the district court might no longer be able to compel record evidence lo- cated in Hungary while plaintiffs could still bring that evi- dence (their testimony) to Hungary weighs in favor of dis- missal. And without some of the bank records in this case, the already difficult task of determining the succession of Nos. 13-3073 & 14-1319 41 bank property would become even more difficult if not im- possible. It is not clear from the record how the costs of obtaining evidence will shift if this litigation moves to Hungary. Plain- tiffs indicate that the witnesses are widely dispersed. For that reason, perhaps moving from the United States to Hun- gary will not increase costs as much as it might if witnesses were concentrated in the United States. But for the same rea- son, it is also not clear that moving the litigation will reduce travel burdens imposed on witnesses. Either way, because witnesses are not concentrated in the United States or Hun- gary, some will need to travel to testify. In terms of docu- ment costs, though, it stands to reason that most of the rele- vant bank records and other documents will be in Hungari- an. Though neither party has provided an estimate of the costs of translating such materials, see Abad, 563 F.3d at 669, it seems obvious that otherwise heavy translation burdens will be greatly reduced if the case were litigated in Hungary. In addition to the private interest factors tipping in favor of litigation in Hungary, public interest concerns make clear that Hungary is the better forum. As discussed at length above, Hungary has a significant legal interest in hearing these claims in Hungarian courts. That is why plaintiffs should be required to exhaust in Hungary before bringing suit here against the national defendants. Allowing plaintiffs to sue a private bank on substantially similar claims might unduly prejudice a foreign sovereign in a way that under- mines the reason we dismissed the national bank for lack of exhaustion of domestic remedies. Dismissal for forum non conveniens accommodates international law norms and Hun- gary’s interest in being able to address these claims first. By 42 Nos. 13-3073 & 14-1319 contrast, while the United States has a general interest in en- suring that international law norms are enforced, in this par- ticular case the executive branch has recommended “dismis- sal of claims against Erste Bank … on any valid legal ground(s).” EA 70 (United States statement of interest). This suggests that the United States’ interest in this litigation is modest. And the international comity concerns that would be raised if United States courts retained the case say nothing of the practical judicial problems. The public interest factors “point towards dismissal where the court would be required to ‘untangle problems in conflict of laws, and in law foreign to itself.’” Piper, 454 U.S. at 251 (citation omitted). It is at best uncertain what the basis for subject matter jurisdiction over Erste Bank might be. Even if jurisdiction in the United States were secure, it is likely that Hungarian law would apply to questions not governed by international law. See Abad, 563 F.3d at 669–72.5 A United States court would have to apply Hungarian law to a host of delicate issues, especially those concerning remedies. The application of foreign law— particularly that of a civil law system—favors dismissal in favor of a Hungarian forum. Stroitelstvo, 589 F.3d at 426; Abad, 563 F.3d at 670. In this case, a Hungarian court would be far better able to apply its own law than any United States court would be. Plaintiffs argue that whatever the outcome of the balance might have been, the district court failed to give their choice 5 At oral argument, plaintiffs indicated that if jurisdiction were based on 28 U.S.C. § 1332, Hungarian law would apply. Nos. 13-3073 & 14-1319 43 of forum sufficient weight. A plaintiff’s choice is presumed convenient to a plaintiff who chooses his home forum. Kamel, 108 F.3d at 803. And because at least some of the plaintiffs here are United States residents, we assume they have opted for the most convenient forum by suing in the United States. The district court emphasized this presump- tion in its first ruling, and it did nothing to suggest it had shifted the burden when it considered these factors in its motion for reconsideration. That presumption of conven- ience has been rebutted by the strength of the private and public factors discussed above. At bottom, the district court did not abuse its discretion when it concluded that the more reasonable and convenient forum for the suit against Erste Bank is Hungary. Though it may seem unfair that Erste Bank was dismissed on forum non conveniens grounds in large part due to the prior dismissals of its co-defendants, this result should not be unexpected. When plaintiffs seek joint and several liability against multi- ple foreign defendants, dismissal of any one defendant can shift the balance. Thus, we uphold the district court’s dismis- sal of Erste Bank. IV. Conclusion Because plaintiffs have not exhausted their Hungarian remedies and have not yet provided a legally compelling reason for their failure to do so, their claims against the na- tional defendants were properly dismissed without preju- dice. The district court did not abuse its discretion by dis- missing the claims against Erste Bank without prejudice based on forum non conveniens. The judgments of the district court are AFFIRMED.