Norfolk Southern Railway Co. v. James N. Kirby, Pty Ltd.

delivered the opinion of the Court.

This is a maritime case about a train wreck. A shipment of machinery from Australia was destined for Huntsville, Alabama. The intercontinental journey was uneventful, and the machinery reached the United States unharmed. But the train carrying the machinery on its final, inland leg derailed, causing extensive damage. The machinery’s owner sued the railroad. The railroad seeks shelter in two liability limitations contained in contracts that upstream carriers negotiated for the machinery’s delivery.

I

This controversy arises from two bills of lading (essentially, contracts) for the transportation of goods from Australia to Alabama. A bill of lading records that a carrier has received goods from the party that wishes to ship them, *19 states the terms of carriage, and serves as evidence of the contract for carriage. See 2 T. Schoenbaum, Admiralty and Maritime Law 58-60 (3d ed. 2001) (hereinafter Schoenbaum); Carriage of Goods by Sea Act (COGSA), 49 Stat. 1208, 46 U. S. C. App. § 1303. Respondent James N. Kirby, Pty Ltd. (Kirby), an Australian manufacturing company, sold 10 containers of machinery to the General Motors plant located outside Huntsville, Alabama. Kirby hired International Cargo Control (ICC), an Australian freight forwarding company, to arrange for delivery by “through” (i e., end-to-end) transportation. (A freight forwarding company arranges for, coordinates, and facilitates cargo transport, but does not itself transport cargo.) To formalize their contract for carriage, ICC issued a bill of lading to Kirby (ICC bill). The bill designates Sydney, Australia, as the port of loading, Savannah, Georgia, as the port of discharge, and Huntsville as the ultimate destination for delivery.

In negotiating the ICC bill, Kirby had the opportunity to declare the full value of the machinery and to have ICC assume liability for that value. Cf. New York, N. H. & H. R. Co. v. Nothnagle, 346 U. S. 128, 135 (1953) (a carrier must provide a shipper with a fair opportunity to declare value). Instead, and as is common in the industry, see Sturley, Carriage of Goods by Sea, 31 J. Mar. L. & Com. 241, 244 (2000), Kirby accepted a contractual liability limitation for ICC below the machinery’s true value, resulting, presumably, in lower shipping rates. The ICC bill sets various liability limitations for the journey from Sydney to Huntsville. For the sea leg, the ICC bill invokes the default liability rule set forth in the COGSA. The COGSA “package limitation” provides:

“Neither the carrier nor the ship shall in any event be or become liable for any loss or damage to or in connection with the transportation of goods in an amount exceeding $500 per package lawful money of the United States . . . unless the nature and value of such goods *20 have been declared by the shipper before shipment and inserted in the bill of lading.” 46 U. S. C. App. § 1304(5).

For the land leg, in turn, the bill limits the carrier’s liability to a higher amount. 1 So that other downstream parties expected to take part in the contract’s execution could benefit from the liability limitations, the bill also contains a so-called “Himalaya Clause.” 2 It provides:

“These conditions [for limitations on liability] apply whenever claims relating to the performance of the contract evidenced by this [bill of lading] are made against any servant, agent or other person (including any independent contractor) whose services have been used in order to perform the contract.” App. to Pet. for Cert. 59a, cl. 10.1.

*21 Meanwhile, Kirby separately insured the cargo for its true value with its co-respondent in this case, Allianz Australia Insurance' Ltd. (formerly MMI General Insurance, Ltd.).

Having been hired by Kirby, and because it does not itself actually transport cargo, ICC then hired Hamburg Südamer-ikanische Dampfschifffahrts-Gesellschaft Eggert & Amsinck (Hamburg Süd), a German ocean shipping company, to transport the containers. To formalize their contract for carriage, Hamburg Süd issued its own bill of lading to ICC (Hamburg Süd bill). That bill designates Sydney as the port of loading, Savannah as the port of discharge, and Huntsville as the ultimate destination for delivery. It adopts COGSA’s default rule in limiting the liability of Hamburg Süd, the bill’s designated carrier, to $500 per package. See 46 U. S. C. App. § 1304(5). It also contains a clause extending that liability limitation beyond the “tackles” — that is, to potential damage on land as well as on sea. Finally, it too contains a Himalaya Clause extending the benefit of its liability limitation to “all agents . . . (including inland) carriers . . . and all independent contractors whatsoever.” App. 63, cl. 5(b).

Acting through a subsidiary, Hamburg Süd hired petitioner Norfolk Southern Railway Company (Norfolk) to transport the machinery from the Savannah port to Huntsville. The Norfolk train carrying the machinery derailed en route, causing an alleged $1.5 million in damages. Kirby’s insurance company reimbursed Kirby for the loss. Kirby and its insurer then sued Norfolk in the United States District Court for the Northern District of Georgia, asserting diversity jurisdiction and alleging tort and contract claims. In its answer, Norfolk argued, among other things, that Kirby’s potential recovery could not exceed the amounts set forth in the liability limitations contained in the bills of lading for the machinery’s carriage.

The District Court granted Norfolk’s motion for partial summary judgment, holding that Norfolk’s liability was lim *22 ited to $500 per container. Upon a joint motion from Norfolk and Kirby, the District Court certified its decision for interlocutory review pursuant to 28 U. S. C. § 1292(b).

A divided panel of the Eleventh Circuit reversed. It held that Norfolk could not claim protection under the Himalaya Clause in the first contract, the ICC bill. It construed the language of the clause to exclude parties, like Norfolk, that had not been in privity with ICC when ICC issued the bill. 300 F. 3d 1300, 1308-1309 (2002). The majority also suggested that “a special degree of linguistic specificity is required to extend the benefits of a Himalaya clause to an inland carrier.” Id., at 1310. As for the Hamburg Siid bill, the court held that Kirby could be bound by the bill’s liability limitation “only if ICC was acting as Kirby’s agent when it received Hamburg Süd’s bill.” Id., at 1305. And, applying basic agency law principles, the Court of Appeals concluded that ICC had not been acting as Kirby’s agent when it received the bill. Ibid. Based on its opinion that Norfolk was not entitled to benefit from the liability limitation in either bill of lading, the Eleventh Circuit reversed the District Court’s grant of summary judgment for the railroad. We granted certiorari to decide whether Norfolk could take shelter in the liability limitations of either bill, 540 U. S. 1099 (2004), and now reverse.

II

The courts below appear to have decided this case on an assumption, shared by the parties, that federal rather than state law governs the interpretation of the two bills of lading. Respondents now object. They emphasize that, at bottom, this is a diversity case involving tort and contract claims arising out of a rail accident somewhere between Savannah and Huntsville. We think, however, borrowing from Justice Harlan, that “the situation presented here has a more genuinely salty flavor than that.” Kossick v. United Fruit Co., 365 U. S. 731, 742 (1961). When a contract is a maritime *23 one, and the dispute is not inherently local, federal law controls the contract interpretation. Id., at 735.

Our authority to make decisional law for the interpretation of maritime contracts stems from the Constitution’s grant of admiralty jurisdiction to. federal courts. See Art. III, § 2, cl. 1 (providing that the federal judicial power shall extend to “all Cases of admiralty and maritime Jurisdiction”). See 28 U. S. C. § 1333(1) (granting federal district courts original jurisdiction over “[a]ny civil case of admiralty or maritime jurisdiction”); R. Fallon, D. Meltzer, & D. Shapiro, Hart and Wechsler’s The Federal Courts and the Federal System 733-738 (5th ed. 2003). This suit was properly brought in diversity, but it could also be sustained under the admiralty jurisdiction by virtue of thé maritime contracts involved. See Pope & Talbot, Inc. v. Hawn, 346 U. S. 406, 411 (1953) (“[Substantial rights ... are not to be determined differently whether [a] ease is labelled ‘law side’ or ‘admiralty side’ oh a district court’s docket”). Indeed, for federal common law to apply in these circumstances, this suit must also be sustainable under the admiralty jurisdiction. See Stewart Organization, Inc. v. Ricoh Corp., 487 U. S. 22, 28 (1988). Because the grant of admiralty jurisdiction and the power to make admiralty law are mutually dependent, the two are often intertwined in our cases.

Applying the two-step analysis from Kossick, we find that federal law governs this contract dispute. Our cases do not draw clean lines between maritime and nonmaritime contracts. We have recognized that “[t]he boundaries of admiralty jurisdiction over contracts — as opposed to torts or crimes — being conceptual rather than spatial, have always been difficult to draw.” 365 U. S., at 735. To ascertain whether a contract is a maritime one, we cannot look to whether a ship or other vessel was involved in the dispute, as we would in a putative maritime tort case. Cf. Admiralty Extension Act, 46 U. S. C. App. § 740 (“The admiralty and maritime jurisdiction of the United States shall extend to *24 and include all cases of damage or injury . . . caused by a vessel on navigable water, notwithstanding that such damage or injury be done or consummated on land”); 1 R. Force & M. Norris, The Law of Seamen § 1:15 (5th ed. 2003). Nor can we simply look to the place of the contract’s formation or performance. Instead, the answer “depends upon . . . the nature and character of the contract,” and the true criterion is whether it has “reference to maritime service or maritime transactions.” North Pacific S. S. Co. v. Hall Brothers Marine Railway & Shipbuilding Co., 249 U. S. 119, 125 (1919) (citing Insurance Co. v. Dunham, 11 Wall. 1, 26 (1871)). See also Exxon Corp. v. Central Gulf Lines, Inc., 500 U. S. 603, 611 (1991) (“[T]he trend in modern admiralty case law ... is to focus the jurisdictional inquiry upon whether the nature of the transaction was maritime”).

The ICC and Hamburg Süd bills are maritime contracts because their primary objective is to accomplish the transportation of goods by sea from Australia to the eastern coast of the United States. See G. Gilmore & C. Black, Law of Admiralty 31 (2d ed. 1975) (“Ideally, the [admiralty] jurisdiction [over contracts ought] to include those and only those things principally connected with maritime transportation” (emphasis deleted)). To be sure, the two bills call for some performance on land; the final leg of the machinery’s journey to Huntsville was by rail. But under a conceptual rather than spatial approach, this fact does not alter the essentially maritime nature of the contracts.

In Kossick, for example, we held that a shipowner’s promise to assume responsibility for any improper treatment his seaman might receive at a New York hospital was a maritime contract. The seaman had asked the shipowner to pay for treatment by a private physician, but the shipowner, preferring the cheaper public hospital, offered to cover the costs of any complications that might arise from treatment there. We characterized his promise as a “fringe benefit” to a shipowner’s duty in maritime law to provide “ ‘maintenance and *25 cure.’ ” 365 U. S., at 736-737. Because the promise was in furtherance of a “peculiarly maritime concer[n],” id., at 738, it folded into federal maritime law. It did not matter that the site of the inadequate treatment — which gave rise to the contract dispute — was in a hospital on land. Likewise, Norfolk’s rail journey from Savannah to Huntsville was a “fringe” portion of the intercontinental journey promised in the ICC and Hamburg Süd bills.

We have reiterated that the “ ‘fundamental interest giving rise to maritime jurisdiction is “thé protection of maritime commerce”’”’ Exxon, supra, at 608 (emphasis added) (quoting Sisson v. Ruby, 497 U. S. 358, 367 (1990), in turn quoting Foremost Ins. Co. v. Richardson, 457 U. S. 668, 674 (1982)). The conceptual approach vindicates that interest by focusing our inquiry on whether the principal objective of a contract is maritime commerce. While it may once have seemed natural to think that only contracts embodying commercial obligations between .the “tackles” (i. e., from port to port) have maritime objectives, the shore is now an artificial place to draw a line. Maritime commerce has evolved along with the nature of transportation and is often inseparable from some land-based obligations. The international transportation industry “clearly has moved into a new era — the age of multimodalism, door-to-door transport based on efficient use of all available modes of transportation by air, water, and land.” 1 Schoenbaum 589 (4th ed. 2004). The cause is technological change: Because goods can now be packaged in standardized containers, cargo can move easily from one mode of transport to another. Ibid. See also NLRB v. Longshoremen, 447 U. S. 490, 494 (1980) (“ ‘[C]on-tainerization may be said to constitute the single most important innovation in ocean transport since the steamship displaced the schooner’”); G. Muller, Intermodal Freight Transportation 15-24 (3d ed. 1995).

Contracts reflect the new technology, hence the popularity of “through” bills of lading, in which cargo owners can con *26 tract for transportation across oceans and to inland destinations in a single transaction. See 1 Schoenbaum 595. Put simply, it is to Kirby’s advantage to arrange for transport from Sydney to Huntsville in one bill of lading, rather than to negotiate a separate contract — and to find an American railroad itself — for the land leg. The popularity of that efficient choice, to assimilate land legs into international ocean bills of lading, should not render bills for ocean carriage non-maritime contracts.

Some lower federal courts appear to have taken a spatial approach when deciding whether intermodal transportation contracts for intercontinental shipping are maritime in nature. They have held that admiralty jurisdiction does not extend to contracts which require maritime and nonmaritime transportation, unless the nonmaritime transportation is merely incidental — and that long-distance land travel is not incidental. See, e. g., Hartford Fire Ins. Co. v. Orient Overseas Containers Lines (UK) Ltd., 230 F. 3d 549, 555-556 (CA2 2000) (“Transport by land "under a bill of lading is not ‘incidental’ to transport by sea if the land segment involves great and substantial distances,” and land transport of over 850 miles across four countries is more than incidental); Sea-Land Serv., Inc. v. Danzig, 211 F. 3d 1373, 1378 (CA Fed. 2000) (holding that intermodal transport contracts were not maritime contracts because they called for “substantial transportation between inland locations and ports both in this country and in the Middle East” that was not incidental to the transportation by sea); Kuehne & Nagel (AG & Co.) v. Geosource, Inc., 874 F. 2d 283, 290 (CA5 1989) (holding that a through bill of lading calling for land transportation up to 1,000 miles was not a traditional maritime contract because such “extensive land-based operations cannot be viewed as merely incidental to the maritime operations”). As a preliminary matter, it seems to us imprecise to describe the land carriage required by an intermodal transportation contract as “incidental”; realistically, each leg of the journey *27 is essential to accomplishing the contract’s purpose. In this case, for example, the bills of lading required delivery to Huntsville; the Savannah port would not do.

Furthermore, to the extent that these lower court decisions fashion a rule for identifying maritime contracts that depends solely on geography, they are inconsistent with the conceptual approach our precedent requires. See Kossick, supra, at 735. Conceptually, so long as a bill of lading requires substantial carriage of goods by sea, its purpose is to effectuate maritime commerce — and thus it is a maritime contract. Its character as a maritime contract is not defeated simply because it also provides for some land carriage. Geography, then, is useful in a conceptual inquiry only in a limited sense: If a bill’s sea components are insubstantial, then the bill is not a maritime contract.

Having established that the ICC and Hamburg Slid bills are maritime contracts, then, we must clear a second hurdle before applying federal law in their interpretation. Is this case inherently local? For. not “every term in every maritime contract can only be controlled by some federally defined admiralty rule.” Wilburn Boat Co. v. Fireman’s Fund Ins. Co., 348 U. S. 310, 313 (1955) (applying state law to maritime contract for marine insurance because of state regulatory power over insurance industry). A maritime contract’s interpretation may so implicate local interests as to beckon interpretation by state law. See Kossick, 365 U. S., at 735. Respondents have not articulated any specific Australian or state interest at stake, though some are surely implicated. But when state interests cannot be accommodated without defeating a federal interest, as is the case here, then federal substantive law should govern. See id., at 739 (the process of deciding whether federal law applies “is surely ... one of accommodation, entirely familiar in many areas of overlapping state and federal concern, or a process somewhat analogous to the normal conflict of laws situation where two sov-ereignties assert divergent interests in a transaction”); 2 *28 Schoenbaum 61 (“ ‘Bills of lading issued outside the United States are governed by the general maritime law, considering relevant choice of law rules’ ”).

Here, our touchstone is a concern for the uniform meaning of maritime contracts like the ICC and Hamburg Süd bills. We have explained that Article Ill’s grant of admiralty jurisdiction “ ‘must have referred to a system of law coextensive with, and operating uniformly in, the whole country. It certainly could not have been the intention to place the rules and limits of maritime law under the disposal and regulation of the several States, as that would have defeated the uniformity and consistency at which the Constitution aimed on all subjects of a commercial character affecting the intercourse of the States with each other or with foreign states.’ ” American Dredging Co. v. Miller, 510 U. S. 443, 451 (1994) (quoting The Lottawanna, 21 Wall. 558, 575 (1875)). See also Yamaha Motor Corp., U. S. A. v. Calhoun, 516 U. S. 199, 210 (1996) (“[I]n several contexts, we have recognized that vindication of maritime policies demanded uniform adherence to a federal rule of decision” (citing Kossick, supra, at 742; Pope & Talbot, 346 U. S., at 409; Garrett v. Moore-McCormack Co., 317 U. S. 239, 248-249 (1942))); Romero v. International Terminal Operating Co., 358 U. S. 354, 373 (1959) (“[Sjtate law must yield to the needs of a uniform federal maritime law when this Court finds inroads on a harmonious system[,] [b]ut this limitation still leaves the States a wide scope”).

Applying state law to cases like this one would undermine the uniformity of general maritime law. The same liability limitation in a single bill of lading for international inter-modal transportation often applies both to sea and to land, as is true of the Hamburg Süd bill. Such liability clauses are regularly executed around the world. See 1 Schoenbaum 595; Wood, Multimodal Transportation: An American Perspective on Carrier Liability and Bill of Lading Issues, 46 Am. J. Comp. L. 403, 407 (Supp. 1998). See also *29 46 U. S. C. App. § 1307 (permitting parties to extend the COGSA default liability limit to damage done “prior to the loading on and subsequent to the discharge from the ship”). Likewise, a single Himalaya Clause can cover both sea and land carriers downstream, as is true of the ICC bill. See Part III-A, infra. Confusion and inefficiency will inevitably result if more than one body of law governs a given contract’s meaning. As we said in Kossick, when “a [maritime] contract . . . may well have been made anywhere in the world,” it “should be judged by one law wherever it was made.” 365 U. S., at 741. Here, that one law is federal.

In protecting the uniformity of federal maritime law, we also reinforce the liability regime Congress established in COGSA. By its terms, COGSA governs bills of lading for the carriage of goods “from the time when the goods are loaded on to the time when they are discharged from the ship.” 46 U. S. C. App. § 1301(e). For that period, COGSA’s “package limitation” operates as a default rule. §1304(5). But COGSA also gives the option of extending its rule by contract. See §1307 (“Nothing contained in this chapter shall prevent a carrier or a shipper from entering into any agreement, stipulation, condition, reservation, or exemption as to the responsibility and liability of the carrier or the ship for the loss or damage to or in connection with the custody and care and handling of goods prior to the loading on and subsequent to the discharge from the ship on which the goods are carried by sea”). As COGSA permits, Hamburg Süd in its bill of lading chose to extend the default rule to the entire period in which the machinery would be under its responsibility, including the period of the inland transport. Hamburg Süd would not enjoy the efficiencies of the default rule if the liability limitation it chose did not apply equally to all legs of the journey for which it undertook responsibility. And the apparent purpose of COGSA, to facilitate efficient contracting in contracts for carriage by sea, would be defeated.

*30 III

A

Turning to the merits, we begin with the ICC bill of lading, the first of the contracts at issue. Kirby and ICC made a contract for the carriage of machinery from Sydney to Huntsville, and agreed to limit the liability of ICC and other parties who would participate in transporting the machinery. The bill’s Himalaya Clause states:

“These conditions [for limitations on liability] apply whenever claims relating to the performance of the contract evidenced by this [bill of lading] are made against any servant, agent or other person (including any inde-' pendent contractor) whose services have been used in order to perform the contract.” App. to Pet. for Cert. 59a, cl. 10.1 (emphasis added).

The question presented is whether the liability limitation in Kirby’s and ICC’s contract extends to Norfolk, which is ICC’s sub-subcontractor. The Circuits have split in answering this question. Compare, e. g., Akiyama Corp. of America v. M. V. Hanjin Marseilles, 162 F. 3d 571, 574 (CA9 1998) (privity of contract is not required in order to benefit from a Himalaya Clause), with Mikinberg v. Baltic S. S. Co., 988 F. 2d 327, 332 (CA2 1993) (a contractual relationship is required).

This is a simple question of contract interpretation. It turns only on whether the Eleventh Circuit correctly applied this Court’s decision in Robert C. Herd & Co. v. Krawill Machinery Corp., 359 U. S. 297 (1959). We conclude that it did not. In Herd, the bill of lading between a cargo owner and carrier said that, consistent with COGSA, “ ‘the Carrier’s liability, if any, shall be determined on the basis of $500 per package.’” Id., at 302. The carrier then hired a stevedor-ing company to load the cargo onto the ship, and the steve-doring company damaged the goods. The Court held that the stevedoring company was not a beneficiary of the bill’s liability limitation. Because it found no evidence in COGSA *31 or its legislative history that Congress meant COGSA’s liability limitation to extend automatically to a carrier’s agents, like stevedores, the Court looked to the language of the bill of lading itself. It reasoned that a clause limiting “ ‘the Carrier’s liability’” did not “indicate that the contracting parties intended to limit the liability of stevedores or other agents. ... If such had been a purpose of the contracting parties it must be presumed that they would in some way have expressed it in the contract.” Ibid. The Court added that liability limitations must be “strictly construed and limited to intended beneficiaries.” Id., at 305.

The Eleventh Circuit, like respondents, made much of the Herd decision. Deriving a principle of narrow construction from Herd, the Court of Appeals concluded that the language of the ICC bill’s Himalaya Clause is too vague to clearly include Norfolk. 300 F. 3d, at 1308. Moreover, the lower court interpreted Herd to require privity between the carrier and the party seeking shelter under a Himalaya Clause. 300 F. 3d, at 1308. But nothing in Herd requires the linguistic specificity or privity rules that the Eleventh Circuit attributes to it. The decision simply says that contracts for carriage of goods by sea must be construed like any other contracts: by their terms and consistent with the intent of the parties. If anything, Herd stands for the proposition that there is no special rule for Himalaya Clauses.

The Court of Appeals’ ruling is not true to the contract language or to the intent of the parties. The plain language of the Himalaya Clause indicates an intent to extend the liability limitation broadly — to “any servant, agent or other person (including any independent contractor)” whose services contribute to performing the contract. App. to Pet. for Cert. 59a, cl. 10.1 (emphasis added). “Read naturally, the word ‘any’ has an expansive meaning, that is, ‘one or some indiscriminately of whatever kind.’ ” United States v. Gonzales, 520 U. S. 1, 5 (1997) (quoting Webster’s Third New International Dictionary 97 (1976)). There is no reason to con *32 travene the clause’s obvious meaning. See Green v. Biddle, 8 Wheat. 1, 89-90 (1828) (“[W]here the words of a law, treaty, or contract, have a plain and obvious meaning, all construction, in hostility with such meaning, is excluded”). The expansive contract language corresponds to the fact that various modes of transportation would be involved in performing the contract. Kirby and ICC contracted for the transportation of machinery from Australia to Huntsville, Alabama, and, as the crow flies, Huntsville is some 366 miles inland from the port of discharge. See G. Fitzpatrick & M. Modlin, Direct-Line Distances 168 (1986). Thus, the parties must have anticipated that a land carrier’s services would be necessary for the contract’s performance. It is clear to us that a railroad like Norfolk was an intended beneficiary of the ICC bill’s broadly written Himalaya Clause. Accordingly, Norfolk’s liability is limited by the terms of that clause.

B

The question arising from the Hamburg Süd bill of lading is more difficult. It requires us to set an efficient default rule for certain shipping contracts, a task that has been a challenge for courts for centuries. See, e. g., Hadley v. Baxendale, 9 Exch. 341, 156 Eng. Rep. 145 (1854). ICC and Hamburg Süd agreed that Hamburg Süd would transport the machinery from Sydney to Huntsville, and agreed to the COGSA “package limitation” on the liability of Hamburg Süd, its agents, and its independent contractors. The second question presented is whether that liability limitation, which ICC negotiated, prevents Kirby from suing Norfolk (Hamburg Süd’s independent contractor) for more. As we have explained, the liability limitation in the ICC bill, the first contract, sets liability for a land accident higher than this bill does. See n. 1, supra. Because Norfolk’s liability will be lower if it is protected by the Hamburg Süd bill too, we must reach this second question in order to give Norfolk the full relief for which it petitioned.

*33 To interpret the Hamburg Slid bill, we turn to a rule drawn from our precedent about common carriage: When an intermediary contracts with a carrier to transport goods, the cargo owner’s recovery against the carrier is limited by the liability limitation to which the intermediary and carrier agreed. The intermediary is certainly not automatically empowered to be the cargo owner’s agent in every sense. That would be unsustainable. But when it comes to liability limitations for negligence resulting in damage, an intermediary can negotiate reliable and enforceable agreements with the carriers it engages.

We derive this rule from our decision about common carriage in Great Northern R. Co. v. O’Connor, 232 U. S. 508 (1914). In Great Northern, an owner hired a transfer company to arrange for the shipment of her goods. Without the owner’s express authority, the transfer company arranged for rail transport at a tariff rate that limited the railroad’s liability to less than the true value of the goods. The goods were lost en route, and the owner sued the railroad. The Court held that the railroad must be able to rely on the liability limitation in its tariff agreement with the transfer company. The railroad “had the right to assume that the Transfer Company could agree upon the terms of the shipment”; it could not be expected to know if the transfer company had any outstanding, conflicting obligation to another party. Id., at 514. The owner’s remedy,- if necessary, was against the transfer company. Id., at 515.

Respondents object to our reading of Great Northern, and argue that this Court should fashion the federal rule of decision from general agency law principles. Like the Eleventh Circuit, respondents reason that Kirby cannot be bound by the bill of lading that ICC negotiated with Hamburg Slid unless ICC was then acting as Kirby’s agent. Other Courts of Appeals have also applied agency law to cases similar to this one. See, e. g., Kukje Hwajae Ins. Co. v. The M/V Hyundai Liberty, 294 F. 3d 1171, 1175-1177 (CA9 2002) (an *34 intermediary acted as a cargo owner’s agent when negotiating a bill of lading with a downstream carrier).

We think reliance on agency law is misplaced here. It is undeniable that the traditional indicia of agency, a fiduciary relationship and effective control by the principal, did not exist between Kirby and ICC. See Restatement (Second) of Agency § 1 (1957). But that is of no moment. The principle derived from Great Northern does not require treating ICC as Kirby’s agent in the classic sense. It only requires treating ICC as Kirby’s agent for a single, limited purpose: when ICC contracts with subsequent carriers for limitation on liability. In holding that an intermediary binds a cargo owner to the liability limitations it negotiates with downstream carriers, we do not infringe on traditional agency principles. We merely ensure the reliability of downstream contracts for liability limitations. In Great Northern, because the intermediary had been “entrusted with goods to be shipped by railway, and, nothing to the contrary appearing, the carrier had the right to assume that [the intermediary] could agree upon the terms of the shipment.” 232 U. S., at 514. Likewise, here we hold that intermediaries, entrusted with goods, are “agents” only in their ability to contract for liability limitations with carriers downstream.

Respondents also contend that any decision binding Kirby to the Hamburg Slid bill’s liability limitation will be disastrous for the international shipping industry. Various participants in the industry have weighed in as amici on both sides in this case, and we must make a close call. It would be idle to pretend that the industry can easily be characterized, or that efficient default rules can easily be discerned. In the final balance, however, we disagree with respondents for three reasons.

First, we believe that a limited agency rule tracks industry practices. In intercontinental ocean shipping, carriers may not know if they are dealing with an intermediary, rather *35 than with a cargo owner. Even if knowingly dealing with an intermediary, they may not know how many other intermediaries came before, or what obligations may be outstanding among them. If the Eleventh Circuit’s rule were the law, carriers would have to seek out more information before contracting, so as to assure themselves that their contractual liability limitations provide true protection. That task of information gathering might be very costly or even impossible, given that goods often change hands many times in the course of intermodal transportation. See 1 Schoenbaum 589; Wood, 46 Am. J. Comp. L., at 404.

Second, if liability limitations negotiated with cargo owners were reliable while limitations negotiated with intermediaries were not, carriers would likely want to charge the latter higher rates. A rule prompting downstream carriers to distinguish between cargo owners and intermediary shippers might interfere with statutory and decisional law promoting nondiscrimination in common carriage. Cf. ICC v. Delaware, L. & W. R. Co., 220 U. S. 235, 251-256 (1911) (common carrier cannot “sit in judgment on the title of the prospective shipper”); Shipping Act, 46 U. S. C. App. § 1709 (nondiscrimination rules). It would also, as we have intimated, undermine COGSA’s liability regime.

Finally, as in Great Northern, our decision produces an equitable result. See 232 U. S., at 515. Kirby retains the option to sue ICC, the carrier, for any loss that exceeds the liability limitation to which they agreed. And indeed, Kirby has sued ICC in an Australian court for damages arising from the Norfolk derailment. It seems logical that ICC— the only party that definitely knew about and was party to both of the bills of lading at issue here — should bear responsibility for any gap between the liability limitations in the bills. Meanwhile, Norfolk enjoys the benefit of the Hamburg Slid bill’s liability limitation.

*36 IV

We hold that Norfolk is entitled to the protection of the liability limitations in the two bills of lading. Having undertaken this analysis, we recognize that our decision does no more than provide a legal backdrop against which future bills of lading will be negotiated. It is n