with whom Justice Kennedy joins, dissenting.
I agree with much of the Court’s reasoning. Our precedents fully support the Court’s reasoning that the Eleventh Circuit’s test is overinclusive; that the subjective intentions of a watercraft’s owner or designer play no role in the vessel analysis of 1 U. S. C. §3; and that an objective assessment of a watercraft’s purpose or function governs whether that structure is a vessel. The Court, however, creates a novel and unnecessary “reasonable observer” reformulation of these principles and errs in its determination, under this new standard, that the craft before us is not a vessel. Given the underdeveloped record below, we should remand. Therefore, I respectfully dissent.
I
The relevant statute, 1 U. S. C. § 3, “sweeps broadly.” Stewart v. Dutra Constr. Co., 543 U. S. 481, 494 (2005). It provides that “[t]he word ‘vessel’ includes every description of watercraft or other artificial contrivance used, or capable of being used, as a means of transportation on water.” This broad phrasing flows from admiralty law’s long recognition that vessels come in many shapes and sizes. See E. Benedict, American Admiralty §218, p. 121 (1870 ed.) (“ ‘[VJessel, is a general word, many times used for any kind of navigation’ ”); M. Cohen, Admiralty Jurisdiction, Law, and Practice 232 (1883) (“ ‘[T]he term “vessel” shall be understood to comprehend every description of vessel navigating on any sea or channel, lake or river . . . ’ ”).
Our test for vessel status has remained the same for decades: “Under § 3, a ‘vessel’ is any watercraft practically capable of maritime transportation . . . .” Stewart, 543 U. S., at 497; see also Evansville & Bowling Green Packet Co. v. Chero Cola Bottling Co., 271 U. S. 19, 22 (1926); Cope v. Valletta Dry Dock Co., 119 U. S. 625, 627 (1887). At its core, vessel status has always rested upon the objective physical characteristics of a vessel (such as its structure, shape, and *135materials of construction), as well as its usage history. But over time, several important principles have guided both this Court and the lower courts in determining what kinds of watercraft fall properly within the scope of admiralty jurisdiction.
Consider the most basic of requirements. Por a watercraft to be “practically capable” of maritime transportation, it must first be “capable” of such transportation. Only those structures that can simultaneously float and carry people or things over water are even presumptively within § 3⅛ reach. Stopping here, as the Eleventh Circuit essentially did, results in an overinclusive test. Section 3, after all, does not drag every bit of floating and towable flotsam and jetsam into admiralty jurisdiction. Rather, the terms “capable of being used” and “practical” have real significance in our maritime jurisprudence.
“[A] water craft is not ‘capable of being used’ for maritime transport in any meaningful sense if it has been permanently moored.” Stewart, 543 U. S., at 494. So, to take an obvious example, a floating bridge over water does not constitute a vessel; such mooring is clearly permanent. Cf. The Rock Island Bridge, 6 Wall. 213, 216 (1867). Less dramatically, a watercraft whose objective physical connections to land “evidence a permanent location” does not fall within § 3's ambit. See, e. g., Evansville, 271 U. S., at 22 (“[The wharfboat] served at Evansville as an office, warehouse and wharf, and was not taken from place to place. The connections with the water, electric light and telephone systems of the city evidence a permanent location”); Dunklin v. Louisiana Riverboat Gaming Partnership, No. 00-31455, 2001 WL 650209, *1, n. 1 (CA5, May 22, 2001) (per curiam) (describing a fully functional casino boat placed “in an enclosed pond in a cofferdam”). Put plainly, structures “permanently affixed to shore or resting on the ocean floor,” Stewart, 543 U. S., at 493-494, have never been treated as vessels for the purposes of § 3.
*136Our precedents have also excluded from vessel status those watercraft “rendered practically incapable of transportation or movement.” Id., at 494. Take the easiest case, a vessel whose physical characteristics have been so altered as to make waterborne transportation a practical impossibility. Ibid, (explaining that a “floating processing plant was no longer a vessel where a ‘large opening [had been] cut into her hull,’ rendering her incapable of moving over the water” (quoting Kathriner v. UNISEA, Inc., 975 F. 2d 657, 660 (CA9 1992))). The longstanding admiralty exception for “dead ships,” those watercraft that “require a major overhaul” for their “reactivation,” also falls into this category. See Roper v. United States, 368 U. S. 20, 21 (1961) (finding that a liberty ship “deactivated from service and ‘mothballed’” is not a “vessel in navigation”); see generally Rutherglen, Dead Ships, 30 J. Maritime L. & Comm. 677 (1999).1 Likewise, ships that “have been withdrawn from the water for extended periods of time” in order to facilitate repairs and reconstruction may lose their status as vessels until they are rendered capable of maritime transport. Stewart, 543 U. S., at 496. Cf. West v. United States, 361 U. S. 118, 120, 122 (1959) (noting: “[T]he Mary Austin was withdrawn from any operation whatever while in storage with the ‘moth-ball fleet’ ” and that “[t]he Mary Austin, as anyone could see, was not in maritime service. She was undergoing major repairs and complete renovation . . . ”).
Finally, our maritime jurisprudence excludes from vessel status those floating structures that, based on their physical characteristics, do not “transport people, freight, or cargo *137from place to place” as one of their purposes. Stewart, 543 U. S., at 493. “Purpose,” in this context, is determined solely by an objective inquiry into a craft’s function. “[Neither size, form, equipment nor means of propulsion are determinative factors upon the question of [vessel status],” though all may be considered. The Robert W. Parsons, 191 U. S. 17, 30 (1903). Moreover, in assessing a particular structure’s function, we have consistently examined its past and present activities. Stewart, 543 U. S., at 495; Cope, 119 U. S., at 627. Of course, a seaborne craft is not excluded from vessel status simply because its “primary purpose” is not maritime transport. Stewart, 543 U. S., at 497. We held as much in Stewart when we concluded that a dredge was a vessel notwithstanding that its “primary purpose” was “dredging rather than transportation.” Id., at 486, 495. So long as one purpose of a craft is transportation, whether of cargo or people or both, § 3’s practical capability requirement is satisfied.
Certainly, difficult and marginal cases will arise. Fortunately, courts do not consider each floating structure anew. So, for example, when we were confronted in Stewart with the question whether a dredge is a §3 vessel, we did not commence with a clean slate; we instead sought guidance from previous cases that had confronted similar structures. See id., at 490, and n. 5; see also Norton v. Warner Co., 321 U. S. 565, 571-572 (1944) (likewise surveying earlier cases).
In sum, our precedents offer substantial guidance for how objectively to determine whether a watercraft is practically capable of maritime transport and thus qualifies as a § 3 vessel. First, the capacity to float and carry things or people is an obvious prerequisite to vessel status. Second, structures or ships that are permanently moored or fixed in place are not § 3 vessels. Likewise, structures that are practically incapable of maritime transport are not vessels, whether they are ships that have been altered so that they may no longer be put to sea, dead ships, or ships removed from navigation for extended periods of time. Third, those water*138craft whose physical characteristics and usage history reveal no maritime transport purpose or use are not § 3 vessels.
I—{ I—I
The majority does not appear to disavow the legal principles described above. The majority apparently accepts that permanent mooring suffices to take a ship out of vessel status, ante, at 125,129,2 and that “[a] craft whose physical characteristics and activities objectively evidence a waterborne transportation purpose or function may still be rendered a nonvessel by later physical alterations,” ante, at 129.3 No one argues that Lozman’s craft was permanently moored, see App. 32 (describing the “deteriorated” ropes holding the craft in place), or that it had undergone physical alterations sufficient to take it out of vessel status, see Tr. of Oral Arg. 13 (Lozman’s counsel arguing that the craft was never a vessel in the first place). Our precedents make clear that the Eleventh Circuit’s “anything that floats” test is overinclusive and ignores that purpose is a crucial factor in determining whether a particular craft is or is not a vessel. Accordingly, the majority is correct that determining whether Lozman’s craft is a vessel hinges on whether that craft had any maritime transportation purpose or function.
*139The majority errs, though, in concluding that the purpose component of the § 3 test is whether “a reasonable observer, looking to the [craft] ⅛ physical characteristics and activities, would not consider it to be designed to any practical degree for carrying people or things on water.” Ante, at 118. This phrasing has never appeared in any of our cases and the majority’s use of it, despite its seemingly objective gloss, effectively (and erroneously) introduces a subjective component into the vessel-status inquiry.
For one thing, in applying this test the majority points to some characteristics of Lozman’s craft that have no relationship to maritime transport, such as the style of the craft’s rooms or that “those inside those rooms looked out upon the world, not through watertight portholes, but through French doors or ordinary windows.” Ante, at 122. The majority never explains why it believes these particular esthetic elements are important for determining vessel status. In fact, they are not. Section 3 is focused on whether a structure is “used, or capable of being used, as a means of transportation on water.” By importing windows, doors, room style, and other esthetic criteria into the §3 analysis, the majority gives our vessel test an “I know it when I see it” flavor. Jacobellis v. Ohio, 378 U. S. 184, 197 (1964) (Stewart, J., concurring). But that has never been nor should it be the test: A badly designed and unattractive vessel is different from a structure that lacks any “practical capacity” for maritime transport. In the majority’s eyes, the two appear to be one and the same.
The majority’s treatment of the craft’s past voyages is also strange. The majority notes that Lozman’s craft could be and was, in fact, towed over long distances, including over 200 miles at one point. Ante, at 118. But the majority determines that, given the design of Lozman’s craft, this is “far too little actual ‘use’ to bring the floating home within the terms of the statute.” Ante, at 130. This is because “when it moved, it carried, not passengers or cargo, but at the very *140most (giving the benefit of any factual ambiguity to the City) only its own furnishings, its owner’s personal effects, and personnel present to ensure the home’s safety.” Ibid.
I find this analysis confusing. The majority accepts that the record indicates that Lozman’s craft traveled hundreds of miles while “carrying people or things.” Ante, at 118. But then, in the same breath, the majority concludes that a “reasonable observer” would nonetheless conclude that the craft was not “designed to any practical degree for carrying people or things on water.” Ibid. The majority fails to explain how a craft that apparently did carry people and things over water for long distances was not “practically capable” of maritime transport.
This is not to say that a structure capable of such feats is necessarily a vessel. A craft like Lozman’s might not be a vessel, for example, if it could only carry its owner’s clothes and personal effects, or if it is only capable of transporting itself and its appurtenances. Jerome B. Grubart, Inc. v. Great Lakes Dredge & Dock Co., 513 U. S. 527, 535 (1995) (“[Mjaritime law . . . ordinarily treats an ‘appurtenance’ attached to a vessel in navigable waters as part of the vessel itself”). But if such a craft can carry large appliances (like an oven or a refrigerator) and all of the other things we might find in a normal home in addition to the occupants of that home, as the existing record suggests Lozman’s craft may have done, then it would seem to be much more like a mobile home (and therefore a vessel) than a firmly rooted residence. The simple truth is that we know very little about the craft’s capabilities and what did or did not happen on its various trips. By focusing on the little we do know for certain about this craft (i. e., its windows, doors, and the style of its rooms) in determini jig whether it is a vessel, the majority renders the §3 inquiry opaque and unpredictable.
Indeed, the little we do know about Lozman’s craft suggests only that it was an unusual structure:. A surveyor was *141unable to find any comparable craft for sale in the State of Florida. App. 43. Lozman’s home was neither obviously a houseboat, as the majority describes such ships, ante, at 122, nor clearly a floating home, ante, at 126-127. See App. 13, 31, 79 (sale, lease, and surveying documents describing Loz-man’s craft as a “houseboat”). The only clear difference that the majority identifies between these two kinds of structures is that the former are self-propelled, while the latter are not. Ante, at 122. But even the majority recognizes that self-propulsion has never been a prerequisite for vessel status. Ibid, (citing The Robert W. Parsons, 191 U. S., at 31); see Norton, 321 U. S., at 571. Consequently, it is unclear why Lozman’s craft is a floating, home, why all floating homes are not vessels,4 or why Lozman’s craft is not a vessel. If windows, doors, and other esthetic attributes are what take Loz-man’s craft out of vessel status, then the majority’s test is completely malleable. If it is the craft’s lack of self-propulsion, then the majority’s test is unfaithful to our longstanding precedents. See The Robert W. Parsons, 191 U. S., at 30-31. If it is something else, then that something is not apparent from the majority’s opinion.
Worse still, in straining to find that Lozman’s craft was a floating home and therefore not a vessel, the majority calls into question the conclusions of numerous lower courts that have found houseboats that lacked self-propulsion to be §3 vessels. See ante, at 126 (citing Miami River Boat Yard, *142Inc. v. 60’ Houseboat, 390 F. 2d 596, 597 (CA5 1968); Sea Village Marina, LLC v. A 1980 Carlcraft Houseboat, No. 09-3292, 2009 WL 3379923, *5-*6 (D NJ, Oct. 19, 2009); Hudson Harbor 79th Street Boat Basin, Inc. v. Sea Casa, 469 F. Supp. 987, 989 (SDNY 1979)). The majority incorrectly suggests that these cases applied an “ ‘anything that floats’ ” test. Ante, at 126. These cases suggest something different. Many of these decisions in assessing the crafts before them looked carefully at these crafts’ structure and function, and determined that these ships had capabilities similar to other long-established vessels, suggesting a significant maritime transportation function. See Miami River Boat Yard, 390 F. 2d, at 597 (likening houseboat at issue to a “barg[e]”); Sea Village Marina, 2009 WL 3379923, *7 (“According to the available evidence, [the houseboats in question] float and can be towed to a new marina without substantial effort ...”); Hudson Harbor, 469 F. Supp., at 989 (houseboat “was capable of being used at least to the extent that a ‘dumb barge’ is capable of being used” and comparable to a “yach[tj”). Their holdings are consistent with older cases, see, e. g., The Ark, 17 F. 2d 446, 447 (SD Fla. 1926), and the crafts at issue in these cases have been widely accepted as vessels by most treatises in this area, see 1 S. Friedell, Benedict on Admiralty § 164, p. 10-6, n. 2 (rev. 7th ed. 2012); 1 T. Schoenbaum, Admiralty & Maritime Law §3-6, p. 153, n. 10 (5th ed. 2011); 1 R. Force & M. Norris, Law of Seamen §2:12, p. 2-82 (5th ed. 2003). The majority’s suggestion that rejecting the Eleventh Circuit’s test necessitates jettisoning these other precedents is simply wrong. And, in its rejection, the majority works real damage to what has long been a settled area of maritime law.5
*143HH HH HH
With a more developed record, Lozman’s craft might be distinguished from the houseboats in those lower court cases just discussed. For example, if Lozman’s craft’s previous voyages caused it serious damage, then that would strongly suggest that it lacked a maritime transportation purpose or function. There is no harm in remanding the case for further factfinding along the lines described above, cautioning the lower courts to be aware that features of Lozman’s “incomparable” craft, see App. 43, may distinguish it from previous precedents. At most, such a remand would introduce a relatively short delay before finally ending the years-long battle between Lozman and the city of Riviera Beach.
On the other hand, there is great harm in stretching the facts below and overriding settled and likely correct lower court precedents to reach the unnecessary conclusion that Lozman’s craft was not a vessel. Without an objective application of the § 3 standard, one that relies in a predictable fashion only on those physical characteristics of a craft that are related to maritime transport and use, parties will have no ex ante notion whether a particular ship is a vessel. As a wide range of amici have cautioned us, numerous maritime industries rely heavily on clear and predictable legal rules for determining which ships are vessels.6 The majority’s *144distorted application of our settled law to the facts of this case frustrates these ends. Moreover, the majority’s decision reaches well beyond relatively insignificant boats like Lozman’s craft, id., at 79 (listing purchase price of Lozman’s craft as $17,000), because it specifically disapproves of lower court decisions dealing with much larger ships, see ante, at 126 (questioning Holmes v. Atlantic Sounding Co., 437 F. 3d 441 (CA5 2006) (finding a 140-foot-long and 40-foot-wide dormitory barge with 50 beds to be a § 3 vessel)).
IV
It is not clear that Lozman’s craft is a §3 vessel. It is clear, however, that we are not in a good position to make such a determination based on the limited record we possess. The appropriate response is to remand the case for further proceedings in light of the proper legal standard. See Brief for United States as Amicus Curiae 29-31. The Court resists this move and in its haste to christen Lozman’s craft a nonvessel delivers an analysis that will confuse the lower courts and upset our longstanding admiralty precedent. I respectfully dissent.
The converse category of ships “not yet borji” is another historical exclusion from vessel status. See Tucker v. Alexandroff, 183 U. S. 424, 438 (1902) (“A ship is born when she is launched, '^.nd lives so long as her identity is preserved. Prior to her launching sh$ is a mere congeries of wood and iron—an ordinary piece of personal property—as distinctly a land structure as a house, and subject only to mechanics’ liens created by state law and enforceable in the state courts”).
In discussing permanent mooring, as well as Stewards rejection of primary-purpose and state-of-transit tests for vessel status, Stewart v. Dutra Constr. Co., 543 U. S. 481, 495 (2005), the majority states that our holdings “say, and they mean, that the statutory definition [given by § 3] may (or may not) apply—not that it automatically must apply—where a structure has some other primary purpose, where it is stationary at relevant times, and where it is attached—but not permanently attached— to land.” Ante, at 124. This must mean, by negative implication, that a permanently moored structure never falls within § 3⅛ definition.
Presumably, this encompasses those kinds of ships “otherwise rendered practically incapable of transportation or movement.” Stewart, 543 U. S., at 494. That is, ships which have been altered sol‘they cannot travel the seas, dead ships, and ships removed from the water for an extended period of time. Supra, at 135-136.
To be clear, some floating homes are obviously not vessels. For example, some floating homes are structures built upon a large inverted pyramid of logs. Brief for Seattle Floating Homes Association et al. as Amici Curiae 14. Cf. App. 38 (Lozman’s craft was buoyed by an empty bilge space). These kinds of floating homes can measure 4,000 or 5,000 square feet, see Brief for Seattle Floating Homes Association et al. as Amici Curiae 4, and may have connections to land that require the aid of divers and electricians to remove, ibid. These large, immobile structures are not vessels and have physical attributes directly connected to their lack of navigational abilities that suggest as much. But these structures are not before us; Lozman’s craft is.
The majority’s invocation of two state environmental and tax statutes as a reason to reject this well-established lower court precedent is particularly misguided. See ante, at 127. We have repeatedly emphasized that the “regulation of maritime vessels” is a “uniquely federal are[a] of regulation.” Chamber of Commerce of United States of America v. Whiting, 563 U. S. 582, 604 (2011) (plurality opinion) (emphasis added); *143see also United States v. Locke, 529 U. S. 89, 99 (2000) (explaining that “the federal interest [in regulating interstate navigation] has been manifest since the beginning of our Republic and is now well established”). Our previous cases did not turn to state law in determining whether a given craft is a vessel. There are no good reasons to do so now.
For example, without knowing whether a particular ship is a § 3 vessel, it is impossible for lenders to know how properly to characterize it as collateral for a financing agreement because they do not know what remedies they will have recourse to in the event of a default. Brief for National Marine Bankers Association as Amicus Curiae 14-15. Similarly, cities like Riviera Beach provide docking for crafts like Lozman’s on the assumption that such crafts actually are “vessels,” App. 13-21 (Riviera Beach’s wet-slip agreement referring to Lozman’s craft as a “vessel,” “boat,” or “houseboat”), that can be “remove[d]” upon short notice, id., at *14417 (requiring removal of the craft on three days’ notice). The majority makes it impossible for these marinas to know whether the “houseboats” that fill their slips are actually vessels and what remedies they can exercise in the event of a dispute. See id., at 15 (“In addition to any other remedies provided for in this Agreement, the Marina, as a provider of necessities to this vessel, has a maritime lien on the vessel and may bring a civil action in rem, under 46 United States Code 31342 in Federal Court, to arrest the vessel and enforce the lien...” (emphasis added)). Lozman’s behavior over the years is emblematic of this problem. For example, in 2003, prior to his move to Riviera Beach, Lozman had his craft towed from one marina to another after a dispute arose with; the first marina and he was threatened with eviction. Id., at 76-78. ’The possibility that a shipowner like Lozman can depart so easily over water and go beyond the reach of a provider of necessaries like the marina in response to a legal dispute is exactly the kind of problem that the Federal Maritime Lien Act, 46 U. S. C. § 31342, was intended to address. See* Dampskibsselskabet Dannebrog v. Signal Oil & Gas Co. of Cal., 310 U. S. 268, 272-273 (1940).