Petitioners Lockheed Martin Corporation and ACE USA, the insurer for Lock-heed Martin Corporation, petition for review of the February 17, 2004, decision of the Benefits Review Board of the United States Department of Labor ("Board"), reversing the findings of Administrative Law Judge Paul H. Teitler of the United States Department of Labor ("ALJ"), holding Rocco Morganti ("Morganti") to have been covered by the Longshore and Harbor Workers' Compensation Act, 33 U.S.C. §§ 901-950, ("LHWCA" or "Act") at the time of his fatal accident, and awarding death benefits to Morganti's widow, respondent Lorraine Morganti. We deny the petition and uphold the award of benefits to Lorraine Morganti.

We agree with the Board and hold that Cayuga Lake is a navigable waterway of the United States because the appropriate test for navigability depends on physical rather than economic characteristics of the waterway in question. We decline to decide whether to adopt a "transient and fortuitous" exception to the general rule that a person injured on actual navigable waters is engaged in maritime employment because we, like the Board, hold that a person on an object floating in actual navigable waters is on actual navigable waters, so that Morganti was not transiently and fortuitously on actual navigable waters. Finally, we agree with the Board's conclusion, although not its underlying reasoning, and hold that Morganti's employment, either in its totality or as performed on actual navigable waters, was not exclusively data processing. We therefore hold that Morganti was covered by the Act and deny the petition for review.

The underlying facts of this case are undisputed. Lorraine Morganti seeks benefits under a federal workers' compensation act, the LHWCA, for her husband's tragic death in the course of his employment. Lockheed seeks to deny these benefits on the ground that Morganti was not covered by the Act.

Morganti was employed for fifteen years as a test engineer by Lockheed Martin Naval Electronics & Surveillance Systems, a division of petitioner Lockheed Martin Corporation (collectively, "Lockheed") based in Syracuse, New York. His job as a test engineer for Lockheed's work under contracts with the United States Navy required him to spend thirty to forty percent of his time testing Lockheed-manufactured equipment on the Paganelli, a research barge moored on Cayuga Lake. On December 20, 2000, Morganti suffered a fatal accident on the shuttle boat returning him to shore from the Paganelli. While untying the shuttle, he fell into the lake. Despite rescue efforts by the pilot of the shuttle, Morganti drowned, and his body was not recovered until later.

Lockheed manufactures transducers, a component required for sonar equipment used by the United States Navy. To test these transducers underwater, Lockheed uses equipment on the Paganelli. The Paganelli has been moored to two mooring buoys since 1982. Each buoy is secured by a 30-ton anchor and a 4-ton sinker, but the buoys are fully independent of the Paganelli, which could disconnect from the buoys within minutes. The Paganelli has no independent means of propulsion. Access to the Paganelli is achieved via a shuttle boat, the Little Toot II.

Cayuga Lake is one of the finger lakes of Central New York, and is connected at its north end to the Erie Canal system, which means that a boat can theoretically sail from Cayuga Lake to any ocean-connected port in the world. The locks of the Erie Canal are limited in size, but sufficient for ninety percent of the ships currently used for inland commercial freight shipping and passenger services. The Erie Canal is closed for the winter months of the year. At trial, Lockheed explicitly conceded that the Erie Canal is a navigable waterway and that Cayuga Lake can physically support commercial shipping. At this time, no commercial shipping actually takes place on Cayuga Lake other than commercial tour boats, which entered or exited the lake over three hundred times in 2001. There is no record evidence as to whether these boats traveled interstate.

Morganti was responsible for testing Lockheed's transducers in the waters of Cayuga Lake and pinpointing and troubleshooting failures in the transducer production process at an office in Syracuse. His qualifications for this position are not in the record, but a witness with a similar position held an associate degree in electrical technology. Morganti carried a security clearance of "secret" in order to work with the transducers and evaluate the testing data. He received a salary of approximately $63,000 per year.

Morganti was required to spend one-and-a-half to two days per week testing transducers on the Paganelli. He was not required to help moor the Little Toot II, handle the transducers, or physically connect testing equipment to the transducers, but he frequently did so voluntarily. Once a transducer was lowered into the water through an internal well, Morganti would use a computer to run tests for an hour, by loading each testing program from disks as needed and inputting information such as water temperature, water depth, and which test to run. Once the tests were completed, he would review the test output, consisting of numbers and graphs, for facial validity and reasonableness. Typically, further analysis would be carried out in Syracuse, not on the Paganelli.

On December 20, 2000, Morganti was untying Little Toot II to return to shore when he fell into the water and drowned. His widow, respondent Lorraine Morganti, filed a claim for death benefits under the Act with the Office of Workers' Compensation Programs of the United States Department of Labor on July 8, 2001. The ALJ, Judge Teitler, denied her claim, on the grounds that 1) Morganti was not a maritime employee under Section 2(3) of the Act, 33 U.S.C. § 902(3); 2) the Paganelli was a fixed platform categorized as artificial land akin to an island under Herb's Welding, Inc. v. Gray, 470 U.S. 414, 422-24, 105 S. Ct. 1421, 84 L. Ed. 2d 406 (1985), so that Morganti was only "transiently and fortuitously" on navigable waters, and was thus outside the coverage of the Act under the Fifth Circuit's holding in Bienvenu v. Texaco, Inc., 164 F.3d 901, 909 (5th Cir.1999); and 3) Morganti's job was "data processing" and thus specifically excluded from coverage under Section 2(3)(A) of the Act. The ALJ rejected Lockheed's argument that Cayuga Lake is not navigable.

Lorraine Morganti appealed to the Board, and was joined in her appeal by the Director of the Office of Workers' Compensation Programs ("Director") as a party in interest. The Board reversed all of the ALJ's conclusions, except as to the navigability of Cayuga Lake, and remanded to the ALJ. The ALJ resolved the remaining issues on stipulations, and at Lockheed's request, the Board summarily affirmed to make the order final and to permit review by this Court. Lockheed now brings the instant petition for review of the Board's decision.

We review the factual findings of the ALJ to determine if they are supported by substantial evidence. New Haven Terminal Corp. v. Lake, 337 F.3d 261, 265 (2d Cir.2003). We review questions of law de novo. Id. However, because the Director is the official charged with the administration of the Act, we defer to the Director's interpretations of the Act if they are reasonable and consistent with the statute. See American Stevedoring Ltd. v. Marinelli, 248 F.3d 54, 58 (2d Cir.2001); Fleischmann v. Director, Office of Workers' Comp. Programs, 137 F.3d 131, 136 (2d Cir.1998). Interpretations advanced in litigation, which are presented to us in the Director's brief and at oral argument, are more suspect. See Marinelli, 248 F.3d at 58 n. 2 However, the Supreme Court has indicated that agency interpretations presented in litigation may still be given deference so long as they are not post hoc rationalizations of past agency action or otherwise do not reflect the agency's fair and considered judgment. Auer v. Robbins, 519 U.S. 452, 462, 117 S. Ct. 905, 137 L. Ed. 2d 79 (1997). We therefore defer to the ALJ on factual findings, defer in specific circumstances to the Director's litigation interpretations of the Act, and review interpretations of case law and all other legal questions de novo.

I. The Longshore and Harbor Workers' Compensation Act

In 1917, the Supreme Court established that state workers' compensation acts could not cover non-local longshoremen who were injured while within maritime jurisdiction, as application of such acts would destroy the uniformity of regulation achieved by the federal government's maritime law power. Southern Pac. Co. v. Jensen, 244 U.S. 205, 214-16, 37 S. Ct. 524, 61 L. Ed. 1086 (1917), superseded by statute on other grounds. Cases following Jensen made clear that the line beyond which state regulatory power could not extend was the water's edge, a demarcation that became known as the Jensen line. See Director, Office of Workers' Comp. Programs v. Perini N. River Assocs., 459 U.S. 297, 306 n. 14, 103 S. Ct. 634, 74 L. Ed. 2d 465 (1983). In response, Congress enacted the Act in 1927 to provide compensation for injuries occurring on navigable waters in the course of employment by a statutory employer.1 Calbeck v. Travelers Ins. Co., 370 U.S. 114, 124, 82 S. Ct. 1196, 8 L. Ed. 2d 368 (1962). Under this regime, any worker, regardless of his or her duties or activities, was protected by the Act if injured on navigable waters. Perini, 459 U.S. at 311, 103 S. Ct. 634.

In 1972, Congress extended this so-called situs requirement by extending the statutory "navigable waters" to include certain adjoining land areas. Section 3(a) of the Act now provides coverage to employees who suffer disability or death due to an injury incurred on this expanded situs. 33 U.S.C. § 903(a). Congress simultaneously added a status requirement in Section 2(3) of the Act, limiting the Act's coverage to:

[A]ny person engaged in maritime employment, including any longshoreman or other person engaged in longshoring operations, and any harbor-worker including a ship repairman, shipbuilder, and ship-breaker, but such term does not include—(A) individuals employed exclusively to perform office clerical, secretarial, security, or data processing work ....

33 U.S.C. § 902(3). However, the Supreme Court held in Perini that the 1972 amendments were not intended to withdraw coverage from any person who would have been covered before 1972. 459 U.S. at 315, 103 S. Ct. 634. The status requirement of Section 2(3) thus limits coverage for injuries on those land areas brought within the statutory meaning of "navigable waters," but "when a worker is injured on the actual navigable waters in the course of his employment on those waters, he satisfies the status requirement in § 2(3)," unless he is excluded by "any other provision of the Act." Perini, 459 U.S. at 324, 103 S. Ct. 634. Thus, coverage under the Act still requires both situs and status, but certain kinds of situs, i.e., physical presence on actual navigable waters, will fulfill the status requirement.

Lockheed first argues that Morganti cannot meet the situs test because the waters of Cayuga Lake are not navigable waters.2 Under a test of long standing, waterways are navigable in fact "when they are used, or are susceptible of being used, in their ordinary condition, as highways for commerce." The Daniel Ball, 10 Wall. 557, 77 U.S. 557, 563, 19 L. Ed. 999 (1870). Lockheed conceded below that the Erie Canal is navigable, that the Erie Canal includes locks identical to the one that connects Cayuga Lake to the Erie Canal, and that Cayuga Lake is physically capable of supporting commercial water traffic.

Lockheed instead argues that the topography of Cayuga Lake and the economic conditions of the surrounding areas make commercial shipping unprofitable, and that the lake is therefore not susceptible of being used for commercial shipping, and thus is not navigable. In support of its argument, Lockheed relies primarily on LeBlanc v. Cleveland, 198 F.3d 353 (2d Cir.1999). In LeBlanc, we considered whether admiralty jurisdiction extended to a section of the Hudson River that ran through nine dams and was not presently used for commercial activity. Id. at 355. The appellant in that case argued that newer artificial obstructions should not be considered because navigability must be judged on the basis of the river's historic or natural state. Id. at 356-57. Citing several damming cases from other courts of appeals, we rejected the historical navigability argument, reasoning that in the absence of the potential for commercial shipping, there was no federal interest in providing a uniform body of law to govern activities on the relevant section of the Hudson River. Id. at 359; see also Livingston v. United States, 627 F.2d 165, 169-70 (8th Cir.1980);3 Chapman v. United States, 575 F.2d 147, 149-50 (7th Cir. 1978); Adams v. Montana Power Co., 528 F.2d 437, 439 (9th Cir.1975). We therefore held that navigability is limited to waters presently used or presently capable of being used for commercial activity, and that natural and artificial obstructions prohibiting such activity defeat navigability. LeBlanc, 198 F.3d at 359.

LeBlanc undermines rather than supports Lockheed's contention that economic conditions can render a waterway incapable of being used for commerce. LeBlanc clearly focuses on physical characteristics of the waterways themselves, holding that obstructions such as dams prevent any possibility of commercial activity. The Supreme Court, in applying the Daniel Ball test, interpreted the term "natural or ordinary conditions" by focusing on physical conditions: the volume of water as well as the gradients and regularity of flow. United States v. Appalachian Elec. Power Co., 311 U.S. 377, 407, 61 S. Ct. 291, 85 L. Ed. 243 (1940). The question of whether it would be economically difficult to profitably use Cayuga Lake commercially is a fundamentally different question that does not affect whether the lake is capable of being used for commerce.

Nor are we persuaded that the reasoning of LeBlanc supports Lockheed's argument. In that case, we focused on the federal interest in ensuring that events on the waterway in question were governed by a uniform body of law. 198 F.3d at 358 (citing Adams, 528 F.2d at 440-41). In the instant case, where the waterway is not physically obstructed but only economically inactive, the federal interest remains. Because Cayuga Lake is physically capable of supporting shipping, it is possible at any time for an interstate commercial vessel to enter the lake. Indeed, Lockheed's own evidence indicates that companies located near Cayuga Lake have seriously considered commercial shipping on the lake in the recent past. Any such shipping would give rise to the same federal interests as interstate commercial shipping anywhere else. If we were now to withdraw federal jurisdiction from Cayuga Lake, we would have to extend it again the moment that an interstate commercial ship entered the lake. In contrast, basing such jurisdictional changes on major events like damming or other physical alterations that make it physically impossible for a commercial vessel to travel on the waterway presents few difficulties. As Lockheed conceded below that there are no such physical alterations, we conclude that Cayuga Lake is navigable for purposes of the Act.

Lockheed raises two arguments to support its conclusion that Morganti was not a covered "employee" under Section 2(3) of the Act. First, Lockheed argues that Morganti was only transiently and fortuitously on actual navigable waters, and therefore should be excluded from coverage. Second, Lockheed contends that Morganti was employed exclusively to perform data processing work and is therefore explicitly excluded from coverage. We address these arguments in turn.

As discussed above, the Supreme Court has held that all employees injured while physically on actual navigable waters are maritime employees within the status requirement of Section 2(3) of the Act. Perini, 459 U.S. at 323-24, 103 S. Ct. 634. However, the Court "express[ed] no opinion whether such coverage extends to a worker injured while transiently or fortuitously upon actual navigable waters." Id. at 324 n. 34, 103 S. Ct. 634. The Court later "note[d] in passing" that there was "a substantial difference between a worker performing a set of tasks requiring him to be both on and off actual navigable waters, and a worker whose job is entirely land-based but who takes a boat to work." Herb's Welding, 470 U.S. at 427 n. 13, 105 S. Ct. 1421. Based on these "signals," Lockheed now asks us to follow the Fifth and Eleventh Circuits and create a transient and fortuitous exception to the general Perini rule. Under this exception, for an employee to be covered under the Act, a "not insubstantial" amount of the employee's time must take place on the water. Bienvenu v. Texaco, Inc., 164 F.3d 901, 908 (5th Cir.1999); see also Brockington v. Certified Elec., Inc., 903 F.2d 1523, 1528 (11th Cir.1990) (denying coverage to a land-based electrician who was injured on a boat en route to a one-time job site).

If we were to adopt this rule, Lockheed would still need to demonstrate that Morganti's presence on Cayuga Lake was in fact transient and fortuitous in order to prevail. Lockheed does not dispute that if Morganti's substantial time on the Paganelli is counted as presence on water, then Morganti could not be considered to be transiently and fortuitously on water. See Bienvenu, 164 F.3d at 908 (declining to set exact standards but holding that a worker who spent 8.3% of his employment time on actual navigable waters did not fall within the transient and fortuitous exception and was covered by the Act). Lockheed argues instead that the Paganelli is a fixed platform akin to an artificial island, and that Morganti's time on the Paganelli should accordingly be considered time on land. Only Morganti's time on the Little Toot II would then be considered time on water, and the brief commute time would constitute only a transient and fortuitous presence on water.

We need not decide in this case whether to adopt the transient and fortuitous exception, because we hold that as a matter of law,4 the Paganelli is not a fixed platform and that Morganti's time on the Paganelli must be considered time on actual navigable waters. The fixed platform rule was first recognized in Herb's Welding, 470 U.S. at 422-24, 105 S. Ct. 1421. In that case, the Supreme Court confirmed that an oil drilling rig was a fixed platform that should be treated as an artificial island, and that an employee injured on such a rig was therefore not covered by the Act. Id. However, Herb's Welding also notes that oil rigs came in both fixed and floating varieties, and that persons on floating rigs were covered by the Act or other maritime acts. Id. at 416 n. 2, 105 S. Ct. 1421. The obvious implication is that "fixed" means, at a minimum, "not floating." This is underscored by the reliance of Herb's Welding on Rodrigue v. Aetna Casualty & Surety Co., which held that fixed oil rigs were governed by a separate federal act intended to reach "the relevant subsoil and seabed and artificial islands and fixed structures erected thereon." 395 U.S. 352, 357, 89 S. Ct. 1835, 23 L. Ed. 2d 360 (1969) (internal alteration omitted). The clear import of "erected thereon" is that the seabed, rather than the buoyancy principle, must provide the fundamental support for the structure in question. See id. at 363 n. 9, 89 S. Ct. 1835 (noting that fixed platforms do not touch the waters except for legs which go through the water and into the ground). We therefore conclude that as a matter of law, a floating object cannot be a fixed platform or artificial island. Because it is undisputed that the Paganelli floats, we hold that the Paganelli is not a fixed platform, and thus that Morganti's time on the Paganelli was time on actual navigable waters.

We do not agree with Lockheed that the fixed versus floating distinction is arbitrary and unrelated to maritime interests, and would not be so persuaded even in the absence of what we view as controlling precedent. Obviously, the Paganelli is connected to the lakebed via mooring buoys that are themselves firmly anchored. But the chains and ropes that form this connection would just as obviously not prevent the prototypical maritime hazard: sinking. On the other hand, a structure directly supported by the lakebed via fixed, rigid posts or legs poses practically no such danger. This is precisely the distinction that prevents any boat from becoming a fixed platform the moment that it is anchored or moored. Similarly, it may be true that fixed oil rigs are eventually taken off their legs and floated away. But the rig is simply fixed while it is on legs; the fact that it once may have been, or one day will be, floating is of no consequence.

We are also unpersuaded by Lockheed's extensive citations, primarily from the Fifth and Ninth Circuits, uniformly demonstrating that not all floating structures are vessels. These cases are inapposite because the question of whether the Paganelli is a vessel is simply irrelevant. The Act, Perini, Herb's Welding, and even Bienvenu all look to whether the injured employee was on navigable water, not whether he was on a vessel. 33 U.S.C. § 903(a); Perini, 459 U.S. at 299, 103 S. Ct. 634; Herb's Welding, 470 U.S. at 416 n. 2, 105 S. Ct. 1421; Bienvenu, 164 F.3d at 908. The cases that Lockheed cites are drawn from a variety of contexts where vessel status is relevant, including salvage, the Jones Act, warranty of seaworthiness under the pre-1972 version of the Act, and Section 5(b) of the Act, which provides a separate, third-party cause of action based on the "negligence of a vessel."5

The one case from this Court that Lockheed cites is instructive, but not for its definition of vessel. In McCarthy v. The Bark Peking, we considered the question of whether a museum employee who was injured while working on a museum ship, permanently moored at the South Street Seaport Museum with its rudder welded into place, could recover under Section 5(b) of the Act. 716 F.2d 130, 132-33 (2d Cir.1983). We concluded that the ship was a vessel because it retained the residual capacity to transport cargo over water, even if only in tow. Id. at 134-35. However, before even reaching this question, we held that under Perini, the employee met the status test of Section 2(3) of the Act. Id. at 132-33. Our Perini status analysis preceded and was completely independent of our vessel analysis. We see no reason to now conflate these inquiries.

We therefore hold that a person on any object floating in actual navigable waters must be considered to be on actual navigable waters.6 As a result, Morganti was on actual navigable waters whenever he was on either the Little Toot II or the Paganelli, and was thus on actual navigable waters for thirty to forty percent of his work week. We therefore conclude that even if we were to adopt the transient and fortuitous exception to the Perini rule, Morganti would not fall within this exception.

We now address Lockheed's argument that Morganti was employed exclusively to perform data processing work and is thus explicitly excluded from coverage by Section 2(3)(A) of the Act. We note at the outset that Perini does not necessarily vitiate the Section 2(3)(A) exclusion. Perini specifically holds that a person who is injured on actual navigable waters is considered to be engaged in maritime employment. 459 U.S. at 299-300, 324, 103 S. Ct. 634. Section 2(3) provides that "[t]he term `employee' means any person engaged in maritime employment ... but such term does not include" persons excluded by Sections 2(3)(A) through 2(3)(H). 33 U.S.C. § 902(3). Thus, Morganti's presence on actual navigable waters permits him to meet the maritime employment test, but to qualify as an "employee" covered by the Act, he must still avoid the statutory exclusions. See also Perini, 459 U.S. at 324, 103 S. Ct. 634 (holding that a person injured on actual navigable waters meets the status test "providing, of course, that he ... is not excluded by any other provision of the Act"); but see Fleischmann, 137 F.3d at 135 (observing in dicta that a person can establish coverage by presence on actual navigable waters "without having to make any further showing... under § 902(3)"). We must therefore consider the data processing exception.

Because the question of whether Morganti was employed exclusively to perform data processing is a question of fact, we must uphold the ALJ's finding if it is supported by substantial evidence. However, the ALJ based his determination on a legal conclusion that the lack of any nexus between Morganti's employment and the loading or unloading of ships made Morganti an "office worker" excluded from coverage by the Act. Because this is a question of statutory interpretation, we do not defer to the ALJ.

Instead, we agree, regardless of any deference that we might owe the Director, with the Director's interpretation of the statute. The Director urges that for the Section 2(3)(A) exclusion to apply, "(i) the work must fit one of the enumerated positions; and (ii) the individual must perform that work `exclusively.'" Under Perini, Morganti must be considered to have been engaged in maritime employment, making the connection to loading or unloading cargo irrelevant.7 The plain language of the statute is then clear: as a maritime employee, Morganti must be considered an "employee" within the meaning of Section 2(3) unless excluded by Section 2(3)(A) or one of the other statutory exclusions. Section 2(3)(A), in turn, plainly requires both of the Director's urged elements.

Under this corrected test, there is no substantial evidence in the record to support a conclusion that Morganti was exclusively employed to perform data processing. We have not found any helpful case law on the meaning of "data processing." All parties rely instead on the definition of "data entry clerk" found in the Dictionary of Occupational Titles published by the Department of Labor, which reads:

Operates keyboard or other data entry device to enter data into computer or onto magnetic tape or disk for subsequent entry: Enters alphabetic, numeric, or symbolic data from source documents into computer, using data entry device, such as keyboard or optical scanner, and following format displayed on screen. Compares data entered with source documents, or re-enters data in verification format on screen to detect errors. Deletes incorrectly entered data, and re-enters correct data. May compile, sort, and verify accuracy of data to be entered. May keep record of work completed.

Dictionary of Occupational Titles ¶ 203.582-054 (4th ed.1991). That is, "data processing" means entering data into a computer, and checking and correcting such entries. The given definition does not seem to encompass any analysis of data, in the sense of examining the meaning of the data or drawing conclusions from the data.

It is not clear whether the question of whether Morganti was exclusively employed as a data processor turns on the totality of his employment, including not only on his activities on the Paganelli, but also his work in Syracuse. We note that this interpretation accords with a natural understanding of the term "exclusively employed" and with the legislative history of the provision. See H.R.Rep. No. 98-570, at 3, reprinted in 1984 U.S.C.C.A.N. 2734, 2736 ("The committee intends that this exclusion be read very narrowly ... if a clerical worker ... performed clerical work both in the office, and on the piers, wharfs or warehouses ... that worker would be covered by the Longshore Act in all phases of employment."). Lockheed itself identifies Morganti as an "engineer." Morganti's work in Syracuse, including full analysis of the test results obtained in his work on the Paganelli, identification of failures in the production process, and troubleshooting of such failures, fully bears out this title. It is clear that, looking at the totality of Morganti's employment, Morganti was not "exclusively" a data processor.

Even if we were to ignore Morganti's work in Syracuse, we would still conclude that there is no substantial evidence that Morganti was employed exclusively to perform data processing. On the contrary, Morganti's activities on the Paganelli, standing alone, clearly indicate his professional status. It is true that Morganti exercised little, if any, discretion in choosing which test to run or what parameters to enter. But taking the most obvious point, there is no factual dispute that Morganti verified the facial validity of each test that he ran by inspecting the test results. If Morganti were, for example, comparing a printout of the data he entered against the source data, that would be consistent with the work of a data processor. Instead, Morganti was examining the numerical and graphical test results to determine whether the readings supported the conclusion that the test had been correctly administered. This task is clearly an analytical one, and as a result, Morganti cannot be considered to have been exclusively employed to perform data processing.

We conclude that Morganti died in the actual navigable waters of Cayuga Lake, that his presence there was neither transient nor fortuitous, and that the employment that took him to the lake was not exclusively data processing. We therefore hold that Morganti satisfied both the situs and status requirements for coverage under the Act, and that respondent Lorraine Morganti is entitled to death benefits under the Act. For the foregoing reasons, the petition for review of the decision of the Benefits Review Board is denied.