Dennis Fryman v. United States

EASTERBROOK, Circuit Judge. A nasty flood of the Mississippi River in 1927 led to the Flood Control Act of 1928. Congress authorized a complex system of dams and diversionary channels to reduce the damage high water produces. Any system of flood control shifts the incidence of loss even if it substantially reduces the net damage. People upstream of the dams will *80 be inundated although safe before; people downstream face spectacular loss if a dam fails or the floodgates must be opened at an inopportune time. These persons, new victims of floods at the same time flood-control programs rescue others, may be expected to complain. One part of the package in 1928, 33 U.S.C. § 702c, closes courthouse doors to them. It provides that “[n]o liability of any kind shall attach to or rest upon the United States for any damage from or by floods or flood waters at any place”. Although § 702c eliminates any remedies the courts might otherwise provide (in 1928 there were none, given sovereign immunity), Congress regularly enacts special legislation offering succor to victims of floods, whether or not the injuries were caused or augmented by federal flood-control works. Assistance in the billions of dollars has become routine. People living in flood plains must count the exchange a gain. Judicial relief would take much longer and would not match the generosity Congress has displayed. Section 702c now plays a role no one could have anticipated in 1928. After the Federal Tort Claims Act of 1946 waived a big chunk of sovereign immunity, § 702c became an important defense. It matters in a class of cases for which Congress does not enact special legislation: boating accidents on the lakes impounded by flood-control dams, drownings, and in this case a diving accident that rendered a teenager quadriplegic. Dennis Fryman and his brother Terry went swimming on August 7, 1983, near an island in Lake Shelbyville in Illinois. The lake was created between 1963 and 1970 as part of a flood-control project, and like many other artificial bodies of water was put to recreational use. The Army Corps of Engineers, which runs the flood-control operations, established the Wilborn Creek Recreational Area along the lake. People use the area for boating, water skiing, swimming, fishing, and picnicking. The island Fryman visited is submerged when the water level in the lake is high, but when it appears it is a popular site for picnics. Fryman attempted a shallow, racing dive into the water off the island; his head struck a concealed hazard, which broke his neck and left him paralyzed. The complaint (the only source of facts in the case) does not reveal what kind of object Fryman struck or whether he looked before leaping. Fryman’s administrative claim under the Tort Claims Act was rejected; his suit contends that the Corps of Engineers acted not only negligently but also maliciously in not posting warning signs or closing the island to recreation. The district court dismissed the suit under Fed.R. Civ.P. 12(b)(1), holding that in light of § 702c there is no federal jurisdiction (because the United States has not waived its sovereign immunity). Section 702c is written broadly in the sense that it covers “any” damage, but it might have been possible to understand the scope of the statute — “floods or flood waters” — as referring only to the escape or overflowing of water. Recreational injuries within flood-control projects then would be covered by the Tort Claims Act rather than § 702c. United States v. James, 478 U.S. 597, 106 S.Ct. 3116, 92 L.Ed.2d 483 (1986), considers and rejects this possibility. Boaters were swept to their deaths when waters within a flood-control and recreation project were released. The Court concluded that § 702c applies to such injuries, because it is broadly written. It read the critical phrase, “floods or flood waters”, broadly indeed, 478 U.S. at 605, 106 S.Ct. at 3121: The Act concerns flood control projects designed to carry floodwaters. It is thus clear from § 702c’s plain language that the terms “flood” and “flood waters” apply to all waters contained in or carried through a federal flood control project for purposes of or related to flood control, as well as to waters that such projects cannot control. Accordingly, every drop of water in Lake Shelbyville is “flood water” within the meaning of § 702c, and the statute blocks recovery if the injury is “damage from or by” these waters. Fryman says that the injury was inflicted by a combination of the hidden *81 object and the Corps of Engineers’ failure to warn him off. Water, Fryman insists, has nothing to do with his paralysis. By contrast, the water in James was the culpable actor, seizing the boat and dragging the passengers under. Attributing motives and deeds to water is not a way around James. In both cases the responsible actors were federal employees, and the negligence (if any) lay in not taking precautions. Water played some role in both — in James creating the fatal current, in this case concealing the obstacle and creating the motive for the dive. The government could have kept boats off the lakes involved in James, or it could have kept boaters farther back from the sluices; decisions concerning the recreational use of the lakes led to the deaths, just as (Fryman contends) decisions concerning the recreational aspects of Lake Shelbyville led to his injuryRecharacterizing causes as Fryman proposes would enable parties to avoid James by playing word games. It was not a pleading defect that led the plaintiffs to lose James; it was the fact that the deaths occurred as a result of a risk of running a flood-control project. The Court rejected both the argument that injuries from the recreational components of a flood-control project are outside § 702c, 478 U.S. at 609-10, 106 S.Ct. at 3123-24, and a contention that losses attributable to failure to warn of hazards are not covered, id. at 610-12, 106 S.Ct. at 3123-25. The Court described § 702e as barring “ ‘any’ liability associated with flood control”, 478 U.S. at 608, 106 S.Ct. at 3122, and any liability arising out of the “ ‘management’ of a flood control project”, id. at 610, 106 S.Ct. at 3123. The Court found it hard to conceive how § 702c could have been more broadly written; we find it hard to conceive how a decision interpreting this statute could have been more broadly written. James was so broadly written that it cannot be applied literally. The “management of a flood control project” includes building roads to reach the beaches and hiring staff to run the project. If the Corps of Engineers should allow a walrus-sized pothole to swallow tourists’ cars on the way to the beach, or if a tree-trimmer’s car should careen through some picnickers, these injuries would be “associated with” flood control. They would occur within the boundaries of the project, and but for the effort to curtail flooding the injuries would not have happened. Yet they would have nothing to do with management of flood waters, and it is hard to conceive that they are “damage from or by floods or flood waters” within the scope of § 702c. See also James, 478 U.S. at 605 n. 7, 106 S.Ct. at 3121 n. 7, implying that § 702c is irrelevant to some injuries within the boundaries of a project by quoting from two cases saying that when the injury is “wholly unrelated” to flood control, the statute does not apply. The Tenth Circuit employed this rationale in Boyd v. United States, 881 F.2d 895, 899-900 (1989), to hold that § 702c does not bar liability if the Corps of Engineers should have excluded boats from a flood-control lake. A boat speeding through the lake struck and killed a snorkeler; this death, the court thought, had nothing to do with flood control. Fryman relies on Boyd and Hayes v. United States, 585 F.2d 701 (4th Cir.1978), one of the cases the Court cited in note 7 of James. (We make nothing of the Court’s citation to Hayes; the same footnote also cites Morici Corp. v. United States, 681 F.2d 645, 647-48 (9th Cir.1982), which expressly rejects the holding of Hayes. It is unlikely that the Justices read these cases, and citing inconsistent holdings does not endorse either.) Just as some language in James is too broad, so Boyd has gone too far in the opposite direction. According to the Tenth Circuit, § 702c does not apply at all to “liability associated with operating a recreational facility”, 881 F.2d at 900. If that is so, then James was wrong on its own facts, for it grew out of recreational boating, and in James no less than in Boyd the deaths could have been prevented by closing the lakes to recreational boating. If § 702c has limits, they have to do with causation. In Boyd the court could have asked, not whether boating and snorkeling are “recreational”, but whether the injury *82 stemmed from features that differentiate flood-control lakes from purely recreational lakes. Boaters in James drowned because the imperatives of flood control led to the release of water from the lake. Collisions between swimmers and powerboats, by contrast, occur on natural lakes and purely recreational artificial lakes; none of the flood-control activities increased the probability of such a run-in. Perhaps, however, an accident is “wholly unrelated” to flood control only when the existence of the project does not increase its probability. In that case, simple drownings are encompassed by § 702c because, but for the project, there would be no water, and but for water no one would swim and drown. Two courts of appeals have reached this conclusion. McCarthy v. United States, 850 F.2d 558, 562 (9th Cir.1988) (diving accident) (unless the injury is “wholly unrelated” to the flood-control project, § 702c applies); Mocklin v. Orleans Levee District, 877 F.2d 427, 430 & n. 6 (5th Cir.1989) (drowning; same rationale). Two others, although sympathetic to this approach, rely in addition on the fact that particular injuries were made more likely by the use of the lake in flood control. De Witt Bank & Trust Co. v. United States, 878 F.2d 246 (8th Cir.1989) (diving accident); Dawson v. United States, 894 F.2d 70 (3d Cir.1990) (drowning). James did not adopt the “wholly unrelated” language from Morid; it only quoted it in a footnote. Courts assuming that these words are “the test” may be jumping the gun, and at all events wholly unrelated to what —the project, or the risk that led to the particular accident? — is an unsettled question. Flood-control activities increase the probability of injuries such as Fryman’s. Lake Shelbyville rises as flood waters enter in the spring and falls as water is released in the summer and fall to maintain constant rates of flow through the rivers downstream. During the spring and part of the summer, the island from which Dennis Fry-man leapt is under water. After it emerges (in 1983, two weeks before the accident) its shoreline changes with the lake level. This makes it hard to identify places hazardous to divers; which spots are safe and which dangerous depends on the level of the lake. Constantly changing shorelines also are hard to plaster with signs. Unless the Corps of Engineers were to move the signs frequently, it would have to establish ranks of signs at different depths, which would appear as the lake fell. Signs themselves might be dangerous. The signs at higher elevations might need to warn divers of other signs still beneath the waves. (“DANGER! Concealed Warning Signs!”) Only preventing all visits to the island would solve the problem. Yet because "the problem” is attributable in substantial part to the nature of a flood-control project, a claim based on the theory that the Corps of Engineers should have closed the island to swimming in order to protect recreational users is the same in principle as the claim that the Corps should have closed to boaters the lakes in James. Although the district court dismissed Fryman’s case on the pleadings, there is no dispute that the water level of Lake Shel-byville is constantly changing, and that the island is indeed submerged for part of the year. Because Fryman concedes this (a concession his lawyer repeated at oral argument), and because a plaintiff may plead himself out of court, we affirm the judgment on this basis. We do not decide whether a recreational injury on a flood-control lake could be actionable under the FTCA when the flood-control activities did not increase the probability of the injury compared with a natural lake devoted to recreational use. Resolution of that question must wait until the answer makes a difference. Affirmed.