The Katrina litigation

In 1965, Hurricane Betsy flooded New Orleans. "Levees for the Mississippi River Gulf Outlet along Florida Avenue in the Lower Ninth Ward and on both sides of the Industrial Canal failed," with predictable consequences: almost half of New Orleans is at or below sea level, and so the effect of the levee breach was to flood a significant area of the city. The same year, Congress passed the Flood Control Act of 1965, for (among other purposes) “the control of destructive floodwaters” and “hurricane-flood protection on Lake Pontchartrain.”1 This authorized, inter alia, a “joint federal, state, and local effort” to “construct a series of control structures, concrete floodwalls, and levees to provide hurricane protection to areas around Lake Pontchartrain” under the principal authority of the Army Corps of Engineers.2

Forty years later, in late summer 2005, Hurricane Katrina made landfall close to New Orleans. During the storm, a portion of the 17th Street Canal wall broke, with even more catastrophic consequences than those attending Hurricane Betsy: “By August 31, … eighty percent (80%) of New Orleans was flooded due to [levee] breaches, with some parts under 15 feet of water.” As people began to return to pick up the pieces of shattered homes and lives, they called on insurance policies they expected to provide funds to begin the process of rebuilding. But several insurance companies said no dice: their policies did not cover damage caused by “flood.”

Such is the background for the litigation by victims of the Katrina flooding to recover moneys they feel the insurance companies owe them. Two weeks ago, the Fifth Circuit heard an appeal in the case that made it onto the MSM’s radar; I don’t feel I can add anything in terms of legal analysis to the case, so this post - written mainly to help me understand the litigation myself - is aimed at explaining to interested lay readers the story so far. I invite corrections, with the caveat that for the intended audience, some of the finer legal points have been glossed over, and so an omission is not necessarily an error.

(Fair warning: A significant part of this post pertains to Civil Procedure; you have been told that Civil Procedure is death, that it is dry and dull, but I tell you that it is fun and sexy, and alas, this post requires only the showing of a bit of leg, not a centerfold.)

In October 2005, eight plaintiffs filed suit in State Court against their insurers and the Board of Commissioners for the Orleans Levee District for the Parish of Orleans. The claims against the insurance companies boil down to “give us the money you owe us,” while the claim against the Commissioners was that they had been negligent “in failing to discover, correct and notify others including [plaintiffs] of the break in the levee wall,” rendering them liable to the plaintiffs.3

Two months later, exercising their option to remove to the federal court system an action brought against them in state court that could have been filed in federal court,4 the insurer defendants removed the case to the U.S. District Court for the Eastern District of Louisiana. The insurers then moved on Valentine’s Day 2006 to sever the claims against them from the claims against the Commissioners,5 a motion the court granted on June 1st, “and finding no jurisdiction over [the Commissioners], remanded those claims back to” the Louisiana Courts.6

Although the foregoing is a slightly longer walk than normal to the doors of the federal courts, the same principles apply as would if they had simply filed there in the first place. The defendants moved for judgment on the pleadings (which the court treats as a motion to dismiss for failure to state a claim on which relief can be granted7): defendants can seek to prevent a case going to trial if it involves no disputed question of fact for a jury to find, and that no matter what facts the plaintiffs might prove at trial, they cannot prevail, and the defendants should thus be spared the difficulty and expense of a trial.8 The insurers so moved, arguing that whatever facts the plaintiffs might hope to prove at trial, the fact would remain that the policies precluded recovery for flood, and the damage to their property was caused by flooding. The District Court concluded that with two exceptions, the insurers were not entitled to a dismissal on the pleadings.

The appeals in the Fifth Circuit stem from this 12(b)(6) motion: an interlocutory appeal from the insurers on the refusal to grant their motions for summary judgment and a cross appeal from the plaintiffs on the district court’s grant of summary judgment to the defendants State Farm insurance. It’s important to understand the procedural posture of the case. First, in terms of evidence: when a party moves for a 12(b)(6) dismissal, proceedings on that motion work from several presumptions: courts are obliged to assume that the facts alleged in the complaint are true, to view the pleadings in the light most favorable to the non-moving party (i.e. the plaintiffs in this case), and draw every reasonable inference in favor of that party. Second, in terms of result: the question here is not “who wins” as such, but whether there is a dispute that can be resolved by allowing the litigation to proceed to discovery and trial. That is, is this purely a question of applying law to undisputed facts, or are there facts that may come out at trial that change the outcome?

The key issue is the wording of the exceptions to the insurance policies. There are five defendants, but only three insurance policies at issue. The defendants are Standard Fire, Hanover, Unitrim, State Farm, and Hartford, but the former three use a “standardized insurance policy language” provided by the Insurance Service Office.9 The so-called “ISO policies” provide the exclusion that the policies

do not insure for loss caused directly or indirectly by any kind of the following. Such loss is excluded regardless of any other cause or event contributing concurrently or in any sequence to the loss.

. . .

(c) Water damage, meaning:
Flood, surface water, waves, tidal water, overflow of a body of water, or spray from any of these, whether or not driven by wind[.]10

State Farm’s policy is comparable, but more specific and explicit in its exclusions: they reserve that the

do not insure under any coverage for loss which would not have occurred in the absence of one or moreof the following excluded events. We do no insure for such loss regardless of: (a) the cause of the excluded event; or (b) other causes of the loss; or (c) whether other causes acted concurrently or in any sequence with the excluded event to produce the loss; or (d) whether the event occurs suddenly or gradually, involves isolated or widespread damage, arises from natural or external forces, or occurs as a result from any combination of these:

. . .

c. (1) flood, surface water, waves, tidal water, overflow of a body of water, or spray from any of these all whether driven by wind or not.11

Lastly, Hartford’s policy includes the ISO policy, but supplements it:

Even if any of the terms of this policy might be construed otherwise, the following [causes of physical loss or damage to property] … are SPECIFICALLY EXEMPTED FROM THIS POLICY. WE DO NOT COVER [OR] INSURE AGAINST LOSS OR DAMAGE DIRECTLY OR INDIRECTLY CAUSED BY, RESULTING FROM, CONTRIBUTED TO OR AGGRAVATED BY, OR WHICH WOULD NOT HAVE OCCURRED BUT FOR … [t]he design, specifications, workmanship repair, construction, renovation, remodeling, grading or compaction of … [inter alia] levees[,] … [or] the maintainance of such facilities[,] [or the] COLLAPSE, CRACKING OR SHIFTING of buildings, other structures or facilities, or their parts, if [such damage] … [o]ccurs during … [inter alia] flood conditions … [and] [w]ould not have occurred but for … [the] flood. 12

The Hartford policy goes on to define flood as including (as the common usage of the term would suggest) “flood, surface water, waves, tides, tidal water, tidal waves, high water, and overflow of any body of water, or their spray, all whether driven by wind or not; … [and the] [r]elease of water held by a dam, levy (sic) or dike by a water or flood control device….”13

At first glance, it seems counterintuitive at best that policies specifically excluding damage caused by “flood” could not exclude damages from the flooding that followed the levee breaches in New Orleans. Indeed, one struggles to find an alternative term to describe the sudden presence of several feet of water in a usually dry area, and most media outlets have not exactly gone to the mats to change the meaning of the term in ordinary discourse to accommodate the plaintiffs’ theory. In the immediate aftermath of the levee breaches, America's paper of record was not shy with the F word, and National Geographic reported - under the headline "New Orleans Flooded in Wake of Hurricane Katrina" - that "[b]roken levees are allowing floodwaters to pour into New Orleans." At the same time, the New Orleans Times-Picayune produced a flash animation titled "Flash flood: Hurricane Katrina's Inundation of New Orleans" (it helpfully shades the areas inundated with water in blue, labeling this in the legend "flooded land"). MSNBC warned that the Katrina aftermath posed "health risks [a]ris[ing] with flooding," and observed that "[i]n some of the low-lying areas most heavily hit by flooding it could be weeks before residents can return to their neighborhoods"; and NIOSH produced "Hazard Based Interim Guidelines: Protective Equipment for Workers in Hurricane Flood Response." Even a Washington Post article giving credence to the plaintiffs’ allegations of negligence on the part of the Commissioners took as axiomatic that the consequences of the alleged negligence was "the flooding of most of New Orleans," and reported this under the headling "Experts Say Faulty Levees Caused Much of Flooding" (emphases added).

Nor is the understanding that a breached levee results in a flood specific to New Orleans: shortly after the flood, USA Today noted that other areas of America are at "flood risk" should various "deteriorating" levees fail, and referred to the inundation of 12,000 acres of California farmland following a levee breach the year before Katrina as a "flood[]." This understanding has not changed in the intervening two years; barely a month ago, USA Today noted that "[h]omes in the lower ninth ward [of New Orleans] were destroyed by flood waters from a levee breach at the Industrial Canal in the wake of Hurricane Katrina." This is merely to scratch the surface of the coverage. Even one sympathetic to the plaintiffs’ position must feel daunted by the mountain they might appear to have to climb.

However, all is not lost for the plaintiffs. The basis for federal jurisdiction over the parties in this case is diversity of citizenship (i.e. the plaintiffs live in Louisiana, while the defendants are incorporated in other states), which means that although the case is heard in Federal court, the law that is applied to the case is the substantive law of the forum state,14 so in this case, the controlling law is Louisiana principles of contract, wherein, “[a]s a general rule, insurance policies should be construed to effect, not deny coverage,”15 and “[w]ords [in a contract] susceptible of different meanings must be interpreted as having the meaning that best conforms to the object of the contract.”16 “[U]nder Louisiana law, unless there is a specific exclusion for the type of water damage that an insured has incurred, coverage is presumed under [such policies as] these … [and] the insurer bears the burden of proving the applicability of an exclusionary clause within a policy.”17

As we saw in Part II, the policies at issue disclaim liability in the event of “flood.” However, while “[i]f the policy wording at issue is clear and unambiguous[] … the insurance contract must be enforced as written,”18 and although generally “[w]ords and phrases used in an insurance policy are to be construed using their plain, ordinary and generally prevailing meaning,”19 nevertheless, “if the text is susceptible to two or more reasonable interpretations,”20 then such “[a]mbiguous policy provisions are generally construed against the insurer and in favor of coverage.”21 This applies with particular force when faced with a standardized agreement such as an insurance contract: a “contract executed in a standard form of one party must be interpreted, in case of doubt, in favor of the other party.”22

Thus, at least in the context of rebutting a 12(b)(6) motion where every reasonable inference will necessarily be drawn in favor of (in this case) the plaintiffs, all the plaintiffs need do to prevail at this stage is to inject ambiguity into the word “flood.” If they can provide a reasonable interpretation of “flood” that is both within the ordinary meaning of the term yet excludes the deluge from the failure of the levees,23 then a row of dominoes falls under state law, because “‘[i]f there is any doubt or ambiguity as to the meaning of a provision in an insurance policy, it must be construed in favor of the insured and against the insurer….”24

The plaintiffs hope to do this by arguing that the flood was caused by human negligence. As their petition for leave to appeal puts it,

New Orleans was destroyed not by an Act of God, but rather from the greed and negligence of man. Lake Pontchartrain and the other surrounding waterways did not flood the City. It was the failure of … [the Commissioners that flooded the city, and] [d]amage caused to property from a third-party’s negligence is covered under the … [policies] at issue.25

Or as the District Court put it:

Simply put, the question before the court is whether it is reasonable to find that in the absence of further definition or pvovision in [the exclusions for “flood”] … - one which encompasses both a “flood” which occurs solely because of natural causes and a “flood” which occurs because of the negligent or intentional act of man and one which limits itself only to a flood which occurs solely because of natural causes.26

In other words, the plaintiffs argue that “flood” should be read only to include flooding that resulted from events over which humans had no possible control, but the levees failed as a result of human failure, and since the flood was thus caused by human hand, was no flood at all.

Occasionally, it becomes necessary for a textualist to recognize that “dictionaries have a limited capacity to record the nuances of usage,”27 and thus to read a word more broadly than its dictionary definition would suggest, since in popular usage it has acquired a somewhat different meaning. This case presents almost a photonegative of that situation: plaintiffs argue that the ordinary or common meaning of the term “flood” should be set aside in favor of the inscription on its casket in Judge Easterbrook’s mausoleum.28

In trying to make sense of the case, the district court turned to several dictionaries and prior rulings in other states, and concluded that yes, there is a reasonable alternative definition of “flood” that would (a) exclude flood resulting from human negligence and (b) include only inundation arising from the “overflow” (i.e. rising and overtopping) of a body of water, which the court assumes (although does not assess) requires a rise in water level rather than a falling of the level of its container.29 “Most of the definitions of the noun [“flood”],” says the court, “imply encroachment caused by an act of nature,”30 dwelling in particular on the presence in most definitions of “flood” of the term “overflow.”

The court adds that to conclude otherwise “leads to absurd results,”31 but fails to elaborate on what exactly it means, leaving the reader to wonder if the court understands the absurdity doctrine.32 A result that is arguably unjust is not absurd. Absurdity in this sense would generally mean an internal logical inconsistency, or at least a result plainly and beyond any possible dispute at odds with the purpose of the contract, not merely an unpalatable result. For example, it would be absurd to construe a flood insurance contract to exclude flood damage, but it is hardly absurd to construe a general insurance contract to exclude flood damage when it explicitly excludes flood damage, and a fortiori when flood damage insurance is available and widely taken out.33

In any event, applying these principles to the three policies noted in Part II, the court found that the term “flood” appearing in the ISO policies was ambiguous, and the domino trail noted in Part III(b) duly falls for these defendants; thus, the court denied motions for judgment on the pleadings by Standard Fire, Hanover, and Unitrim, leading to their appeal.34 But State Farm’s policy, “does precisely what the ISO Water Exclusion Policy fails to do. It makes clear that regardless of the cause of the flooding, there is no coverage provided for any flooding….”35 This language rendered the meaning of flood, in the context of State Farm’s coverage, unambiguous, and the court duly granted its motion for judgment on the pleadings, leading to the plaintiffs’ cross appeal. The Hartford policy goes even further; the court concluded that it is “unequivocal[]” and “leaves nothing to the imagination.”36 Since their policy explicitly disclaims liability in the precise circumstances at issue here, the court granted Hartford’s motion to dismiss, and the correctness of this ruling seems to be born out in the failure of plaintiffs to cross appeal this point.

Thus, in a manner of speaking, the District Court has dared the Fifth Circuit to find that an alternative definition that arguably exists in the dictionary can possibly be unreasonable (recall the plaintiffs need only “two or more reasonable interpretations” to topple that first domino in the chain).

Update, 8/7/07: The 5th Cir doesn't take the bait.

1. 1. 79 Stat. 1073 §204 (1965 U.S.C.A.A.N. 1065, 1079 (89th Cong. 1st Sess.)).
2. 2. See http://www.gao.gov/new.items/d051050t.pdf (mirrored at http://www.levees.org/research/sources/gao%20Testimony.pdf).
3. 3. Complaint at ¶ 8.
4. 4. 28 U.S.C. § 1441.
5. 5. Defendants agued that “[t]he claims against the Insurer Defendants and the Levee Board are improperly joined. The allegedly totious conduct by the Levee Board is entirely separate and distinct from the alleged breaches of contract by the Insurer defendants … [and] Plaintiffs do not allege that the liability of the Indurer defendants is premised … on the alleged negligence of the Levee Board.” Even if this were not so, defendants argued that in the service of “judicial economy and efficiency, it made a great deal of sense to sever the pure question of law presented by the claims against the insurers from the “myriad factual and legal issues” presented by the claims against the commissioners. Motion to sever at 2-3.
6. 6. District Court Order and Opinion on Defendants’ motions for summary judgment11/27/06 (“Op”). at 4.
7. 7. Id. at 6-7.
8. 8. Federal Rule of Civil Procedure 12(b)(6).
9. 9. Op. at 15 n.4 and accompanying text. The ISO promulgates such products “to help the insurance industry to measure, manage and reduce risk.” Ibid.
10. 10. Id. at 15 (emphasis in original)
11. 11. Id. at 16 (footnote omitted).
12. 12. Id. at 16-7.
13. 13. Id. at 18.
14. 14. Erie RR. Co. v. Tompkins, 304 U.S. 64 (1938).
15. 15. Op. at 40 (citing Holden v. Connex-Metalna, 2001 WL 40994 (E.D. La. 2001)).
16. 16. Ibid. (quoting La. Civ. Code. Art. 2048).
17. 17. Id. at 9-10.
18. 18. Id. at 8 (quoting Cadwallader v. Allstate Ins. Co, 848 So. 2d. 577 (La. 2003) (citations omitted).
19. 19. Ibid.
20. 20. Id. at 10 (emphasis in original).
21. 21. Id. at 10 (citing La. Civ. Code Art. 2056).
22. 22. Ibid.
23. 23. The verbal contortions required to find an alternative way to write “that excludes the flooding that resulting from the failure of the levees” in this sentence demonstrates just how weak the plaintiffs’ case would be were the bar not so extraordinarily low (i.e. the existence of any reasonable alternative interpretation will suffice). As a matter of ordinary usage of the term “flood” in the English language, the plaintiffs’ premise is preposterous on its face. Of course, it's far from impossible to design language in a policy that explicitly shuts out flooding from levee failure, as the Hartford policy demonstrates. The insurance companies certainly know that the levees exist, that they are man-made, and that they might fail. Indeed, the possibility of levee failure was far from unthinkable: it was thought, almost ten years ago. "A 1998 ruling, by an administrative judge for the Corps' Board of Contract Appeals, shows that the
    contractor, Pittman Construction, told the Corps that the soil and the foundation for the walls were
    'not of sufficient strength, rigidity and stability' to build on."
24. 24. Id. at 41 (quoting ).
25. 25. Plaintiffs’ Petition for leave to cross-appeal, 2/1/07 (“Pet.”) at 2.
26. 26. Op. at 24.
27. 27. J. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387, 1258-9 (2003).
28. 28. The allusion is to F. Easterbrook, Text, History and Structure in Statutory Interpretation, 17 Harv. J. L. & Pub. Pol. 61, 67 (1994) (“a dictionary … [is] a museum of words, an historical catalogue….”).
29. 29. Op. at 18-39.
30. 30. Id. at 39
31. 31. Id. at 41.
32. 32. See generally Manning, supra; A. Gold, Absurd Results, Scrivener's Errors, and Statutory Interpretation, 75 U. Cinn. L. Rev. 25 (2006).
33. 33. “[n]early half the [Katrina] victims did not even have flood insurance.” If approximately half of the victims spent extra time and money obtaining flood insurance, is it really reasonable to claim that no reasonable person could possibly have thought that a general insurance contract would exclude flood damage? The fact that I spend extra money each month on earthquake insurance implies my understanding that my regular policy does not cover earthquake damage.”
34. 34. Op. at 39-49.
35. 35. Id. at 50.
36. 36. Id. at 53.