"Cracking the Floodwall Code"
Stanley A. Millan
20 Tulane Environmental Law Journal 169

General – When Levees Fail

“Where there are men, there can be no peace”[1].  Wars do not have to be fought with weapons of mass annihilation.  Lies, broken promises, and doubt can bring their own seeds of destruction.  Were New Orleans’ levee systems enigmas or props?  Is New Orleans the new Pompeii?  Only hope and perseverance can survive.  This essay may bring a ray of hope to the legal minded in the flood fights of both present and future. 

People are looking for someone or some entity to blame, scapegoat or real culprit, for the failed levees, water works and floodwalls in Orleans and adjacent parishes during Hurricane Katrina’s aftermath in 2005.  The storm became a meteorological enema to the cultural, social and physical environment of this area.  But who inserted the poison pill?  This is not just about anger but about economic survival.  Class actions, FEMA, Write Your Own (Flood) Policy Companies, homeowner insurers, the United States Army Corps of Engineers, public works contractors, and local and state governments have turned flood victims into predators, and these entities are their prey.  No one knows yet how successful plaintiffs will be legally or if they just will have to rely on good will and the public fisc for help.

However, of the many entities that could be named, certainly the Corps of Engineers tops most public lists, from newspapers and scientists, to lawyers.  The Corps is providing most of the funds for what flood control measures can be fixed in the near future, but longer term measures remain uncertain.  The Federal Tort Claims Act (FTCA) and inverse condemnation are two causes of action that I will address preliminarily and that relate primarily to the Corps.  Certainly state analogs, maritime law and private tort claims can apply to others, but the focus here will be on the Corps.  (An environmental statute is another option and it is mentioned in footnote 37, but is otherwise beyond the scope of this essay.)


This Act[2] provides a limited waiver of federal sovereign immunity when federal employees are negligent and acting within the scope of their employment.  The United States can be sued under circumstances where the United States, if a private person, would be liable in accordance with applicable state law. 

To sue, one must first timely file an administrative claim within two years after a claim accrues.[3]  The government then has a six-month time period to review the claim before suit in federal district court can be filed by a plaintiff.[4]  Failing in this administrative claim step can result in dismissal of a FTCA lawsuit.  Alleging a sum certain in property and personal injury damages is a key part of an administrative claim or else a subsequent FTCA lawsuit is in jeopardy.

The Act focuses on negligence and excepts from its coverage intentional torts, such as assault, battery, deceit, negligent and deliberate misrepresentation, etc.[5] 


Further, the Act excludes certain claims arising out of an act or omission of a federal employee exercising due care in the execution of a statute or regulation, valid or not, and discretionary functions.[6]  The former exception has limited application, but the latter discretionary function exception is broader.  The federal government is thus not liable for its actions and decisions based on public policy considerations and judgmental choices.[7] 

If governmental judgment or choice is involved based upon social, economic or political considerations, an FTCA claim will be barred.  For example, the government (and its contractors) were immune from a tort suit for designing a “death trap”, e.g., it was apparently too costly to design a helicopter emergency escape that would open under water.[8] 

Negligence is irrelevant to discretion.[9]  However, if the agency fails to act in accord with a mandatory statutory or regulatory directive, the discretionary function exemption does not apply.[10]  Formal agency policy may also suffice.[11]  Guidelines are not legally enough here.[12] 

Therefore, federal tort claims based upon allegations that the Corps did not build higher levees than authorized, not implement levee repairs at a quicker or more even pace, not elect to armor all levees, not sooner close certain public waterways or canals, not quickly provide enough pumping capacity for closed or narrowed drainage canals, and not favor some communities’ risk over navigational or “efficiency” interests, could face a discretionary function challenge by the United States in court, unless the government violated statutes, rules, or binding policy in its decisions or inaction.

Levees are not under the tight scrutiny of a law like the National Dam Safety Program Act.[13]  It is still open to question whether parts of the New Orleans area federal flood control projects were designed and built according to Congressionally mandated standards.[14]  The recently resigned Chief of Engineers had admitted that “We have now concluded we had problems with the design of the structure [e.g., the 17th Street Canal Floodwall].  We had hoped that wasn’t the case, but we recognize it is the reality”, and that “…we had a catastrophic failure with one of our projects”. [15]  For instance, were floodwalls and levees vigorously designed and constructed to substantially meet the Congressionally approved plans and recommendations of the Chief of Engineers in the Flood Control Act of 1965, for the Lake Pontchartrain and Vicinity Hurricane Protection Project, [16] e.g., the project storm, now called a Saffir-Simpson Hurricane Scale Category 3 hurricane?[17]  This is a statute that the Corps has no discretion to violate.  There lies the rub.  The debate persists over whether flooding was caused by levees being overtopped or pushed over by a greater hurricane (category 4) or was most of the flooding caused by levees and floodwalls that were not Category 3 survivable.[18]  Failing to take into account area subsidence or water pressure on floodwalls certainly sounds like a negligent failure to meet the standards of the 1965 Act.  [19] 


More pointed, the Flood Control Act of 1928,[20] bars claims against the United States “…for any damage from or by floods or flood waters at any place”.  Though derived from the great Mississippi River flood of 1927,[21] courts have, rightly or wrongly, broadly applied this immunity geographically.[22]  Courts had also focused on immunizing the Federal government exclusively from damages by all flood control projects.  In 2001, the United States Supreme Court in Central Green Co. v. U.S.,[23] changed the Act’s immunity analysis from the character of the project (flood control or not) to the character and purpose of release of the waters that caused damage, e.g., a release of flood waters from a reservoir.  In Central Green, it was argued that it was irrigation, not flood  water that allegedly caused subsurface property damage, but the Court remanded the issue to see if immunizing flood waters were also involved.  Simply put, not all damaging water is “flood or flood waters” under the 1928 Act; but non-flood waters can still cause FTCA damages.  Some waters allow for governmental immunity, and others create governmental liability.

Thus, the government’s blowing a levee to flood area A and spare area B, a la Rising Tide, is likely an immunized “flood,” but storm waters from canal seepage eroding subsurface soils may allow for FTCA damages.  Tidal surges do not automatically flood developed land, and storm waters are not automatically flood waters.  The logical problem with this analysis is that once a tidal surge or storm water erodes a levee or scours in flood wall banks, the inland character of the lake or storm water appears to become flood waters.  I would rather just call the resulting occurrence property damages.[24]  Under Central Green’s analysis, there was still no planned purpose of the waters released here, just a failure.

Reservoir water was not released; levees eroded and flood walls broke.  Perhaps all fugitive lake water and storm water are not immunized “flood or flood waters” under this Act, and the FTCA’s liability applies.  This argument will require lower courts to rethink the 1928 Act and also consider the following issues:

1.  Was the Flood Control Act of 1928 for the Mississippi River and its tributaries intended to apply to another federal undertaking, i.e., the Lake Pontchartrain and Vicinity Hurricane Protection Project?

2.  If so, was the quid pro quo satisfied, e.g., adequate Mississippi River main line levees versus Lake Pontchartrain inadequate levees and flood walls, or promise of compensated flowage easements along the Mississippi River versus no compensation for inadvertently created flowage easements in areas near Lake Pontchartrain that flooded in 2005?[25]

3.  Is non-flood water, i.e., storm water, captured in drainage canals and which is not intended to be released inland, statutory “flood or flood waters” when it breaks through and causes damage?

4. How much flooding was due to, not flood, but navigation waters (the Mississippi River Gulf Outlet)?

5. Was the 1928 Act repealed by implication of the FTCA in 1945 or the Flood Control Act of 1965 (especially in light of Dolan, supra, note 22)?[26]


On the other side of the coin, we can turn  from torts to compensable takings under the Fifth Amendment to the United States Constitution.  Property damage and taking are close but are not the same, legally or morally.  A constitutional taking claim should not be barred by an Act of Congress, like the Flood Control Act of 1928.[27]  Temporary takings are compensable.[28]  Federal flooding is certainly eminent domain at the apex of abuse.

Inverse condemnation is a taking claim against the government to recover the value of property taken by the agency, though no formal exercise of eminent domain has been completed.[29]  Some takings are per se, such as those involving a physical invasion of land or those involving confiscatory regulations (e.g., zoning) that deprive owners of all economically viable use of their property.[30]  Otherwise, a balancing test is used to determine if public action is a taking, including the character of the government action, interference with reasonable investment back expectations, and the economic impact of government actions.[31]  The effect of the flooding here is on private property.  Flooding is a physical invasion of property and should constitute a per se taking.  The questions become was the flooding predictable and what is the value of property interest taken – e.g., easement or fee title.

Unintended invasion of private property from government activity, e.g., a flowage easement, may thus amount to a foreseeable appropriation or inverse condemnation.[32]  The public purpose of the taking would be the political illusion of flood control up to a point after the inevitable flooding event occurs.  Ridge Line, dealing with increased flooding from a federal project, held to constitute a compensable taking, 1) the government must intend to invade a protected interest or the asserted invasion is the direct, natural or probable result of an authorized activity, and 2) an invasion must either appropriate a benefit to the government at the expense of the property owner, or preempt the owner’s right to enjoy his property for an extended period of time (rather than inflict incidental or consequential injury that reduces its value).  The effect does not have to be infinite to be compensable.[33]  In other words, the flooding must be predictable, and the landowner must be deprived of beneficial uses of his or her land.

We have seen the Corps teams and National Science Foundation scientists dispute the obvious, e.g., whether it was predictable that a poorly designed or built flood wall could cause floods.[34]  This is a question of fact and expert opinion, but certainly predictability or foreseeability is arguable.  Merely because flooding was not anticipated by some Corps officials does not equate to flooding being unpredictable by other reasonable engineers.  On the magnitude issue, we had one Katrina flood in 2005, but the extent and duration of that flooding reclaimed parish lands that were once swamps many decades ago (“…the worst urban natural [and man-made] disaster in American history”[35]) continues.  Both issues of direct and indirect causation in the taking analysis are involved.[36]

If direct government action did cause physical invasions by flooding that is substantial, even if the frequency is not fully known yet, this is still arguably more than a mere stigma or tort in scope.  The delay in private land use is further exacerbated by uncertainty in funding, flood maps, insurance and contractors, etc., which all are in the ripple of the indirect taking in federally-caused flooding, as well as by possibly new contamination in many areas.[37]  More permanent taking would transcend a flowage easement and perhaps require payment for fee title, based on the property’s market value, including its highest and best use at the time of the taking.[38]  Value would be a jury question.

Of course, unlike personal injury and property damages in the FTCA, taking claims entitle property owners to just compensation only.  Suit would not be filed in the United States District Court, like the FTCA, but in the United States Court of Federal Claims if more than $10,000.00 is involved.[39]  Time limitation are more generous than the FTCA, six years[40], and only the tort actions against the United States are without benefit of a jury.[41]  The property owner must also retain title at the time of taking to recover.[42] 


Litigating with the federal government is no easy task.  Other litigation options exist at the state, private and local levels, and all are starting to play out.  Running the course here is like tiptoeing through the undulations of a Loch Ness monster – expensive to even locate.  Bail-outs and buy-outs may be the best option in the long run, even if they leave New Orleans and vicinity in a Davy Jones-like “death trap” should people stay, which many will.

Now a parting thought as I go beyond legal.  Perhaps our biggest shortcoming is, with our heads still bowed, we are gazing too low to find the bigger solutions.  We should also look up with the cloud physicists for answers and not focus solely on engineers working at ground level (and perhaps recreating the same errors from the past).[43]  Atmospheric modification has long been within the realm of science but is now ignored for political reasons.[44]  Feasible seeding techniques may slow the leading edge of tropical energy sails and weaken their vortex.  At least that option should not be ignored.  Levees, even good levees, should not be left to do their work alone any more.  Federal funds (and damages), flood walls, coastal restoration, barriers, shore stabilization, a sea change in flood protection and oversight and smart re-growth[45] are all part of the salvation needed.  Amen.

* S.J.D., special counsel at Jones, Walker, Associate Adjunct Professor of Law at Tulane Law School and lecturer at Loyola Law School.  This essay is an expanded version with footnotes of a presentation at the 11th Annual Tulane Environmental Law Conference in New Orleans on April 1, 2006.  The views herein are for discussion and are not admissions of any sort nor binding on any Jones, Walker clients. The author is not a litigant but is a flood victim.

[1]  Charles Beaumont, Elegy, 1960.

[2]  28 U.S.C. §1346(b).

[3]  Standard Form 95, attached.  28 U.S.C. §2401(b).  See Federal Tort Claims Handbook (JAGC 1999) at page 1, et seq., Requirements for Administrative Filing.

[4]  28 U.S.C. §§2675(a), 2401(b).  Suit must be filed within six months of final claim denial by an agency.

[5]  28 U.S.C. §2680(a) through (h).

[6]  28 U.S.C. §2680(a).

[7]  U.S. v. Gaubert, 499 U.S. 315, 111 S.Ct. 1267  (1991) (e.g., management of banking affairs).

[8]  Boyle v. United Technologies Corp., 487 U.S. 500, 108 S.Ct. 2510 (1988).

[9]  Barnson v. U.S., 816 F.2d 549 (10th Cir. 1987), cert. denied, 484 U.S. 896 (1987). 

[10]  Berkovitz by Berkovitz v. U.S., 486 U.S. 531, 108 S.Ct. 1954 (1988). 

[11]  Id.

[12]  Id.

[13]  33 U.S.C. 467, et seq.

[14]  See, Marshall, Levees Built Using Obsolete Standards, The Times-Picayune, May 1, 2006, Section 1, page A-1, and Walsh, Category 3 Pledges Ring Hollow, The Times-Picayune, May 3, 2006, Section 1, page A-1.  The Corps’ Interagency Performance Evaluation Task Force has recently admitted that the system used to determine flood protection need, risks and funding is flawed, and flooding of much of New Orleans would have occurred without the breaches.  Marshall, et al, Flood Policy Flawed, Report Says, The Times-Picayune, June 2, 2006, Section 1, pages A-1 and 8.  This mea culpa seems to be hinting at flaws in Congress more so than the Corps and begs the discretionary function exemption.  Was the flood protection Category 3 survivable or not?

[15]  Lt. Gen. Carl Strock, The Times-Picayune, April 6, 2006, Section 1, page A-1; and June 2, 2006, Section 1, page A-8.

[16]  Pub. L. 89-298, §204.

[17]  Lake Pontchartrain and Vicinity Hurricane Protection Project (GAO-05-1050T, September, 2005).

[18]  Overtopping, Says Corps; Breaches, Says LSU Analyst, The Times-Picayune, April 26, 2006, Section 1, page A-4.

[19]  Borenstein, Study:  N.O. Sinking Faster than Thought, The Times-Picayune, June 1, 2006, Section 1, page A-11.  See also Schwartz, Army Corps Admits Flaws in New Orleans Levees, The New York Times, June 1, 2006

[20]  33 U.S.C. §702(c).  This is but a small portion of a more detailed statute.

[21]  See, e.g., Barry, Rising Tide (Simon and Schuster, 1997)

[22]  See Hofman, An Enduring Anachronism:  Arguments for Repeal of the 702c Immunity Provision of the Flood Control Act of 1928, 79 Tex. L. Rev. 791 (2001), at Part VI.  Irrigation canals, navigation canals, lake reservoirs as well as rivers have been argued by the government to be covered by this immunity.  However, one court, noting the Act’s vague legislative history, has excluded navigation projects from the 1928 Act.  See Graci v. United States, 456 F.2d 20 (5th Cir. 1971), writ denied, 93 S.Ct. 2752 (1973).  This holding may still stand, notwithstanding Central Green Co., infra.  In light of the remedial nature of the FTCA and strict construction of sovereign immunity waiver in that FTCA context, flood water exceptions should be narrowly construed.  See e.g., Dolan v. United States Postal Service, 126 S.Ct. 1252 (2006).

[23]  121 St. Ct. 1005 (2001).

[24]  Interesting, by analogy, FEMA regulations at 44 C.F.R. 61, Appendix A(3), limit what is called flooding, e.g., overflow of land or tidal waters, unusual or rapid accumulation of runoff, or mudflow.  Broken floodwalls and subsurface seepage do not appear to be covered.

[25]  See, e.g., 33 U.S.C. §702(c), second paragraph. 

[26]  Implied repeals are not favored.  See Lockhart v. United States, 126 S.Ct. 699, 704 (2005) per concurring opinion of Justice Scalia.  Implied repeals exist when two statutes conflict or where a latter Act covers the whole subject of the earlier one and is clearly intended as a substitute.  Certainly, the Flood Control Act of 1965 covers Lake Pontchartrain while the 1928 Act does not.

[27]  See Hofman, supra, note 21 at Part V.B.  See also, Lewis v. Porters Creek Watershed Dist., 586 F.2d 1081, 1088 (6th Cir. 1978).

[28]  First English Evangelical Lutheran Church of Glendale v. County of Los Angeles, 482 U.S. 304, 318 (1987).  That case involved a regulatory taking.  Here we are talking about a direct physical taking which should be even more actionable.

[29]  King, The Fifth Amendment Takings Implications of Air Force Overflight and the Air Installation Compatibility Use Zone Program, 43 A.F. L. Rev. 197, n. 2 (1997).

[30]  Lucas v. South Coastal Council, 505 U.S. 1003 (1992).

[31]  Penn Central Transportation Co. v. New York, 438 U.S. 104 (1978).

[32]  Ridge Line, Inc. v. U.S., 346 F.3d 1346 (Fed. Cir. 2003).  Predictability, regardless of negligence, is the issue here.  See also, U.S. v. Causby, 328 U.S. 256, 66 S.Ct. 1062 (1946).  The fact that a plaintiff has alternate or distinct theories of recovery – negligence in tort and perhaps overlapping inverse condemnation – should not bar the latter.  See Warner/Electra/Atlantic Corporation v. City of DuPage, Illinois, 771 F.Supp. 911, 915-916 (N.D. Ill. 1991), affirmed, 991 F.2d 1280 (7th Cir. 1993).

[33]  Loretto v. Teleprompter Manhattan CATV Corp., 458 U.S. 419, 441, 102 S. Ct. 3164 (l982).

[34]  See Corps final draft reports at  See contra, Grisset, Team Pushes for Levee System Overhaul, The Times-Picayune, at A-1, May 23, 2006.  See also Van Heerden, The Storm:  What Went Wrong and Why During Hurricane Katrina - The inside Story from One Louisiana Scientist (Penguin USA 2006).

[35]  Russell, et al., It’s Nagin, The Times-Picayune, May 21, 2006, A-1.

[36]  See Applegate v. United States, 35 Fed. Cl. 406 (1996).

[37]  This raises the specter of federal clean-up liability under CERCLA.  42 U.S.C. §9601, et seq.  CERCLA waives federal sovereign immunity.  §9620(a)(1).  Liability under CERCLA is based on a release or threatened release of a hazardous substance from a facility which causes the occurrence of response costs (not private damages).  §9607.  Responsible parties include facility owners, facility operators, arrangers of waste disposal, and certain transporters.  §9607.  The levees and flood walls are arguably the “facility” which was owned or operated by the United States.  The United States could also have arranged the CERCLA release or threatened release, by participating in faulty levee inspection with local interests.  See, e.g., Gen. Corp. v. Olin Corp., 390 F.3d 433 (6th Cir. 2004).  Their failure resulted in the release of waters containing hazardous substances (benzene, heavy metals, etc.).  The waters also threatened and caused release of pollutants or contaminants (salt water, mold and bacteria, which are injurious or lethal to fresh water and land based life) and which are covered by CERCLA.  §9601(33).  Response costs could include decontamination, gutting, relocation, etc.  §9601(23).  Following the National Contingency Plan may be problematic.  §9607(a)(4)(A) and (B).  CERCLA coverage of clean-up costs for pollutants and contaminants is less clear than for hazardous substances.  See Jastram v. Phillips, 844 F.Supp. 1139 ( E.D. La. 1994).  Although EPA gave New Orleans a clean bill of health after Katrina, see Brown, Final EPA Report Deems N.O. Safe, The Times-Picayune, August 19, 2006, at A-1, EPA still notes widespread lead elevations around the city that were not caused by Katrina but, some argue, could have been exacerbated and released during the flooding.

[38]  Cooley v. United States, 46 Fed. Cl. 538, 550-553 (2000), affirmed in part, reversed in part, 324 F.3d 1297 (Fed. Cir. 2003).

[39]  28 U.S.C. §§1346(a)(2), 1491(a)(1).

[40]  28 U.S.C. 2401(a).

[41]  28 U.S.C. §1346(b), 2402.

[42]  See Applegate v. United States, 35 Fed. Cl. 406 (1996).

[43]  Poor levee rebuilding may have its own share of federal immunity.  42 U.S.C. 5148.  This immunity would not be from CERCLA.  United States v. City of New Orleans, 2003 U.S. Dist. Lexis 16765 ( U.S.D.C. E.D. La., 2003).

[44]  See, e.g.,

[45]  Houck, Can We Save New Orleans?, 19 Tul. E.L.R. 1 (2006).  Alas, no one has conceived of the “promised land” of a definitive smart map yet.  Go up, go back, or just go away, are our only guides now.