St. Bernard Parish Government v. United States

  United States Court of Appeals
      for the Federal Circuit
               ______________________

      ST. BERNARD PARISH GOVERNMENT,
 GWENDOLYN ADAMS, HENRY ADAMS, CYNTHIA
   BORDELON, STEVEN BORDELON, STEVE'S
 MOBILE HOME AND RV REPAIR, INC., EDWARD
 ROBIN, SR., EDWARD “PETE” ROBIN, JR., BRAD
ROBIN, ROBIN SEAFOOD COMPANY, INC., ROBIN
  YSCLOSKEY DEVELOPMENT #1, LLC, ROBIN
  YSCLOSKEY DEVELOPMENT #2, LLC, ROBIN
  YSCLOSKEY DEVELOPMENT #3, LLC, ROBIN
  YSCLOSKEY DEVELOPMENT #4, LLC, ROCCO
 TOMMASEO, TOMMOSO “TOMMY” TOMMASEO,
 ROCKY AND CARLO, INC., PORT SHIP SERVICE,
INC., AND OTHER OWNERS OF REAL PROPERTY
IN ST. BERNARD PARISH OR THE LOWER NINTH
     WARD OF THE CITY OF NEW ORLEANS,
             Plaintiffs-Cross-Appellants

                         v.

                 UNITED STATES,
                Defendant-Appellant
               ______________________

                2016-2301, 2016-2373
               ______________________

    Appeals from the United States Court of Federal
Claims in No. 1:05-cv-01119-SGB, Chief Judge Susan G.
Braden.
                ______________________
2         ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES




                 Decided: April 20, 2018
                 ______________________

    CHARLES J. COOPER, Cooper & Kirk, PLLC, Washing-
ton, DC, argued for plaintiffs-cross-appellants. Also
represented by BRIAN W. BARNES, VINCENT J.
COLATRIANO, MICHAEL W. KIRK, WILLIAM C. MARRA;
CARLOS A. ZELAYA, II, Mumphrey Law Firm, LLC, Chal-
mette, LA.

    BRIAN HALLIGAN FLETCHER, Office of the Solicitor
General, United States Department of Justice, Washing-
ton, DC, argued for defendant-appellant. Also represent-
ed by JOHN C. CRUDEN, BRIAN C. TOTH, Environment and
Natural Resources Division, United States Department of
Justice, Washington, DC.

    MARK F. HEARNE, II, Arent Fox, LLP, Clayton, MO,
for amici curiae National Federation of Independent
Business Small Business Legal Center, Reason Founda-
tion, Southeastern Legal Foundation, Property Rights
Foundation of America, Inc., National Association of
Reversionary Property Owners, James W. Ely, Jr. Also
represented by STEPHEN SHARP DAVIS, MEGHAN SUE
LARGENT.
                ______________________

      Before NEWMAN, LOURIE, and DYK, Circuit Judges.
DYK, Circuit Judge.
    Saint Bernard Parish Government and various other
owners of real property in St. Bernard Parish or in the
Lower Ninth Ward of the City of New Orleans (collective-
ly “plaintiffs”) brought suit in the Court of Federal Claims
(“Claims Court”) under the Tucker Act, 28 U.S.C.
§ 1491(a)(1), alleging a taking. They claimed that the
government was liable for flood damage to their proper-
ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES         3



ties caused by Hurricane Katrina and other hurricanes.
Plaintiffs’ theory was that the government incurred
liability because of government inaction, including the
failure to properly maintain or to modify the Mississippi
River-Gulf Outlet (“MRGO”) channel, and government
action (the construction and operation of the MRGO
channel). The Claims Court found a taking occurred and
awarded compensation. The government appeals, and
plaintiffs cross-appeal alleging that the Claims Court’s
compensation award was inadequate.
    We conclude that the government cannot be liable on
a takings theory for inaction and that the government
action in constructing and operating MRGO was not
shown to have been the cause of the flooding. This is so
because both the plaintiffs and the Claims Court failed to
apply the correct legal standard, which required that the
causation analysis account for government flood control
projects that reduced the risk of flooding. There was
accordingly a failure of proof on a key legal issue. We
reverse.
                      BACKGROUND
    New Orleans has a long history of flooding. The geo-
graphic location of the city makes it “particularly vulner-
able to hurricanes.” J.A. 25035. The city was hit by major
storms in 1909 and 1915, and much of the city flooded due
to the Fort Lauderdale Hurricane in 1947. In 1955, Con-
gress authorized the Army Corps of Engineers (“Corps”) to
study the need for additional hurricane protection in the
Lake Ponchartrain area. This resulted in a comprehensive
report known as the “Barrier Plan,” which recommended
a system of floodgates, levees, and floodwalls to protect
the area from hurricanes.
    In 1956, Congress authorized the Corps to construct
the MRGO navigation channel in New Orleans. The
purpose of the channel was to increase commerce by
providing a direct connection between the port of New
4         ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES



Orleans and the Gulf of Mexico. Construction was com-
pleted in 1968.
     Plaintiffs allege that over the course of the next sev-
eral decades, the construction, operation, and improper
maintenance of the MRGO channel caused various ad-
verse impacts that increased storm surge along the chan-
nel as follows. The construction, operation, and failure to
maintain MRGO increased salinity in the water by
providing a direct route for salt water to flow into the area
from the Gulf of Mexico. The saltwater changed the
character of the marshes and destroyed wetlands in the
area that previously acted as a natural buffer against
flooding. Moreover, the “failure of the Army Corps to
maintain the banks” caused erosion along the banks,
which allowed more water to pass through the channel at
higher velocities. MRGO also created the potential for a
funnel effect, which increased flooding during storms by
compressing storm surge into the channel and causing it
to rise faster and higher.
     In 1965, while MRGO was still under construction,
Congress authorized funding to implement the Barrier
Plan through the Lake Pontchartrain and Vicinity Hurri-
cane Protection Project (“LPV project”) to control flooding
resulting from hurricanes. See Flood Control Act of 1965,
Pub. L. No. 89-298, 79 Stat. 1073, 1077 (1965). At an
estimated cost of approximately $56 million ($447 million
in today’s money), the LPV project included construction
of levees and floodwalls in the St. Bernard basin along the
banks of MRGO utilizing dredged material from the
MRGO channel. The levee system was designed to, and
did, reduce the risk of flooding in New Orleans, including
specifically along the banks of MRGO. Construction
began around the same time that construction of MRGO
was concluding.
   Plaintiffs own properties located in the St. Bernard
Parish and Lower Ninth Ward areas. These properties
ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES          5



were catastrophically flooded during Hurricane Katrina
in 2005. Hurricane Katrina was “one of the most devas-
tating hurricanes that has ever hit the United States,
generating the largest storm surge elevations in the
history of the United States.” In re Katrina Canal Breach-
es Consol. Litig., 647 F. Supp. 2d 644, 678 (E.D. La. 2009),
aff’d in part, rev’d in part, 696 F.3d 436 (5th Cir. 2012).
Storm surge is a “wind generated process,” so “a storm of
such intensity creates an immense storm surge.” Id. at
679. During Hurricane Katrina, as a result of the storm
surge, levees from the LPV project around St. Bernard
Parish and the Lower Ninth Ward breached, contributing
to flooding in the area. Increased storm surge also con-
tributed to subsequent damage from other hurricanes.
    Plaintiffs brought an action in the Claims Court un-
der the Tucker Act, alleging that construction and opera-
tion of MRGO and failure to properly maintain or modify
it constituted a taking by causing flooding damage to
their properties. Plaintiffs made no effort to show that the
combination of MRGO and the LPV levees caused more
flooding than would have occurred without any govern-
ment action, arguing that the court should limit its con-
sideration to MRGO in isolation.
    After a bench trial in December 2011, the Claims
Court held that a temporary taking occurred. The Claims
Court found, as plaintiffs alleged, that a causal link
existed between increased storm surge and MRGO. The
construction of, continued operation of, and failure to
maintain or modify MRGO caused erosion, increased
salinity, wetlands loss, and a funnel effect, which in turn
caused increased storm surge. The Claims Court found
that “the substantially increased storm surge-induced
flooding of Plaintiffs’ properties that occurred during
Hurricane Katrina and subsequent hurricanes and severe
storms was the direct result of the Army Corps’ cumula-
6         ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES



tive actions, omissions, and policies regarding the MR-GO
that occurred over an extended period of time.” 1 St. Ber-
nard Par. Gov’t v. United States, 121 Fed. Cl. 687, 741
(2015); see also id. at 745 (“The evidence in this case
established that the substantial increase in storm surge
and flooding was the ‘direct, natural or probabl[e] result’
of the construction, expansions, operation, and failure to
maintain the MR-GO.”). The plaintiffs presented no
evidence, and the Claims Court made no findings, as to
whether the combination of these MRGO-related effects
and the LPV levees caused flooding on plaintiffs’ proper-
ties greater than would have occurred had the govern-
ment engaged in no action at all.
    The Claims Court also determined that these envi-
ronmental effects were foreseeable at least by 2004. The
Claims Court found that “it was foreseeable to the Army
Corps that the construction, expansions, operation, and
failure to maintain the MR-GO would increase salinity,
increase habitat/land loss, increase erosion, and increase
storm surge that could be exacerbated by a ‘funnel effect’
and likely cause flooding of Plaintiffs’ properties in a
hurricane or severe storm.” Id. at 723.
    After a separate trial on compensation in November
2013, the Claims Court awarded compensation of $5.46
million based primarily on the replacement cost of im-
provements to the properties and lost rental value during
the temporary taking period. The Claims Court also sua


    1   However, the Claims Court somewhat inconsist-
ently noted that some evidence suggested that MRGO
“did not significantly impact the height of Katrina’s storm
surge, not because the ‘funnel’ effect was nonexistent, but
because the storm was so great it nullified the impact of
either the wetlands or the intersection of MRGO and the
GIWW—the funnel—at the height of the surge.” J.A.
18361.
ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES         7



sponte awarded lost real-estate taxes to the New Orleans
city government, a non-party. The Claims Court then
certified a class under Court of Federal Claims Rule 23(a)
for purposes of liability and two subclasses for purposes of
just compensation.
   The government appeals both the finding of liability
and the compensation award. Plaintiffs cross-appeal the
amount of the compensation award. We have jurisdiction
pursuant to 28 U.S.C. § 1295(a)(3).
                        DISCUSSION
    Whether a taking under the Fifth Amendment has oc-
curred is a question of law with factual underpinnings.
Ridge Line, Inc. v. United States, 346 F.3d 1346, 1352
(Fed. Cir. 2003). We review the trial court’s legal deter-
minations de novo and its fact-findings for clear-error. Id.
                              I
     This is an inverse condemnation case. Inverse con-
demnation is the means by which a landowner may
recover just compensation under the Fifth Amendment for
a physical taking of his property when condemnation
proceedings have not been instituted. United States v.
Clarke, 445 U.S. 253, 257 (1980). The inverse condemna-
tion claim here is based on a taking of a flowage ease-
ment. It is well established that the government cannot
take such an easement without just compensation. United
States v. Dickinson, 331 U.S. 745, 748 (1947); Ridge Line,
346 F.3d at 1352. Most recently, in Arkansas Game &
Fish Commission v. United States, 568 U.S. 23, 34 (2012),
the Supreme Court held that temporary, government-
induced flooding may give rise to a claim for the taking of
a flowage easement. Thus, the issue presented is whether
the increased flooding from MRGO constituted a tempo-
rary taking. Proof of such a claim requires the plaintiffs to
establish that government action caused the injury to
their properties—that the invasion was the “direct, natu-
8         ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES



ral, or probable result of an authorized activity.” Ridge
Line, 346 F.3d at 1355; see also Arkansas Game, 568 U.S.
at 38–40; Sanguinetti v. United States, 264 U.S. 146, 149
(1924); Moden v. United States, 404 F.3d 1335, 1342–43
(Fed. Cir. 2005). Establishing liability for a temporary
taking also requires proof that the invasion was either
intentional or foreseeable. Arkansas Game, 568 U.S. at
39; Moden, 404 F.3d at 1343.
                             II
     The Claims Court’s finding of liability here is based in
large part on the failure of the government to take action,
particularly on its failure to maintain MRGO or to modify
it. The Claims Court determined that the government’s
decisions not to armor the banks and not to repair erosion
along the banks caused the channel to widen, which
allowed MRGO to “carry significantly more water at
higher velocities.” St. Bernard Par. Gov’t, 121 Fed. Cl. at
730. The Claims Court also found that “the failure of the
Army Corps to maintain the banks” caused expansion
that “affected the severity of wave attack on the Chal-
mette levee.” Id. at 731. Indeed, each of the Claims
Court’s causation findings mentions a causal link to the
“failure to maintain” MRGO. Id. at 726, 729, 731, 733,
738. In finding a causal link between MRGO and storm
surge, the Claims Court cited expert testimony noting
that there was “no evidence that the MRGO project was
ever modified to reduce the predictable excess surge
stresses and wave attack.” Id. at 737. In particular, the
Claims Court noted that MRGO’s lack of armoring or
foreshore protection contributed to erosion along the
banks. 2 Id. at 700, 729. Thus, the government’s failure to


    2  The Corps added foreshore protection in the
1980s, but the Claims Court noted that the decision to
defer erosion protection allowed the channel to widen
considerably. Id. at 692, 729.
ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES          9



properly maintain or to modify the banks played a signifi-
cant role in plaintiffs’ takings theory and the Claims
Court’s analysis.
    While the theory that the government failed to main-
tain or modify a government-constructed project may
state a tort claim, it does not state a takings claim. A
property loss compensable as a taking only results when
the asserted invasion is the direct, natural, or probable
result of authorized government action. Sanguinetti, 264
U.S. at 149–50; Ridge Line, 346 F.3d at 1355.
    On a takings theory, the government cannot be liable
for failure to act, but only for affirmative acts by the
government. “The government’s liability for a taking does
not turn, as it would in tort, on its level of care.” Moden,
404 F.3d at 1345. Instead, takings liability arises from an
“authorized activity.” Id. at 1338 (citing Ridge Line, 346
F.3d at 1355); see also Lingle v. Chevron USA, Inc., 544
U.S. 528, 543 (2005) (explaining that “the Takings Clause
presupposes that the government has acted in pursuit of a
valid public purpose”); Cary v. United States, 552 F.3d
1373, 1377, 1379 (Fed. Cir. 2009); Acadia Tech., Inc. v.
United States, 458 F.3d 1327, 1331 (Fed. Cir. 2006); All. of
Descendants of Tex. Land Grants v. United States, 37 F.3d
1478, 1481 (Fed. Cir. 1994) (“A claimant under the Fifth
Amendment must show that the United States, by some
specific action, took a private property interest for public
use without just compensation.”). 3 In both physical tak-
ings and regulatory takings, government liability has


   3    See also 1 George Cameron Coggins & Robert L.
Glicksman, Public Natural Resources Law § 12:14 (“Tak-
ings result from authorized acts by government officials,
whereas ‘[c]hallenges to the propriety or lawfulness of
government actions sound in tort.’”) (quoting Thune v.
United States, 41 Fed. Cl. 49, 52 (1998)); 3-8 Nichols on
Eminent Domain § 8.01.
10        ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES



uniformly been based on affirmative acts by the govern-
ment or its agent. 4
    In the flooding context, in particular, both Supreme
Court precedent and our own precedent have uniformly
based potential takings claims on affirmative government
acts. For example, in Arkansas Game, the Supreme Court
found a temporary taking claim could be based on affirm-
ative actions by the government in releasing water from a
government constructed and operated dam that caused
downstream flooding on the plaintiff’s property. 568 U.S.
at 27–28. 5 In Ridge Line, the United States Postal Service



     4  See, e.g., Armstrong v. United States, 364 U.S. 40,
48 (1960) (taking occurred when government required
transfer of title of an unfinished boat, making a lien
unenforceable); United States v. Pewee Coal Co., 341 U.S.
114, 116–17 (1951) (taking occurred when government
temporarily seized and operated coal mines during war);
United States v. Causby, 328 U.S. 256, 267 (1946) (taking
occurred when military flew planes over a chicken farm at
low altitudes); see also Dolan v. City of Tigard, 512 U.S.
374, 396 (1994) (taking occurred when city granted build-
ing permit on condition that owner dedicate public green-
way along stream and allow public access); Lucas v. S.C.
Coastal Council, 505 U.S. 1003, 1015–19 (1992) (taking
occurred when government regulation completely elimi-
nated economic use of a property); Nollan v. Cal. Coastal
Comm’n, 483 U.S. 825, 841–42 (1987) (taking occurred
when state conditioned grant of a building permit on
recording an easement to allow public access to the
beach); Loretto v. Teleprompter Manhattan CATV Corp.,
458 U.S. 419, 441 (1982) (taking occurred when state
statute required landlords to allow installation of cable
TV equipment on the premises).
    5   See also United States v. Kan. City Life Ins. Co.,
339 U.S. 799, 801, 811–12 (1950) (taking occurred when
ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES         11



built a facility that increased storm water runoff onto the
plaintiff’s property. 346 F.3d at 1351. We held that this
government action may form the basis for an inverse
condemnation claim. Id. at 1355.
    In contrast, other cases establish that takings liability
does not arise from government inaction or failure to act.
In United States v. Sponenbarger, 308 U.S. 256 (1939),
the government built a flood protection system. The
plaintiff complained that the system was inadequate to
protect the plaintiff’s property from flooding. In finding
that there was no taking, the Supreme Court held that
“[w]hen undertaking to safeguard a large area from
existing flood hazards, the Government does not owe
compensation under the Fifth Amendment to every land-
owner which it fails to or cannot protect.” Id. at 265.
Similarly, in Georgia Power Co. v. United States, 633 F.2d
554 (Ct. Cl. 1980), cert. denied, 450 U.S. 981 (1981),
Georgia Power brought a takings claim asserting the
government took its power line easement by permitting
sailboat masts to intrude into the portion of the easement
over a government reservoir. Finding no takings liability
for failure to regulate sailboat mast heights, the Court of
Claims explained that “issuance of such regulations is
merely a discretionary act, and a taking may not result
from this discretionary inaction” absent a duty to act. Id.
at 527.




construction of federal lock and dam raised water level in
a river and caused flooding); Dickinson, 331 U.S. at 747,
751 (taking occurred when the government built a dam
that increased water levels, which flooded plaintiffs’
property); United States v. Cress 243 U.S. 316, 330 (1917)
(same); United States v. Welch, 217 U.S. 333, 338–39
(1910) (same); United States v. Lynah, 188 U.S. 445, 469,
474 (1903) (same).
12        ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES



    Plaintiffs point to no case where the government in-
curred takings liability based on inaction. Takings liabil-
ity must be premised on affirmative government acts. The
failure of the government to properly maintain the MRGO
channel or to modify the channel cannot be the basis of
takings liability. Plaintiffs’ sole remedy for these inac-
tions, if any, lies in tort. 6
                            III
    Here, the sole affirmative acts involved were the con-
struction of MRGO, which was completed by 1968, and
the continued operation of the channel. 7 The parties
debate whether the injury to the plaintiffs was foreseea-
ble as a result of these actions. We need not reach that
question because we conclude that the plaintiffs have


     6  Here, another group of plaintiffs, who owned land
in the St. Bernard polder, originally sued in tort but lost.
Those plaintiffs brought a lawsuit against the United
States under the Federal Tort Claims Act (“FTCA”), 28
U.S.C. § 2671. The plaintiffs initially succeeded on the
tort theory at the U.S. District Court for the Eastern
District of Louisiana. See In re Katrina Canal Breaches
Consol. Litig., 647 F. Supp. 2d 644 (E.D. La. 2009). The
Fifth Circuit then reversed that holding on the ground
that the government was immune from liability under the
discretionary-function exception of the FTCA. See In re
Katrina Canal Breaches Litig., 696 F.3d 436 (5th Cir.
2012).
    7   The extent to which plaintiffs contend that the
operation of MRGO caused their injury is unclear. At
least by 2009, it appears that plaintiffs concede that
MRGO’s operation was causing them no injury because
they alleged that the closure of the channel in that year
“made at most a negligible contribution to protecting
Plaintiffs’ properties from the risk of recurring future
flooding.” Appellee Br. 78.
ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES          13



failed to establish that the construction or operation of
MRGO caused their injury.
     It is well established that a takings plaintiff bears the
burden of proof to establish that the government action
caused the injury. Causation requires a showing of “what
would have occurred” if the government had not acted.
United States v. Archer, 241 U.S. 119, 132 (1916). In order
to establish causation, a plaintiff must show that in the
ordinary course of events, absent government action,
plaintiffs would not have suffered the injury. The burden
of proof is on the plaintiff to establish causation. Thus, for
example, in Archer, plaintiffs claimed that the govern-
ment’s construction of a dike on their property constituted
a taking because the construction of the dike caused
depositing of sand and gravel on their land. 241 U.S. at
128. Due to the possibility that without the dike, a river
may have flowed through plaintiff’s property and perma-
nently submerged the property, the Supreme Court
remanded the case to determine whether the testimony
demonstrated “what would have occurred if the dike had
not been constructed.” Id. at 132. In Sanguinetti, plain-
tiffs brought a takings claim alleging that a canal built by
the government caused flooding damage. 264 U.S. at 147.
The Court noted the relevance of whether “[t]he land
would have been flooded if the canal had not been con-
structed.” Id.; see Danforth v. United States, 308 U.S. 271,
286 (1939) (“The Government could become liable for a
taking . . . by such construction as would put upon this
land a burden, actually experienced, of caring for floods
greater than it bore prior to the construction.”); see also
Arkansas Game, 568 U.S. at 34.
    Our cases are to the same effect. In Accardi v. United
States, the government built a dam, and after a severe
storm with unexpected precipitation, water flowed onto
plaintiffs’ property. 599 F.2d 423 (Ct. Cl. 1979). The court
explained that “plaintiffs have wholly failed to show that
defendant’s construction or operation of the [dam] sub-
14        ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES



jected their lands to any additional flooding above what
would have occurred in consequence of the severe . . .
storm had defendant not constructed the [dam] at all.” Id.
at 429–30. The court then held that “[i]n these circum-
stances, there has been no taking of plaintiffs’ property.”
Id. at 430. 8 Thus, the causation analysis requires the
plaintiff to establish what damage would have occurred
without government action.
    Here, the plaintiffs failed to present evidence compar-
ing the flood damage that actually occurred to the flood
damage that would have occurred if there had been no
government action at all. 9 The plaintiffs’ proof of causa-



     8   See also Bartz v. United States, 633 F.2d 571, 577
(Ct. Cl. 1980) (finding no taking when “excessive precipi-
tation was the root cause of the flooding experienced by
plaintiffs,” and in most instances, flooding “would in all
likelihood have happened without the existence of the
upstream dam”); ARK-MO Farms, Inc. v. United States,
530 F.2d 1384, 1386 (Ct. Cl. 1976) (“No proof was made
that [the dam] or any other consequence of the [govern-
ment] project was the cause of the floods complained of.
This failure of proof would alone dispose of plaintiff’s
case.”); Coates v. United States, 110 F. Supp. 471, 475 (Ct.
Cl. 1953) (no taking where “plaintiffs have not been able
to establish that the land would not have been flooded
had there been no dikes”).
    9    We note that, though it was excluded during the
just-compensation portion of the trial, the government
presented evidence that “[i]f the Mississippi River Gulf
Outlet (MRGO) had never been built, or if the MRGO had
remained at its original design dimensions, the flooding of
the Trial Properties east of Paris Road would have been
virtually identical to the flooding that actually occurred
on those properties during Hurricane Katrina. For the
Trial Properties located west of Paris Road, the maximum
ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES          15



tion rested entirely on the premise that it was sufficient
to establish that the plaintiffs’ injury would not have
occurred absent the construction and operation of the
MRGO channel without taking account of the impact of
the LPV flood control project. Plaintiffs on appeal are
clear that in their view the LPV levees cannot be consid-
ered in the causation analysis. Plaintiffs argue that the
Claims Court “properly analyzed whether Plaintiffs’
properties would have flooded absent MRGO, not whether
they would have flooded absent both MRGO and the LPV
levee system.” Appellee Br. 17.
    The result is that plaintiffs failed to take account of
other government actions—specifically the LPV project
including the construction of a vast system of levees to
protect against hurricane damage—that mitigated the
impact of MRGO and may well have placed the plaintiffs
in a better position than if the government had taken no
action at all. 10 In other words, the plaintiffs addressed the
wrong question—asking not whether the whole of the
government action caused the plaintiffs’ injury, but rather



water elevations would have been 1–3 feet lower.” St.
Bernard Par. Gov’t, 126 Fed. Cl. at 717; see also J.A.
15812–13, 16213–17.
    10  The LPV project apparently prevented the flood-
ing of St. Bernard from other hurricanes before Katrina.
St. Bernard polder did not flood during Hurricane Camille
in 1969, although other areas in New Orleans flooded
during that storm. St. Bernard Par. Gov’t, 121 Fed. Cl. at
704. Moreover, after Hurricane Katrina, minimal flood-
ing occurred inside the levees during Hurricane Gustav in
2008, although that hurricane caused flooding outside the
levees. Id. at 714. Of seventy-seven properties located
within the federal levee system, sixty-two only flooded
once (during Hurricane Katrina) after the LPV project
was authorized. J.A. 10857–58.
16         ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES



whether isolated government actions, the construction
and operation of MRGO, caused their injury. And the
Claims Court’s causation findings took no account of the
risk-decreasing impact of the LPV levee construction. St.
Bernard Par. Gov’t, 121 Fed. Cl. at 724–38.
    The plaintiffs’ approach to causation is simply incon-
sistent with governing Supreme Court and Federal Cir-
cuit authority, particularly in flooding cases. These cases
establish that the causation analysis must consider the
impact of the entirety of government actions that address
the relevant risk. In Sponenbarger, the plaintiff owned
land that was in a contemplated floodway of a govern-
ment flood control plan, and the plaintiff alleged that the
government plan caused flooding on it. 308 U.S. at 260.
The plaintiff’s land was a natural floodway that had
flooded in the past, and the land would have flooded but
for other government work done under the flood control
plan. Id. at 263–64. The Court stated “[t]he Government
has not subjected respondent’s land to any additional
flooding, above what would occur if the Government had
not acted.” Id. at 266. The Court explained that to hold
the government liable for “damages which may result
from conjectural major floods, even though the same
floods and the same damages would occur had the Gov-
ernment undertaken no work of any kind” would “far
exceed even the ‘extremest’ conception of a ‘taking’ by
flooding within the meaning of the Fifth Amendment.” Id.
at 265. It further made clear that “[t]he far reaching
benefits which respondent’s land enjoys from the Gov-
ernment’s entire program” had to be considered. Id. at
266–67. Thus,
     [e]nforcement of a broad flood control program
     does not involve a taking merely because it will
     result in an increase in the volume or velocity of
     otherwise inevitably destructive floods, where the
     program measured in its entirety greatly reduces
     the general flood hazards, and actually is highly
ST. BERNARD PARISH GOVERNMENT        v. UNITED STATES             17



      beneficial to a particular tract of land. . . . [I]f gov-
      ernmental activities inflict slight damage upon
      land in one respect and actually confer great bene-
      fits when measured in the whole, to compensate
      the landowner further would be to grant him a
      special bounty. Such activities in substance take
      nothing from the landowner.
Id.
    Similarly, in John B. Hardwicke Co. v. United States,
467 F.2d 488 (Ct. Cl. 1972), the government built two
dams as part of a flood control project. The first dam
decreased the risk of flooding on the plaintiff’s property,
while the second dam increased the risk of flooding. The
court explained that overall “the expectation of flooding
was still far less than it would have been if there had
been no flood control program at all” and concluded the
government was not liable for a taking. Id. at 490–91; see
also ARK-MO, 530 F.2d at 1386 (finding no taking when a
government flood control program was “at most ‘little
injury in comparison with far greater benefits conferred’”
(quoting Sponenbarger, 308 U.S. at 267)).
    More recently, in the remand decision of Arkansas
Game, we clarified that the appropriate analysis for
causation considers all government actions. Arkansas
Game & Fish Comm’n v. United States, 736 F.3d 1364,
1372 n.2 (Fed. Cir. 2013). As discussed above, in that
case, the government built a dam and then released water
(a deviation from policy) that flooded the plaintiff’s prop-
erty. We explained that “the proper comparison would be
between the flooding that occurred prior to the construc-
tion of [the dam] and the flooding that occurred during
the deviation period,” emphasizing that the causation
analysis considers causation based on the entirety of
18        ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES



government action, not merely the deviation from the
original water-release policy. Id. 11
    This principle has been applied outside the context of
takings induced by flooding. In Cary, a hunter set a fire
when visiting a national forest, and the fire spread onto
plaintiffs’ properties. 552 F.3d at 1375. Plaintiffs alleged
that certain government fire-suppression policies and
visitor policies caused the fire and constituted a taking
because the policies increased the risk of fires. In explain-
ing that a takings plaintiff must show that the injury is
the “direct, natural, and probable result” of government
action, we explained that “government action” includes all
of the government’s actions. Id. at 1377 n.*.
    There, the government action included “a long se-
quence of decisions, some risk-increasing but others risk-
decreasing, spread out over decades.” Id. at 1379. We
rejected plaintiffs’ attempt to “cherry-pick parts of the
[government] policy which they argue[d] ha[d] increased
the risk of wildfire since 1911 without acknowledging that
much of the [government] policy over the last century
ha[d] been devoted to reducing the risk of wildfire.” Id. at
1377 n.*. We explained that “each risk-decreasing action
in the Forest Service’s policies is an intervening act
breaking whatever causal chain would lead from an
accused risk-increasing action to the conflagration which
destroyed the landowners’ property.” Id. at 1379. Thus,
the causation analysis must consider both risk-increasing
and risk-decreasing government actions over a period of


     11 In Arkansas Game, the original water-release pol-
icy (before the deviation) mimicked the pre-dam water
flows. Therefore, comparing the flooding that occurred
with the deviation to the flooding that would have oc-
curred under the original water-release policy, rather
than to what would have occurred before the dam was
built, had no impact on the outcome. 736 F.3d at 1372 n.2.
ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES         19



time to determine whether the totality of the govern-
ment’s actions caused the injury.
    Plaintiffs argue that even if in some circumstances
the totality of government action must be considered in
determining causation, such consideration is unnecessary
if the beneficial government action is unrelated to the
detrimental government action. Here, they contend that
the relevant beneficial government action must be part of
the same project and that “[t]he Government cannot
defeat Plaintiffs’ claim by pointing to benefits provided by
the separate LPV project.” Appellee Br. 54. That is not
correct.
    To be sure, in determining causation, government ac-
tions must be directed to the same risk that is alleged to
have caused the injury to the plaintiffs. Here, there is no
question that the LPV project was directed to decreasing
the very flood risk that the plaintiffs allege was increased
by the MRGO project. The LPV project was directly
concerned with flood control; it was authorized under the
Flood Control Act of 1965. See Pub. L. No. 89-298, 79 Stat.
1073, 1077 (1965). The LPV project included levees along
the banks of MRGO, and the construction of the levees
used some of the material dredged from MRGO. The
levees built under the LPV project decreased the risk of
flooding in the area, including on plaintiffs’ properties.
Avoiding flooding damage was the very objective of the
system of levees.
    The relatedness of the MRGO and LPV projects is re-
inforced by the theory that the taking occurred because
MRGO caused breaches in the levees. The plaintiffs’ claim
rests on the assertion that the MRGO project undermined
the LPV government flood-control project. Plaintiffs argue
that “[a]bsent MRGO, the levees would not have
breached, or at a minimum, would have breached later;
notably, other levee segments that were not located near
MRGO, and thus not exposed to destructive MRGO
20        ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES



waves, did not suffer extensive breaching.” Appellee Br.
11, see also id. at 51–53, 62. 12 In the causation analysis,
the Claims Court noted that a problem with MRGO was
that it rendered flood-control projects at least partially
ineffective, quoting expert testimony that MRGO exposed
the LPV levees to “greater stress . . . for a longer period”
during Hurricane Katrina and that “all of the LPV struc-
tures that breached were adjacent to some part of the
MRGO project.” St. Bernard Par. Gov’t, 121 Fed. Cl. at
737.
    When the government takes actions that are directly
related to preventing the same type of injury on the same
property where the damage occurred, such action must be
taken into account even if the two actions were not the
result of the same project.
    In arguing that the other government actions only
need to be considered if they are part of the same project,
plaintiffs rely on authorities not directed to causation, but
rather concerned with the extent of the economic injury
sustained by the plaintiffs or the amount of a just com-
pensation award. In assessing economic loss for regulato-
ry takings, the entirety of government action must be
considered. Penn Cent. Transp. Co. v. City of New York,
438 U.S. 104, 113–14 (1978). 13 And in determining just


     12 See also J.A. 10344 (“[MRGO] led directly to both
greater and earlier breaching of the levees than would
have otherwise been the case, which in turn had grave
implications for the flooding of the developed areas within
the St. Bernard polder.”); J.A. 10355.
    13  See A&D Auto Sales, Inc. v. United States, 748
F.3d 1142, 1157–59 (Fed. Cir. 2014) (holding that to prove
economic loss based on a government requirement that
automobile manufacturers terminate their dealer fran-
chises as a condition of receiving federal assistance during
bankruptcy, the dealership owners must show that the
ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES          21



compensation, offsetting benefits may be considered. See,
e.g., United States v. Miller, 317 U.S. 369, 375–77 (1942).
But there, benefits generally must flow from the same
government action. Plaintiffs primarily rely on City of
Van Buren v. United States, 697 F.2d 1058, 1061–62 (Fed.
Cir. 1983), for the proposition that offsetting benefits may
be considered only where land is enhanced in value by the
same project for which it is condemned. However, the
discussion of offsetting benefits in Van Buren did not
involve an issue of causation. In that case, the govern-
ment closed a dam, which raised groundwater levels and
caused the plaintiff’s sewer system to flood. Id. at 1059.
The court determined that “closure of the dam in fact
caused [plaintiff’s] having to replace [pipe] lines.” Id. at
1060. The other benefits from closing the dam, such as
bank stabilization, flood control, and lowered costs for
river transportation, did nothing to mitigate the risk of
groundwater levels rising, which caused the flooding. Id.
at 1061. And on the issue of economic injury, the govern-
ment had not shown offsetting benefits specific to the
plaintiff’s property rather than the public at large that
would render the injury to the plaintiff’s property de


franchises would have retained some value without the
government’s financial assistance to the automobile
manufacturers); Cienega Gardens v. United States, 503
F.3d 1266, 1282–83 (Fed. Cir. 2007) (explaining that the
Claims Court erred by “fail[ing] to consider the offsetting
benefits that the statutory scheme afforded which were
specifically designed to ameliorate the impact of the
prepayment restrictions” in a statute); see also Connolly v.
Pension Benefit Guar. Corp., 475 U.S. 211, 225–26 (1986)
(noting that, in determining the economic impact of a
statute alleged to constitute a taking, it is relevant that “a
significant number of provisions in the Act . . . moderate
and mitigate the economic impact of an individual em-
ployer’s liability.”).
22        ST. BERNARD PARISH GOVERNMENT     v. UNITED STATES



minimis compared to the benefits conferred on the proper-
ty. Id. at 1062.
    We are aware of no case, and the plaintiffs have cited
none, where the government has taken action that creates
a risk of flooding and subsequent government action
designed to mitigate that risk can be ignored in the causa-
tion analysis. 14 That is what the plaintiffs have done here.
When government action mitigates the type of adverse
impact that is alleged to be a taking, it must be consid-
ered in the causation analysis, regardless of whether it




     14  John B. Hardwicke suggested that if the risk-
reducing government action preceded the risk-increasing
action, the risk-reducing action would only be considered
in assessing causation if the risk-increasing action was
“contemplated” at the time of the risk-reducing action.
467 F.3d at 490–91. In this respect, the court relied on the
Miller doctrine, which holds that when the project for
which property is condemned also confers benefits, the
plaintiff cannot recover just compensation for those
benefits. Miller, 317 U.S. at 375–77. The Miller rule only
applies to benefits that arise from the same government
project for which the property is condemned. Id. Whether
the John Hardwicke approach is correct or whether the
Miller doctrine is even relevant to determining causation
is not raised in this case. The John Hardwicke approach
would only be relevant if the LPV project had been con-
structed before MRGO. See Van Buren, 697 F.2d at 1061
(“[A]pplication of the ‘Miller doctrine’ necessarily presup-
poses that the property owner seeks to recover the incre-
ment in property value effected by a federal project prior
to condemnation.”) Here, the risk-increasing action
(MRGO) was constructed before the risk-reducing action
(LPV project), and in any event, MRGO was certainly
contemplated when the LPV levees were built.
ST. BERNARD PARISH GOVERNMENT   v. UNITED STATES        23



was formally related to the government project that
contributed to the harm. 15
     Indeed, the plaintiffs themselves admit that other un-
related projects have to be considered in the causation
analysis. In the compensation decision, the Claims Court
determined that the temporary taking ended on the date
that MRGO closed. However, plaintiffs argued, and
continue to argue on appeal, that the taking did not end
when MRGO closed, but instead when the government
built a new “risk reduction” levee system. Plaintiffs argue
that the flooding risk “ended (at the earliest) on June 1,
2011, when the Corps substantially completed the new
HSDRRS ‘risk reduction’ levee system with its new robust
levees and floodwalls and massive multi-billion dollar
surge barrier.” Appellee Br. 78. This risk-reduction levee
system is a new, separate project. Strikingly, the plain-
tiffs’ own characterization of the temporary taking
demonstrates that the totality of government action is
relevant to the takings inquiry, regardless of whether
individual construction projects were authorized under
separate congressional legislation.
    Under the correct legal standard, plaintiffs failed to
establish that government action, including both the
construction of MRGO and the levees, caused their injury.
By their own admission, they have failed to consider the




   15    Plaintiffs also argue the burden was on the gov-
ernment to establish the “offsetting benefit” from the
levees, relying on our decision in CCA Associates v. United
States, 667 F.3d 1239 (Fed. Cir. 2011). That decision did
not change the well-established rule that proof of causa-
tion lies with the plaintiffs, but rather addressed whether
the plaintiffs must demonstrate the absence of mitigating
factors in order to prove economic injury. Id. at 1245.
24        ST. BERNARD PARISH GOVERNMENT    v. UNITED STATES



impact of the risk-reducing LPV project. 16 Thus, there
was a failure of proof on the key issue of causation. Be-
cause plaintiffs failed to show that government action,
including both MRGO and the LPV project, caused their
injury, the government is not liable for a taking under the
Fifth Amendment based on the construction or operation
of MRGO.
                       CONCLUSION
    In summary, we conclude that the allegations of gov-
ernment inaction do not state a takings claim, and that
plaintiffs have not established that the construction or
operation of MRGO caused their injury. In light of our
disposition, we do not reach the other issues.
                       REVERSED




     16  It appears that a few of plaintiffs’ properties are
outside the federal levee system. But even as to those, the
plaintiffs failed to show their properties would not have
flooded absent MRGO or absent the combination of
MRGO and the federal flood-control program. These
properties routinely flooded, even before Hurricane Katri-
na and before the construction of MRGO was completed.
J.A. 10389 (explaining that these properties “flooded
during each of the five hurricanes that have struck the
area since Betsy on September 10, 1965”).