Paterno v. State

6 Cal. Rptr. 3d 854 (2004) 113 Cal. App. 4th 998

Peter PATERNO et al., Plaintiffs and Appellants,
v.
STATE of California et al., Defendants and Respondents.

No. C040553

Court of Appeal, Third District

November 26, 2003 As Modified on Denial of Rehearing December 24, 2003 Review Denied March 17, 2004.

Desmond, Nolan, Livaich & Cunningham, Gary Livaich, David Collins, and Richard F. Desmond, Sacramento, and Law Office of Clifford E. Hirsch; Howard, Rice, Nemerovski, Canady, Falk & Rabkin, Jerome B. Falk, Jr., and Simon J. Frankel, San Francisco, for First Union Real Estate Equity & Mortgage Investments; Kronick, Moskovitz, Tiedemann & Girard and Lloyd Hinkelman; Law Offices Of Stanley Bell, Sally G. Bergman; Robins, Kaplan, Miller & Ciresi, and Scott G. Johnson, Los Angeles; and Frederick A. Jacobsen for Plaintiffs and Appellants.

Bill Lockyer, Attorney General, Andrea Hoch, Chief Assistant Attorney General, Darryl Doke, Supervising Deputy Attorney General, Sterling A. Smith, Deputy Attorney General, for State of California; G. Steven Jones, La Jolla, and Carl R. Lindmark, Yuba City, for Reclamation District 784, Defendants and Respondents.

MORRISON, J.

The environmental aftermath of the Gold Rush continues to plague California. Hydraulic mining debris caused flooding which led to the building of levees at the confluence of the Yuba and Feather Rivers. Almost a century ago the Linda levee was built with uncompacted mining debris, and the use of that debris caused the levee to collapse on February 20, 1986.

About 3,000 plaintiffs sued the State of California (State), Reclamation District 784 (District) and others not now parties, seeking damages. In Paterno v. State of California (1999) 74 Cal. App. 4th 68, 87 Cal. Rptr. 2d 754 (Paterno I), we affirmed a defense jury verdict finding no dangerous condition of public property and reversed an inverse condemnation liability finding against defendants, and remanded for another trial on inverse liability. A new coordination judge (Hon. John J. Golden), conducted a lengthy court trial and issued a defense judgment against sample plaintiffs (collectively, Paterno) who filed this appeal.

Paterno embraces Judge Golden's factual findings, which in his view, create inverse liability on the part of the State as a matter of law. We agree. When a public entity operates a flood control system built by someone else, it accepts liability as if it had planned and built the system itself. A public entity cannot be held liable for failing to upgrade a flood control system to provide additional protection. But the trial court found the levee was built with porous, uncompacted mining debris, in a location which encouraged seepage, leading directly to the failure of the levee, and that long before the failure, feasible cures could have fixed the problems. Use of such technology would not have been an upgrade, but would have ensured the planned flood control capacity was achieved.

Inverse liability stems from the California Constitution and is not dependent on tort or private property principles of fault. (See Albers v. County of Los Angeles (1965) 62 Cal. 2d 250, 261-262, 42 Cal. Rptr. 89, 398 P.2d 129 (Albers ).) California Supreme Court precedent dictates that a landowner should not bear a disproportionate share of the harm directly caused by failure of a flood control project due to an unreasonable plan. Whether the plan is unreasonable is not measured by negligence principles, as in a tort case alleging a dangerous condition of public property, but by balancing a number of specific factors referred to as the Locklin factors. (Locklin v. City of Lafayette (1994) 7 Cal. 4th 327, 27 Cal. Rptr. 2d 613, 867 P.2d 724 (Locklin).) Based on the facts found by the trial court and application of the Locklin factors, we conclude Paterno's damages were directly caused by an unreasonable State plan which resulted in the failure of the Linda levee and the State is liable to pay for Paterno's damages. In large part our conclusion is based on the fact that the levee system benefited all of California and saved billions of dollars and to require Paterno to bear the cost of the partial failure of that system — a failure caused by construction and operation of an unstable levee — would violate Locklin. A basic part of the State's flood plan was to accept existing levees as much as possible, to reduce the cost of an extensive, coordinated, flood-control system. The People benefited from that cost-saving feature. However, the record shows that State never tested the Linda levee, or reviewed the records of its construction, to see if it was as strong as the global plans assumed it was, and the State even ignored specific warnings about the levee's weaknesses. In such circumstance, the costs of the levee failure must be deemed part of the deferred costs of the project. We do not separately address an alternate theory that the State is liable because of an inadequate levee inspection plan, although we discuss the lack of any plan to examine the heart of the levee.

Although in some ways the District is a coparticipant with the State in operating the levee, we conclude it is entitled to judgment. The District was responsible for and only for ordinary maintenance and could not alter the structure of the levee, even if it had the financial means to do so.

We will affirm the judgment in favor of the District, reverse the judgment in favor of the State with directions to enter judgment in favor of Paterno, and remand for further proceedings. In making this order, we realize this case is as hoary as Jarndyce v. Jarndyce. We expedited this appeal, and counsel assisted this court by providing much of the record and the briefs in computerized format. We will direct that this case be given priority in the trial court and that all available means to expedite the remaining triable issues be implemented.

BACKGROUND FACTS

Much of the evidence from the first trial was introduced on retrial and although we set out the trial court's findings in this opinion, the interested reader should review Paterno I. (See Paterno I, supra, 74 Cal.App.4th at pp. 75-91, 96-99, 87 Cal. Rptr. 2d 754.) Judge Golden adopted parts of Judge Lorenzo Sawyer's decision in Woodruff v. North Bloomfield Gravel Mining Co. (C.C.D.Cal.1884) 18 Fed. 753 (Woodruff), which famously declared certain hydraulic gold mining practices to be a nuisance. Woodruff helps explain the origin of Linda levee's problems.

"Hydraulic mining ... is the process by which a bank of gold-bearing earth and rock is excavated by a jet of water, discharged through the converging nozzle of a pipe, under great pressure, the earth and debris being carried away by the same water, through sluices, and discharged on lower levels into the natural streams and water-courses below." (Woodruff, supra, 18 Fed. at p. 756.) The technology improved until large pipes, or monitors, could discharge 185,000 cubic feet of water per hour at a speed of 150 feet per second, and "at the North Bloomfield, several of these Monitors are worked, much of the time, night and day, the several levels upon which they are at work being brilliantly illuminated by electric lights, the electricity being generated by water power. A night scene ... is in the highest degree weird and startling, and it cannot fail to strike strangers with wonder and admiration." (Id. at p. 757.)

But admiration was far from universal. The environmental damage is indescribable, and must be seen at the Malakoff Diggins State Historic Park to be believed. (See Woodruff, supra, 18 Fed. at p. 757 [scale of project "can only be duly appreciated by actual observation"].) Millions of cubic yards of "slickens" (fine wet mining debris) filled up the river beds and ruined vast agricultural tracts. (Id. at pp. 758-760.) Levees built to protect Linda township failed in 1881 and 1883 and the space between the levees filled with debris. (Woodruff, supra, 18 Fed. at pp. 760, 765-767.)

"The California Debris Commission (CDC) was formed by Congress [in 1893] to counter the effects of hydraulic mining" and the Army Corps of Engineers (Corps) worked on the Yuba. (Western Aggregates, Inc. v. County of Yuba (2002) 101 Cal. App. 4th 278, 287, 130 Cal. Rptr. 2d 436; see 33 U.S.C.A. § 661; Gray v. Reclamation District No. 1500 (1917) 174 Cal. 622, 628-630, 163 P. 1024 (Gray ).) In 1911, the State adopted the Jackson Report as its flood control plan. (Reclamation District v. Riley (1923) 192 Cal. 147, 149-150, 218 P. 762 (Riley).) On appeal the State describes the report as "a skeletal or conceptual plan for a flood control system to be developed over time with the benefit of further studies and experience...." (See Gray, supra, 174 Cal. at pp. 629-630, 163 P. 1024 ["the details of it were still to be worked out. The reclamation board ... was called into existence to do these things"].) The Sacramento River Flood Control Project (SRFCP) was based on the Grant Report (a modification of the Jackson Report) approved by California in 1925 and by Congress in 1928. (American Riv. Flood Control Dist. v. Sweet (1932) 214 Cal. 778, 781-782, 7 P.2d 1030; Beckley v. Reclamation Board (1962) 205 Cal. App. 2d 734, 740-741, 23 Cal. Rptr. 428 (Beckley); 11 Ops.Cal.Atty.Gen. 93, 93-94 (1948); Wat.Code, § 8525 [report as modified approved as a plan for flood control].)

"In 1953, the SRFCP works were transferred to the state. A memorandum of understanding confirmed the state's obligation to operate and maintain all completed works of the SRFCP and to hold the federal government harmless. The state turned the levees over to [local reclamation] districts for maintenance and operation but maintained responsibility for the project." (Akins v. State of California (1998) 61 Cal. App. 4th 1, 11, 71 Cal. Rptr. 2d 314 (Akins).) The State agreed to this plan for financial reasons. (See 9 Ops.Cal.Atty.Gen. 87, 89-91 (1947).) At trial the State took responsibility for "policy-making functions as they pertain to all flood control matters of the SRFCP," and by statute it has "supervisory powers" over the SRFCP. (Wat.Code, § 8360; see Paterno I, supra, 74 Cal.App.4th at p. 77, 87 Cal. Rptr. 2d 754.)

As we have said before, the SRFCP "consists of a vast and intricate general plan for levees, bypasses, weirs and other works designed for flood control, reclamation and improvement of navigation. It is a cooperative federal-state venture which has been in the process of design and construction for over half a century. It has been described in so many reported appellate decisions that further exposition is unnecessary." (Sacramento, etc. Drainage Dist. ex rel. State Reclamation Bd. v. Reed (1963) 215 Cal. App. 2d 60, 65, 29 Cal. Rptr. 847.) Before this comprehensive scheme existed, flood control consisted of public or private projects protecting small areas, often conflicting with other projects, or "dog-eat-dog reclamation" as we have called it. (Beckley, supra, 205 Cal.App.2d at p. 740, 23 Cal. Rptr. 428; see People v. Sacramento Drainage Dist. (1909) 155 Cal. 373, 379-381, 103 P. 207.) This followed from the "common enemy" rule which allowed each landowner to fend off flood waters regardless of the effect on other lands. (See In re Sutter-Butte By-Pass Assess. No. 6 (1923) 191 Cal. 650, 656, 218 P. 27; Van Alstyne, Inverse Condemnation: Unintended Physical Damage (1969) 20 Hastings L.J. 431, 499-502 [criticizing doctrine as applied to public projects] (Van Alstyne).)

As stated, 19th-century levee projects failed near Linda. (Woodruff, supra, 18 Fed. at pp. 760, 765-767.) In 1904 Yuba County adopted a resolution authorizing construction of a levee known as the Morrison Grade, which became the Linda levee. It was built by men and horses using scrapers to borrow nearby material, mostly mining debris. The trial court found: "In the process, little or no compaction of the material was attempted or achieved. [¶] As built, Morrison Grade was highly susceptible to seepage failure because of its siting on top of fifteen feet of porous hydraulic mining debris, the porosity of the material of which it was constructed, and the absence of any compaction of that material during construction." The Linda levee was part of the District, formed in 1908, and incorporated into the SRFCP.

Pursuant to the Grant Report the Corps improved the levee in 1934 and 1940 but the trial court found "the existing levee was incorporated into the finished work" without change. The floods of 1955 sorely tested the SRFCP, and exposed many deficiencies, but no problems were revealed in the Linda levee, in particular the south levee on the Yuba between the Southern Pacific Railroad and the E Street Bridge (before the Feather confluence). Although the flood stage exceeded design capacity and water came within a foot of the top, it held. In the 1964 flood year the Linda levee was also subjected to higher waters than in 1986, yet held.

"In February 1986, a tropical weather system brought much warm rain, which in turn caused snow melt" triggering massive flooding in California. (Paterno I, supra, 74 Cal.App.4th at p. 77, 87 Cal.Rptr.2d 754.) For three days the Linda levee held water reaching to 76 feet (U.S.Eng.Datum), but it failed when the water had receded to about 74.3 feet; it is designed to hold up to 80 feet. The State concedes the levee failed at about half its designed capacity. The trial court found "the resulting 150 foot gap in the embankment allowed roughly 20,000 acre-feet of water ... to inundate some 7,000 acres of land situated in the communities of Linda and Olivehurst, lying across the river south of Marysville in a territory which had been protected by the [levee] from flooding for many years. The flooding resulted in damage ... estimated to be in the range of one hundred million dollars." "By 1986, the value of property protected by the levee ... was about $409,400,000. [¶] There was evident no general perception that the area was not a safe place for urban development."

The parties stipulated Paterno's property (real and personal) was damaged as a proximate result of the failure of a public flood control project to function as planned. The trial court found it failed because seepage had eroded soil from the levee, probably over decades, resulting in an impaired foundation which could not withstand the third major flood test. Underneath the levee were channels from prior geologic river configurations, providing watercourses which made the levee vulnerable to seepage. However, the trial court reasoned that the defendants had not built the levee and therefore were not responsible for creating these problems, characterizing proposed available fixes as "upgrades" upon which liability could not be based.

In his objections to the proposed statement of decision, Paterno partly complained that there was no basis for the trial court's "conflation of ``feasible alternatives' and ``upgrades.'" Paterno also urged that liability be predicated on the Jackson and Grant reports, which in effect ratified the Linda levee's configuration on the ground, i.e., as Yuba County had sited and built it. This was based on a theory known as the "acceptance" doctrine. Paterno asked the trial court to balance the Locklin factors — which we discuss in detail later — and determine if the plans were unreasonable. Paterno raised other points which have been abandoned. (See Estate of Randall (1924) 194 Cal. 725, 728-729, 230 P. 445.)

The trial court adopted the tentative decision, with insignificant changes, as the statement of decision.

We disagree with the trial court's legal conclusions, but the thorough yet concise statement of decision has been of enormous assistance. Because the parties dispute its meaning, we quote it at length, and italicize certain critical portions.

STATEMENT OF DECISION EXCERPT

"CAUSE OF FAILURE

"The physical process implicated in the failure of the levee was one in which water ... seeped through the levee's foundation (the ground upon which the levee's embankment was constructed) and so eroded it as to permit the levee's embankment (that portion of the earthen structure which was constructed above the foundation) to collapse into the weakened foundation thereby creating an opening in the embankment through which the river's waters flooded. The precise physical phenomena which accompanied the process were described in considerable detail by well-informed and credible experts whose accounts and opinions varied in some details but were fairly consistent in advancing the general proposition that seepage, attributable to the natural physical properties of the elements implicated in the process — soil and water — produced the failure.

"It is likely that the process of seepage and erosion of the foundation was one which had been underway during high water events in the decades preceding the failure [e.g., in 1955 and 1964] and culminated in collapse on that occasion because the degree of erosion produced during the event of February 1986, when added to that produced during such events in the preceding decades, resulted in a foundation critically impaired.

"Throughout the scenario of the construction and collapse of the Linda levee, seepage had been a constant presence in the lower reach of the Yuba — and elsewhere in the Sacramento River drainage — and was an entity whose vigor and effect were affected by the permeability of the soil in which it existed.

"The site of Linda levee is one characterized by the deposition of hydraulic mining debris to a depth of approximately 15 feet, under which lay natural over-bank deposits, under which were sand channels, under which was very [coarse] gravel with sand seams. The ... debris was highly porous material which could have a profound effect on the stability of a levee, long term. Moreover, in the area where the levee failed, there were 5 different former river channels which had been there in the recent geologic past and were natural courses for the movement of underground water.

"The embankment of the levee had been built with mining debris dug out of borrow pits near the levee and was characterized as very unstable and loose; the upper part of the foundation was characterized as very poor. These conditions rendered the levee susceptible to becoming unstable as a result of seepage.

"One of plaintiff's experts, Meehan, characterized the levee as an inferior, high-risk levee which was poorly constructed and didn't meet any engineering standards that existed any time during its life; it was built on a very unstable foundation which was subject to severe seepage pressure and offered little resistance to seepage over the course of its history; the embankment was composed of loose, sandy material and its composition and construction were not adequate. This is an indictment which the evidence supports. (Italics added.)

"Nevertheless, the cause of the failure of this levee is found to be that which was stated at the outset of this discussion: an interaction of the physical properties of natural elements: soil and water.

"CONTROLLING LEGAL PRINCIPLES

"... [¶] ... [¶]

APPLICATION OF PRINCIPLES TO EVIDENCE

"Pursuant to a stipulation of the parties, a pre-trial order ... determined that [the failure of the Linda levee] proximately caused damage to sample plaintiffs' property.

"The inquiry, then, becomes whether the evidence demonstrates that the failure of Linda levee was substantially caused by a plan adopted by the State or [District] for the design, construction, operation or maintenance of the levee.

"It does not.

"The evidence demonstrates that the failure of the levee was caused entirely by a natural process involving natural elements.

"Moreover, the plan of design and construction of Morrison Grade, out of which Linda levee evolved, was one adopted by [Yuba County] and if that plan were a substantial cause of the levee's failure by reason of its specification of siting, construction materials or ... techniques, such a circumstance would not engage liability of [Defendants].

"PLAINTIFFS' CONTENTIONS

"THE ``PLAN'

"[Paterno advocates] the view that ... the ``plan' which is seen as the substantial cause of the failure of the ... Linda levee, is the ``plan' for the [SRFCP].

"Pursuit of [this] contention requires a consideration of the question whether a plan adopted by the State for the [SRFCP] caused Linda levee to fail. The contention is ... the ``plan' adopted ... in 1911, when the [Jackson Report was adopted] was the genesis of a plan which is not found in any particular document, but is a plan which ... had been constantly evolving and manifesting itself in the many [plans of the Project.]

"The evidence does not demonstrate that any such amorphous ``plan' for the [SRFCP] played a causal role in the failure of Linda levee.

"Moreover, the ``plan' which the relevant legal principle refers to is a discrete plan, considered and adopted by an authorized governmental decision-making entity, for the design, construction, operation or maintenance of a specific flood control project which failed. That project is identified as the Linda levee and a relevant plan must be one having something to do with the physical properties of that project or a prescribed system of its operation or maintenance.

"There are four such plans identified by the evidence.

"The first of them was the plan approved in 1904 by [Yuba County] for the original construction of Morrison Grade, but, for reasons earlier discussed, it cannot form the basis for liability of [defendants.]

"The second was the plan approved ... for the 1934 work ... but no element of that plan ... contributed to the 1986 failure of the levee....

"The same evaluation is made with respect to the third plan, that involving the [1940 improvements].

"A fourth plan was one resulting in work done on Linda levee in 1960 [.] There is no evidence that the work played any role in causing the [levee failure].

"In conclusion, the evidence does not demonstrate the existence of any relevant plan which was a substantial cause of the failure of the levee.

"ALIGNMENT

"[Paterno contends] that a substantial factor in the failure of the levee was the selection of its alignment along a course in the active channel of the river, underlain by permeable hydraulic mining debris from which construction material for the embankment was borrowed, over former channel beds, whereas there was a preferred course southerly of the one selected, lying outside of the area of the active channel of the river, the former channels and the hydraulic mining debris.

"The decision to select the course selected was made by [Yuba County], before [defendants] had any involvement with the levee and cannot be said to represent a plan adopted by either of them. (Italics added.)

"... [¶] ... [¶]

"SEEPAGE CONTROL

"Seepage is the underground movement of water between a stream and adjacent lands. It has been an historic and persistent condition along the Sacramento River and its tributaries, including the Yuba[.] Seepage occurred through and under the levees of the [SRFCP]. Linda levee was in an area in which there was a recorded history of seepage and siting the levee in the Yuba River channel, when it was originally constructed, markedly increased the seepage failure potential. (Italics added.)

"A 1955 State report [citation] recommended a study of modifications of levees in the project system which might mitigate or prevent seepage. After the failure of the Linda levee [an evaluation], of the system by the Corps in 1990 found that the project levees were susceptible to seepage problems which deprived them of their ability to provide design levels of flood protection. The [Corps] report recommended seepage control measures.

"Before the failure of the levee, however, there were seepage control measures available which had been evolving during the 20th century and, in 1978, design standards [citation] adopted on behalf of the Corps for project levees prescribed seepage control measures ... none of which was possessed by Linda levee.... [T]hese were techniques that were employed by the Corps in performing repairs to the levee. They proved to be effective during the bigger high-water event of 1997 and would have been effective in 1986[.]

"When the 1934 and the 1940 work was done on Linda levee, similar seepage control measures were available but not used. Had the available measures been used in the design of the 1934 or 1940 work, it is probable that the levee would not have failed at the site where such measures had been used. Thus, the argument can be made that the plan for the 1934 work and the plan for the 1940 work, each, was a substantial cause of the failure of the levee in 1986 for the reason that each plan failed to provide for the use of available seepage control measures which probably would have prevented the failure. A similar argument can be made that, had the original levee been removed and a new levee segment constructed in accordance with current engineering standards at the time of either of those projects, rather than simply incorporating the old segment into the new work, the 1986 failure probably would have been prevented. Likewise, it could be said that, at the time of the 1934 and 1940 projects, the entire alignment of the Linda levee could have been [fixed].

"Although the factual foundation for each of these contentions is sound, each is rejected. (Italics added.)

"Each of the 1934 and 1940 projects was designed for the purpose of achieving specified and discrete levee design characteristics and the objective of each was achieved. There was no failure of either plan to achieve its objective and it was not the failure of either plan to achieve its objective which was a substantial cause of the failure of the levee in 1986....

"Moreover, the provision of seepage control features and the reconstruction of the levee segments or their realignments would constitute upgrades in the condition of the levee which the State was not required to provide and which the court may not consider as the basis for imposition of inverse condemnation liability ([Paterno I,] supra, 74 Cal.App.4th at pp. 96-98, 87 Cal. Rptr. 2d 754).

"CONCLUSION

"Since the evidence does not demonstrate that the failure of Linda levee on February 20, 1986 was substantially caused by a plan adopted by either defendant for the design, construction, operation or maintenance of the levee, it would be an incongruous exercise to attempt to determine whether any such plan was unreasonable and that effort will not be undertaken.

"The ultimate conclusion resulting from the foregoing discussion is that plaintiffs shall recover nothing from either defendant."

The trial court issued a defense judgment and Paterno filed a timely notice of appeal therefrom.

DISCUSSION

Paterno does not contest the factual findings, he asserts the facts "compel the legal conclusion that the Linda levee's failure to function as intended was a legal cause" of his damages. Defendants did not object to the statement of decision, and they, too, largely embrace the trial court's findings, arguing they do not show liability.

We will first discuss the trial court's causation finding and conclude the trial court found the initial poor construction of the levee caused a seepage failure, and that feasible repairs were not undertaken. We will then discuss the rules applicable when a flood control project fails to function as intended, flooding properties historically subject to flooding. Applying those rules we will conclude the State, but not the District, is liable for Paterno's damages, because of the unreasonable plan within the SRFCP which accepted the levee as built without any measures to ensure it met design standards. Then we will discuss the two legal reasons given by the trial court to absolve the State of liability. We will conclude that the fact Yuba County built the levee does not relieve the State of liability because the State accepted the levee within the SRFCP. Moreover, the State never added seepage controls, and we explain why imposing liability for not doing so does not impose liability for failing to upgrade the system. We conclude there is no basis to apportion any liability to the District.

I. The Cause of the Failure

In Paterno I we emphasized Paterno would have to prove that some unreasonable aspect of an official plan caused the levee to break. (Paterno I, supra, 74 Cal.App.4th at p. 91, 87 Cal. Rptr. 2d 754.) "Where damage results from the acts of employees, and not from a policy decision, there is no taking. Recovery, if any, lies in a tort action, such as negligence. [Citation.] In the case of alleged shoddy maintenance ... it is the plan of maintenance which must be unreasonable to establish a taking. Poor execution of a maintenance plan does not result in a taking." (Id. at pp. 86-87, 87 Cal. Rptr. 2d 754.) Paterno quotes a snippet of transcript to suggest the trial court thought we erred in Paterno I in our "``insistence on focusing on a plan for the purpose of establishing inverse condemnation liability[.]'" The trial court properly implemented Paterno I on retrial, and required Paterno to show an unreasonable plan caused his damages.

The State points to parts of the decision to argue that the trial court found no plan for the levee contributed to its collapse. For example, the trial court concluded that "seepage, attributable to the natural physical properties of the elements implicated in the process — soil and water — produced the failure [;]" "the cause of the failure of this levee is found to be that which was stated at the outset of this discussion: an interaction of the physical properties of natural elements: soil and water[;]" the "evidence demonstrates that the failure of the levee was caused entirely by a natural process involving natural elements."

The State is correct but only in a superficial way. In the language used by the State, the trial court did find no relevant plan caused Paterno's damages, but this was based on erroneous legal premises. A fair reading of the statement of decision shows that factually the trial court found the initial levee construction was abysmal and that feasible technology existed in the 1930's and 1940's which, if implemented, would have brought the levee within engineering standards and averted the failure. As for the former point, the trial court credited Paterno's expert, who "characterized the levee as an inferior, high-risk levee which was poorly constructed and didn't meet any engineering standards that existed any time during its life; it was built on a very unstable foundation which was subject to severe seepage pressure and offered little resistance to seepage over the course of its history; the embankment was composed of loose, sandy material and its composition and construction were not adequate. This is an indictment which the evidence supports." (Italics added.) As for the latter, the trial court found feasible curative measures "were available but not used. Had the available measures been used in the design of the 1934 or 1940 work, it is probable that the levee would not have failed at the site where such measures had been used."

The trial court rejected Paterno's claim that any plan of the defendants contributed to the levee failure for two legal reasons: (1) Defendants were not liable for the original (1904-1905) alignment or construction of the levee because Yuba County built it; and (2) liability could not be predicated on subsequent plans (the repairs in the 1930's and 1940's) because those plans achieved their design goals and there can be no inverse liability for failing to upgrade a project.

The State argues the SRFCP cannot be a relevant plan, but does not explain why simply because a plan incorporates a number of subsidiary plans, the larger plan cannot lead to liability. It was a central cost-saving feature of the early reports (which evolved into the SRFCP) to use existing levees, but there was never any effort to test those levees (or at least, the Linda levee) for structural soundness. The global plans assumed the levee met engineering standards, despite the fact that the records of its construction were public and showed that mining debris was simply scraped up and heaped, without compaction, to form the Morrison Grade, which later was raised and slightly reshaped, retaining the defective core. The State claims "Linda levee as it existed after the work performed in 1934 was incorporated into the finished 1940 work[,] in conformance with the levee design standards of the day," but although the improvements may have been designed to the standards of the day, the trial court found the levee never met those standards.

The State asserts the trial court found the levee failed due to "hydro-consolidation" and defines that theory as "a physical process whereby sandy material consolidates or settles when loaded or reloaded (with weight or additional weight) and exposed to water. Hydro-consolidation began at Linda levee decades before its failure when flood waters entered subsurface flood plain soils and from there, formed a subterranean pathway for water that later flood events progressively extended closer to the levee's foundation. Once underground flood waters eventually permeated the levee's foundation material to reach a point inland of its landside toe in 1986, ``hydro-fracture' caused a sudden, catastrophic failure by rapid evacuation of large amounts of soil from the levee's foundation."

But the statement of decision never uses the terms "hydro-fracture" or "hydro-consolidation." (See Lafayette Morehouse, Inc. v. Chronicle Publishing Co. (1995) 39 Cal. App. 4th 1379, 1384, 46 Cal. Rptr. 2d 542 ["When the record clearly demonstrates what the trial court did, we will not presume it did something different"].) The State gives a lengthy explanation of the evidence in its effort to show the trial court meant "hydro-consolidation" caused the failure. We are confident that if that is what Judge Golden meant, he knew how to say it. Moreover, the State's point appears to be that if we agree a "hydro-fracture" was the immediate cause of the failure we would have to conclude the failure was unforeseeable, a claim we discuss below.

We reject the State's efforts in this court to impliedly relitigate the factual cause of the levee failure.

II. Inverse Liability for Flood Projects

In this section we explain the general liability rules which governed the retrial. We will address the significance of foreseeability, vel non. Because the trial court concluded no relevant plan caused the failure, it did not proceed to balance the Locklin factors. We will do so on appeal.

A. General Rules.

The taking "or damag[ing]" of private property for "public use" must be compensated. (Cal. Const., art. I, § 19; Locklin, supra, 7 Cal.4th at p. 362, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) Generally, "whether or not the public improvement involved was made with care and skill is irrelevant." (29 Cal.Jur. (3d ed.1986) § 304, pp. 454-455, fns. omitted.) The public should pay the costs inherent in public works, including damages, foreseeable or not. (Holtz v. Superior Court (1970) 3 Cal. 3d 296, 310-311, 90 Cal. Rptr. 345, 475 P.2d 441 (Holtz) [since the undertaking at a lower cost created some risk of damage to private property, it was proper to require the public to bear the loss]; Albers, supra, 62 Cal.2d at pp. 261-264, 42 Cal. Rptr. 89, 398 P.2d 129.)

"Inverse condemnation liability ultimately rests on the notion that the private individual should not be required to bear a disproportionate share of the costs of a public improvement" and where liability results, "compensation ``constitutes no more than a reimbursement to the damaged property owners of their contribution of more than their "proper share [to] the public undertaking."'" (Locklin, supra, 7 Cal.4th at pp. 367-368, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) For example, if a project intentionally floods lands not otherwise subject to flooding for the purpose of protecting other lands, it would be unfair to make the flooded property owners subsidize the others. (Akins, supra, 61 Cal.App.4th at pp. 31-33, 71 Cal. Rptr. 2d 314; see Odello Brothers v. County of Monterey (1998) 63 Cal. App. 4th 778, 791-792, 73 Cal. Rptr. 2d 903.)

The following passage of Paterno I, supra, 74 Cal.App.4th at pages 82-83, 87 Cal. Rptr. 2d 754, sets out the law governing the retrial:

"When a flood control project fails to function as intended, causing damage to properties historically subject to flooding, strict liability for a taking does not apply. [Citation.] Instead, a rule of reasonableness must be applied, as Paterno concedes. This rule arose in [Belair v. Riverside County Flood Control Dist. (1988) 47 Cal. 3d 550, 253 Cal. Rptr. 693, 764 P.2d 1070 (Belair)] and [Locklin] and balances the need for flood control projects against the damages occasioned by their failure, by means of weighing a number of specific factors. [Citations.] Here damage was caused by a failure of the levee and the lands were historically subject to flooding, which explains why the levee was built...."

"Foreseeability does not suffice. "Plan or design characteristics that incorporate the probability of property damage under predictable circumstances may later be judicially described as ``negligently' drawn; yet, in the original planning process, the plan or design with its known inherent risks may have been approved by responsible public officers...." (Van Alstyne, [supra, 20 Hastings L.J. at pp. 489-490, fns. omitted, approved, Bunch v. Coachella Valley Water Dist. (1997) 15 Cal. 4th 432, 450, 63 Cal. Rptr. 2d 89, 935 P.2d 796 (Bunch)].)

"``[T]he placement, design, and construction of even the most effective system inherently involve a complex balancing of interests and risks.... The dangers posed to individual lands by the failure of any public flood control project are "potentially enormous" and sometimes deserve compensation. However, strict and "open-ended" liability for the failure of a project whose overall design, construction, operation, and maintenance was "reasonable" would unduly deter the development of these vital bulwarks against common disaster.... [¶] [Bunch concluded:] "... [A] flood control agency does not necessarily exact ``disproportionate,' and thus compensable, contributions from particular landowners simply because it constructs adjacent flood control improvements that may alter how floodwaters will affect those landowners if the improvements fail to contain the flow. When a public flood control system fails to protect land from historic periodic flooding, the only way to determine whether a damaged private landowner has thereby been forced to contribute a compensable ``disproportionate' share of the public undertaking is to determine whether the system, as designed, constructed, operated, and maintained, exposed him to an ``unreasonable' risk of harm, either individually or in relation to other landowners.'"

The Locklin court approved two partially overlapping sets of factors to be used in making the reasonableness calculus, quoting from Albers, supra, 62 Cal.2d at page 263, 42 Cal. Rptr. 89, 398 P.2d 129, and adapting from Van Alstyne's article. (Locklin, supra, 7 Cal.4th at pp. 368-369, 27 Cal. Rptr. 2d 613, 867 P.2d 724; see Akins, supra, 61 Cal.App.4th at pp. 26-27, fn. 18, 71 Cal. Rptr. 2d 314.) When we speak of the "Locklin" factors we will refer to both the "Van Alstyne" factors and the "Albers" factors.

The Van Alstyne factors have been summarized as follows (Bunch, supra, 15 Cal.4th at p. 446, 63 Cal. Rptr. 2d 89, 935 P.2d 796):

"``(1) The overall public purpose being served by the improvement project; (2) the degree to which the plaintiff's loss is offset by reciprocal benefits; (3) the availability to the public entity of feasible alternatives with lower risks; (4) the severity of the plaintiff's damage in relation to risk-bearing capabilities; (5) the extent to which the damage of the kind the plaintiff sustained is generally considered as a normal risk of land ownership; and (6) the degree to which similar damage is distributed at large over other beneficiaries of the project or is peculiar only to the plaintiff.

"In addition, ... ``[r]easonableness ... also considers the historic responsibility of riparian owners to protect their property from damage caused by the stream flow and to anticipate upstream development that may increase that flow.... [P]laintiff must demonstrate that the efforts of the public entity to prevent downstream damage were not reasonable in light of the potential for damage posed by the entity's conduct, the cost to the public entity of reasonable measures to avoid downstream damage, and the availability of and the cost to the downstream owner of means of protecting that property from damage.'"

A footnote says "inquiry into ``reasonable' design, construction, operation, and maintenance is not limited to a narrow examination whether the system's technical specifications, intended capacities, materials, workmanship, and repairs were adequate under all the circumstances. [Citation.] Instead, the inquiry should include specific consideration whether the location and configuration of the system, and its purpose to divert the natural flow, were themselves ``reasonable.'" (Bunch, supra, 15 Cal.4th at p. 446, fn. 3, 63 Cal. Rptr. 2d 89, 935 P.2d 796.)

Before discussing the Van Alstyne factors, Locklin stated (7 Cal.4th at p. 368, 27 Cal. Rptr. 2d 613, 867 P.2d 724):

"The factors which the court identified as important in imposing liability in Albers, supra, 62 Cal. 2d 250, 263, 42 Cal. Rptr. 89, 398 P.2d 129, are also important here: ``First, the damage to [the] property, if reasonably foreseeable, would have entitled the property owners to compensation. Second, the likelihood of public works not being engaged in because of unseen and unforeseeable possible direct physical damage to real property is remote. Third, the property owners did suffer direct physical damage to their properties as the proximate result of the work as deliberately planned and carried out. Fourth, the cost of such damage can better be absorbed, and with infinitely less hardship, by the taxpayers as a whole than by the owners of the individual parcels damaged. Fifth, ... "the owner of the damaged property if uncompensated would contribute more than his proper share to the public undertaking."'"

The trial court proposed that if it got to the Locklin weighing stage, that is, if it found a relevant plan which caused Paterno's damages, it would weigh these Albers factors in addition to the Van Alstyne factors. The State argued the Albers factors were inappropriate because Albers itself involved strict liability and because those factors were not discussed in Bunch or Paterno I. We disagree with the State.

First, as Professor Arvo Van Alstyne himself noted, Albers was not a true "strict liability" case: "Three important qualifications are indicated. First, Albers supports liability absent foreseeability of injury (i.e., without fault) only when inverse liability would obtain in a situation involving the same facts plus foreseeability (i.e., plus fault). Secondly, the rule is limited to instances of ``direct physical damage.' Finally, the damage must be ``proximately caused' by the public improvement as designed and constructed." (Van Alstyne, supra, 20 Hastings L.J. at pp. 434; id. at pp. 434-438 [discussing these points in detail]; see Holtz, supra, 3 Cal.3d at p. 304, 90 Cal. Rptr. 345, 475 P.2d 441.)

Second, in Akins we quoted the portion of Locklin quoting the Albers factors and stated, "We assume all of the foregoing factors could be properly considered." (Akins, supra, 61 Cal.App.4th at pp. 26-27, fn. 18, 71 Cal. Rptr. 2d 314.) We adhere to that view.

Third, in Locklin the California Supreme Court called the Albers factors "also important here" (Locklin, supra, 7 Cal.4th at p. 368, 27 Cal. Rptr. 2d 613, 867 P.2d 724), and neither Paterno I nor the later California Supreme Court case of Bunch purported to give an exhaustive list of factors. Locklin repeatedly emphasized that the purpose of balancing is to determine if a disproportionate burden has been inflicted by a public project. (Locklin, supra, 7 Cal.4th at pp. 366, 369, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) The Locklin factors are not elements of a cause of action for inverse liability, but, when balanced, indicate whether "the owner, if uncompensated would contribute more than his proper share of the public undertaking." (Akins, supra, 61 Cal.App.4th at p. 27, fn. 18, 71 Cal. Rptr. 2d 314; see Belair, supra, 47 Cal.3d at p. 558, 253 Cal. Rptr. 693, 764 P.2d 1070 [decisive consideration]; Barham v. Southern Cal. Edison Co. (1999) 74 Cal. App. 4th 744, 752, 88 Cal. Rptr. 2d 424 ["fundamental policy ... is to spread among the benefiting community any burden disproportionately borne by a member of that community, to establish a public undertaking for the benefit of all"].) This mode of analysis stems from the shift of inverse liability away from tort and private property concepts and towards policy-based constitutional analysis. (See Holtz, supra, 3 Cal.3d at p. 303, 90 Cal. Rptr. 345, 475 P.2d 441 ["``to socialize the burden ... — to afford relief to the landowner in cases in which it is unfair to ask him to bear a burden that should be assumed by society'"], quoted with approval by Locklin, supra, 7 Cal.4th at p. 365, 27 Cal. Rptr. 2d 613, 867 P.2d 724; Clement v. State Reclamation Board (1950) 35 Cal. 2d 628, 642, 220 P.2d 897 (Clement).) Consideration of the Albers factors will help answer the question of disproportionate burden (although admittedly that ultimate issue itself is framed as the fifth Albers factor).

The true cost of a project must include certain deferred costs. (Paterno I, supra, 74 Cal.App.4th at p. 86, 87 Cal. Rptr. 2d 754.) Two inverse liability cases not involving flood control hold that because the cost-savings realized by plans deferring maintenance benefit the public, it is fair for the public to compensate the owners of the property which happens to be damaged when a failure caused by such plan of deferred maintenance takes place.

In McMahan's of Santa Monica v. City of Santa Monica (1983) 146 Cal. App. 3d 683, 194 Cal. Rptr. 582 (McMahan's), a city's plan ensured water pipes would be used past their lifetimes. (Id. at pp. 687-688, 693, 194 Cal. Rptr. 582.) When a landowner sued after a break, the city argued it was due to poor maintenance (for which inverse liability will not lie). But "whether the City's program of water main installation and replacement is characterized as ``construction' or ``maintenance,' the fact remains that it was inadequate and contributed to the break due to corrosion of the [main which failed]. The City's knowledge of the limited life of such mains and failure to adequately guard against such breaks caused by corrosion is [a deliberate act]." (Id. at pp. 695-696, 194 Cal. Rptr. 582.) "[T]he City was taking a calculated risk by adopting a plan ... it knew was inadequate. The City's plan of replacement of the water mains reflected the deferred risks of the project both foreseeable and unforeseeable, and it is proper to require the City to bear the loss when the damage occurs." (Id. at pp. 697-698, 194 Cal. Rptr. 582.) In Pacific Bell v. City of San Diego (2000) 81 Cal. App. 4th 596, 96 Cal. Rptr. 2d 897 (Pacific Bell) the lack of any plan to monitor pipe deterioration saved money, therefore "The burdens attending City's cost-saving approach should be spread to the community benefiting from lower water rates rather than imposing the entire cost on those property owners placed in harm's way by City's program." (Id. at pp. 607-608, 96 Cal. Rptr. 2d 897.)

In these two examples, the harm was foreseeable because the plans guaranteed failures. But we must not conflate inverse liability with tort liability. Inverse liability is generally not based on fault. An exception has been carved out for failures of flood control projects in areas historically subject to flooding, requiring use of the Belair-Locklin reasonableness calculus, but this does not reimport traditional notions of fault or foreseeability into this branch of inverse liability. "A constitutional analysis for determining inverse condemnation liability in the flood control context should not include ``a fruitless search for the somewhat artificial moral elements inherent in the tort concepts of negligence and intentional wrongs.'" (Bunch, supra, 15 Cal.4th at p. 449, 63 Cal. Rptr. 2d 89, 935 P.2d 796, quoting Van Alstyne, supra, 20 Hastings L.J. at p. 495.) Instead, liability is based "on the balancing of interests that [the California Constitution] requires. This balancing of interests serves both the private sector and public improvement efforts by addressing the cost-spreading objective of the just compensation clause while protecting public entities from unlimited, undeserved liability that could well inhibit further construction of public works." (Bunch, supra, 15 Cal.4th at p. 451, 63 Cal. Rptr. 2d 89, 935 P.2d 796; see Belair, supra, 47 Cal.3d at pp. 565-566 ["``Reasonableness, in this context, is not entirely a matter of negligence, but represents a balancing of public need against the gravity of private harm'"; quoting Van Alstyne, supra, 20 Hastings L.J. at p. 455].)

"In the presumably rare instance where substantial damage does in fact eventuate ``directly' from the project,[fn.] and is capable of more equitable absorption by the beneficiaries of the project (ordinarily either taxpayers or consumers of service paid for by fees or charges) than by the injured owner, [fn. omitted] absence of fault may be treated as simply an insufficient justification for shifting the unforeseeable loss from the project that caused it to [the] equally innocent owners. Absence of foreseeability, like the other factual elements in the balancing process, is, in effect, merely a mitigating but not necessarily exonerating circumstance." (Van Alstyne, supra, 20 Hastings L.J. at pp. 493-495.)

On appeal the State (but not the District) asserts the levee broke due to unforeseeable causes (e.g., "hydro-consolidation") and therefore the State cannot be liable to Paterno. We tend to agree with Paterno that the State has waived this argument. At trial the State objected to the trial court's proposal that it consider the Albers factors, which included foreseeability. Elsewhere at trial the State asserted it "has never contended that foreseeability was an issue in this case, ever." The State appears to be improperly changing its theory on appeal by claiming lack of foreseeability as a defense. (Richmond v. Dart Industries (1987) 196 Cal. App. 3d 869, 874, 242 Cal. Rptr. 184.) But the point is important and raises no new factual issues, so we will address it.

The State misapprehends the role of foreseeability. As Paterno points out, foreseeability plays no role in the causation analysis and is not determinative in the balancing step, only informative. This is not a case involving a dangerous condition of public property, and we are not applying tort or water law standards of liability. We are implementing the constitutional command that the State must compensate landowners when it damages their property. As Paterno points out, "Even if the State failed to appreciate the risk of failure, this is not a defense to proximate cause. Paterno I, 74 Cal.App.4th at 87, 87 Cal. Rptr. 2d 754. In other words, the levee's planned design and construction throughout its life ``endangered the levee in a way not adequately valued by the planners.' Id. at 98, 87 Cal. Rptr. 2d 754." We agree with Paterno's interpretation and the point has been made elsewhere. (Arreola v. County of Monterey (2002) 99 Cal. App. 4th 722, 762-763, 122 Cal. Rptr. 2d 38 (Arreola) ["the entity either failed to appreciate the probability that the project would result in some damage to private property, or ... took the calculated risk that damage would result"]; see Akins, supra, 61 Cal.App.4th at pp. 11, 13-14, 71 Cal. Rptr. 2d 314 [plan failed to include measures to close gap in flood control system].)

The "reasonableness" balanced here is not a negligence standard of care, which might turn on foreseeability, but is "determined by balancing the public benefit and private damage in each case." (Locklin, supra, 7 Cal.4th at p. 368, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) "In Belair the Supreme Court refined the ``proximate cause' element, noting Albers ``contained the seeds of confusion through its combination of "proximate cause" terminology with the elimination of foreseeability as an element of inverse condemnation.' [Citation.] The causation element is restated with greater precision in terms of ``substantial causation.'" (Akins, supra, 61 Cal.App.4th at p. 20, fn. 13, 71 Cal. Rptr. 2d 314; see Goebels v. City of Santa Barbara (2001) 92 Cal. App. 4th 549, 555, 111 Cal. Rptr. 2d 901 ["The injury need not be foreseeable, but the public improvement must be a substantial cause" of injury].)

Thus, while foreseeability may weigh in favor of the landowner, lack of foreseeability does not defeat the claim.

In this case the evidence overwhelmingly shows the failure of the levee was foreseeable. The State says "There were no events that reasonably put the State on notice before Linda levee failed that its capability to safely carry flood flows was or was becoming compromised." Paterno does not argue the State actually foresaw the levee failure. But the State must be charged with knowledge of how the levee was built. It operated the levee for decades and had ample opportunity to examine it. If it chose not to do so for fiscal reasons, that would indicate the loss should be absorbed by the State. We explain.

First, the method of construction and available technology in 1904 were detailed in public documents available to the State had it chosen to look at them. In a telling passage the State asserts: "The 1911 Plan is a policy for control of flooding in the Sacramento and San Joaquin Valleys by a unified system in accord with the general parameters the Jackson Report proposed. Its references to the Yuba River and its south bank as geographically part of the proposed system, and recommendation that ``present levees be used as far as practicable' within the proposed project, is not a ``plan' for the design, construction, operation or maintenance of Linda levee, a 3000 foot segment along the Yuba River. [Citations.] The report's proposals, including that ``no work is needed along the Yuba River except the protecting and strengthening of the south levee at a few points,' were subject to change based upon further studies and experience." But this passage concedes the State accepted the levee as built, and by implication that the State failed to undertake any studies to determine its adequacy to meet the waters the State was proposing to route against it. The State's later assertion that the 1986 levee was a "different structure" than the 1911 levee, because of the intervening changes, merely shows that the State made some efforts to improve the levee, (e.g., raising the height and grooming the crown) but never took steps to ensure its basic foundation was sound.

Second, in 1978 the Corps of Engineers adopted levee design standards and these standards discuss the problem of levees made of "uncompacted, or hydraulic fill" or those with "serious underseepage problems, weak foundation soils, or undesirable borrow materials[.]" The standards provide that basic general design procedure involves a geologic study followed by seepage analysis. (Id. p. 1-2.) Even what those standards describe as a mere "office study," that is, review of existing data on the levee in question (id. p. 2-1) rather than field subsurface testing (borings, seismic studies and so forth), would have revealed the poor construction of the Linda levee.

Third, in 1970, well before the levee failure, the District engineer sent a letter to the State, complaining (with specifics) about the sorry condition of the levee: "During past years, the south or left bank of the main channel of the Yuba River, from upstream of the Southern Pacific Railroad bridge to near the ``E' Street highway bridge, has gradually degraded, thereby endangering a shift of the main river channel from its historic location in the north channel to a new location in the south channel immediately adjacent to a substantial reach of Reclamation District No. 784's project levee. [¶] Reclamation District No. 784 views with alarm this potential shift of the main stream of the river to against its levee without adequate levee and bank protection works first being installed to insure the integrity of the levee. The levee at this location, consisting mainly of sand and founded on sandy materials, is believed by the District to be unsafe to withstand violent river flows as may occur." (Italics supplied.) Assuming the State was ignorant of the condition of its own levee up to that point, after this letter the State had 15 years to investigate this detailed warning before the collapse. So far as the record shows, the State did nothing.

In another case where the public entity was warned about the danger which later led to damage, the court held "The ``plan' was the long-term failure to mitigate a known danger. That failure persisted for 20 years." (Arreola, supra, 99 Cal.App.4th at p. 746, 122 Cal. Rptr. 2d 38.) Here, the State failed to acknowledge the danger despite adequate evidence at its disposal and we reject its claim the failure was "unforeseeable."

The State concedes that sometimes "the public entity failed to recognize a risk of harm inherent in its plan but should have, because the risk was foreseeable. In other words, the government failed ``to recognize the probability that, functioning as deliberately conceived, the public undertaking as altered and maintained would result in some damage to private property.'" This flows from the California Supreme Court's early observation that one measure of "fault" in such cases is the "failure to appreciate the probability that, functioning as deliberately conceived, the public improvement as altered and maintained would result in some damage to private property." (Bauer v. Ventura County (1955) 45 Cal. 2d 276, 286, 289 P.2d 1, quoted with approval in Van Alstyne, supra, 20 Hastings L.J. at p. 439; see also Paterno I, supra, 74 Cal.App. 4th at p. 98, 87 Cal. Rptr. 2d 754 [viable theory that "approval of the pit endangered the levee in a way not adequately valued by the planners"]; Arreola, supra, 99 Cal.App.4th at p. 746, 122 Cal. Rptr. 2d 38; Akins, supra, 61 Cal.App.4th at pp. 11, 13-14, 71 Cal. Rptr. 2d 314.) At best for the State, that happened here.

B. Application of the Locklin Factors.

The trial court did not address the Locklin factors because it found no relevant plan of either defendant caused the failure. Paterno asks that we engage in the Locklin reasonableness calculus on appeal, and both he and the State have set out an analysis of some of those factors in their briefs. Because of Judge Golden's detailed factual findings, the application of the Locklin factors to those facts presents a legal question. At bottom, based on the factual findings made by Judge Golden and the evidence at the second trial, we conclude the Locklin factors tilt sharply in Paterno's favor.

1. The Six Van Alstyne Factors.

a. The purpose served by the project.

The overall purpose served by the project is huge; according to the evidence, the levee system protected billions of dollars of property throughout the State. (See Akins, supra, 61 Cal.App.4th at p. 14 & fn. 6, 71 Cal. Rptr. 2d 314.) As the State concedes, the SRFCP purposes include flood control, reclamation of lands subject to flooding, and improvement of navigation. Although both this case and Akins demonstrate the system did not prevent all damage, the SRFCP exists to protect billions of dollars of property and millions of lives and largely accomplished its mission. Tellingly, in response to the trial court's question whether other areas benefited from the levee, the State replied: "Not specifically from the Linda levee, but certainly from the [SRFCP] of which the Linda levee is one component [.]" Just so: The Linda levee is part of a comprehensive system of flood control works and cannot be evaluated in isolation.

b. Offsetting reciprocal benefits.

Paterno (or his predecessors) paid for flood protection by taxes and by assessments to maintain the levee, as did all Californians who own land protected by the SRFCP. As Paterno notes, quoting a finding in another case, "``the longstanding negligent operation of a flood control project, such as is documented here, serves no legitimate purpose, nor does it promote any "reciprocal benefit" which offsets or justifies the damage that was caused by the failure of the Project.'" (Arreola, supra, 99 Cal.App.4th at p. 741, fn. 7, 122 Cal. Rptr. 2d 38.) At best for the State Paterno did benefit to the extent that his land was protected in 1955, 1964 and other high water years, but he shared that benefit with all others protected by the SRFCP. He received no offsetting benefit due to the defective levee.

c. Feasible alternatives.

The trial court found feasible alternatives would have saved the levee, by bringing it into design capacity.

Contrary to the District's view, Paterno's proof of feasible seepage controls is not "just another way or restating their previous claim on appeal from the first trial that Defendants were liable for failing to upgrade the levee." We disagree that introducing seepage controls to counter the poor construction, thus insuring the project actually met design standards, would have been an "upgrade."

We agree that fiscal constraints are a critical part of the feasibility analysis, and feasibility must account for the costs of the project as a whole. (Bunch, supra, 15 Cal.4th at pp. 451-452, 63 Cal. Rptr. 2d 89, 935 P.2d 796.) This is partly because the benefits of the entirety of the project (or plan) which causes damage must be balanced against the harm to those damaged. But "fiscal constraints are never alone determinative of the government's reasonableness in its flood control measures[.]" (Id. at p. 452, 63 Cal. Rptr. 2d 89, 935 P.2d 796.)

Paterno points to evidence in the record showing the curative measures were available at a reasonable cost. At trial Judge Golden examined the State budgets from 1949 to 1986 and indicated the State had adequate funds although the statement of decision is silent on that question, perhaps because the court did not proceed to the balancing step. Although the statement of decision does not recite the cost of seepage controls in the 1930's or 1940's, or at any other time, the tenor of the statement of decision indicates the court found the curative measures were fiscally feasible and the State makes no contrary claim on appeal.

Paterno claims that the existence of feasible alternatives shows the damages inflicted were not "necessary" to accomplish the public purpose and no more need be shown. But Locklin requires a court to weigh all relevant factors and not stop at this one, albeit critical, factor.

d. Risk-bearing capabilities.

The cost of Paterno's damage can better be absorbed, and with far less hardship, by the taxpayers, due to the severity of his damages in relation to risk-bearing capabilities. In a general passage of Paterno I we commented about the availability of flood insurance. (Paterno I, supra, 74 Cal.App.4th at p. 85, 87 Cal. Rptr. 2d 754.) The State asserts this indicates Paterno could adequately bear the risk of flooding by buying flood insurance. To the extent that passage of our prior opinion can be so read, we eschew it for the reasons stated by Paterno: Insurance does not eliminate the loss, it simply shifts the loss from the landowner to the insurer, which is then entitled to assert its subrogation rights. (Aetna Life & Casualty Co. v. City of Los Angeles (1985) 170 Cal. App. 3d 865, 873-875, 216 Cal. Rptr. 831 [rejecting claim that availability of fire insurance eliminates need for inverse liability; allowing insurers to recover]; McMahan's, supra, 146 Cal.App.3d at pp. 690-691, 194 Cal. Rptr. 582.)

Locklin requires consideration of "the availability of and the cost to the downstream owner of means of protecting" the property. (Locklin, supra, 7 Cal.4th at p. 369, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) There is nothing Paterno could have done to avoid the risk, because the project works funnel huge quantities of water through Linda and there is no defensive measure Paterno could have taken to turn away those waters. Moreover, it would be absurd to require landowners, whose taxes have paid for the levee system and whose yearly assessments pay for its maintenance, to construct secondary protective systems, individually or in groups.

e. Whether the damage is a normal risk of land ownership.

Flooding, particularly in the Sacramento Valley, is a normal risk of land ownership, and if the SRFCP (or similar works) did not exist, Paterno's property would have flooded anyway (and would have in 1955, 1964 and in many other years).

But over time artificial works became the natural condition and parties are generally entitled to rely on them. (See Clement, supra, 35 Cal.2d at p. 638, 220 P.2d 897; Paterno I, supra, 74 Cal.App.4th at p. 86, 87 Cal. Rptr. 2d 754; Beckley, supra, 205 Cal.App.2d at p. 751, 23 Cal. Rptr. 428; Van Alstyne, supra, 20 Hastings L.J. at pp. 454, 459, 492.)

As the California Supreme Court said in another case, "By inducing plaintiffs to make substantial improvements in reliance on its providing protection [to a certain capacity], and then failing to provide such protection, the levee plainly constituted a ``substantial cause' of plaintiffs' damages." (Belair, supra, 47 Cal.3d at p. 560, 253 Cal. Rptr. 693, 764 P.2d 1070.) So it is here.

The trial court found there was no perception of lack of safety and, as Paterno points out, the State itself relied on the levee, by locating a CalTrans yard behind it. (Cf. U.S. v. Zenni (E.D.Ky.1980) 492 F. Supp. 464, 466-468 & fn. 18 [nonverbal safety assertions].) The risk was not normal.

f. Distribution of damage across the project.

The flood waters inundated the properties below the levee and in the environs of the Linda and Olivehurst, rather than impacting the vast SFRCP-protected properties generally.

2. The Five Albers factors.

a. The damage, if reasonably foreseeable, would have entitled the property owners to compensation.

Had the State foreseen the levee was in danger of collapsing due to its poor alignment and composition, and refused to correct the problem, it would have been liable to Paterno on a theory of dangerous condition of public property. Indeed, that was Paterno's central theory at the first trial. (See Paterno I, supra, 74 Cal.App.4th at p. 100, 87 Cal. Rptr. 2d 754 [affirming first jury's verdict rejecting this theory].) As we explain elsewhere, despite the jury's verdict on the facts from the first trial, the facts found on retrial compel the conclusion that the failure was foreseeable, if not foreseen.

b. The likelihood of public works not being engaged in because of unforeseeable direct damage to property.

The SRFCP, of which the Linda levee is but one small component, would have been built regardless because even despite its isolated failures, it has saved many lives and billions of dollars by preventing floods, and it has opened or improved thousands of acres of land to productive use throughout the Sacramento Valley. Liability here would not likely deter future beneficial public works.

In Akins we pointed out that it is not always bad policy to discourage deleterious governmental practices. (Akins, supra, 61 Cal.App.4th at pp. 31-32, 71 Cal. Rptr. 2d 314.) We do not think imposing liability for maintaining a physically flawed structure which does not meet basic engineering standards will discourage reasonable planners from engaging in further flood control projects, it will only discourage them from failing to determine if the projects physically meet the designed standards and discourage them from failing to heed warnings about dangers lurking beneath the surface of projects. That, we think, is the appropriate public policy under the California Constitution.

c. The damage was the proximate result of the work as deliberately planned and carried out.

Proximate causation was partly stipulated at trial: It was stipulated that Paterno's damage was proximately caused by the failure of the levee. We have explained why that failure resulted from the project as planned.

d. The damage can better be absorbed, and with less hardship, by the taxpayers as a whole.

This is not a "deep pockets" question, but instead overlaps with the Van Alstyne factor of the relative risk-bearing capability of the landowner. This weighs in Paterno's favor. This does not mean that whenever a number of properties are flooded the taxpayers have to pick up the tab. That "would make flood control projects insurers against floods, a result eschewed by the California Supreme Court." (Paterno I, supra, 74 Cal.App.4th at p. 97, 87 Cal. Rptr. 2d 754.)

e. The owner if uncompensated would contribute more than a proper share to the public undertaking.

This Albers factor restates the object of the Locklin balancing exercise. We now summarize our views.

The public received the benefit from the levee without having to bear the expense of insuring it met the designed standards and was capable of carrying the water channeled to it by upstream features of the project. That the levee did not break in 1955 or 1964 is either miraculous or simply indicates "third time pays for all," meaning that the earlier high water events weakened the levee but not enough to cause a failure. The savings from not correcting the problems with the levee benefited the State and it would be unfair to require Paterno to bear all of the risk of that plan.

There is also a statutory policy to consider. The trial court found seepage caused the failure and the seepage was caused by the poor location and construction of the levee. Inverse liability for constructing water works with or over porous material, causing seepage, is not novel, although the cases usually involve damage from seeping water itself. (See, e.g., Turpen v. Turlock Irrigation Dist. (1903) 141 Cal. 1, 3, 74 P. 295 [canal seepage due to porous bed of sand]; Tormey v. Anderson-Cottonwood I. Dist. (1921) 53 Cal. App. 559, 568, 200 P. 814 (opn. of Supreme Ct.) [where damage "caused directly by seepage of water carried in said canal through the intervening soil on to the adjoining land[,]" plaintiffs need not show negligence].) But by statute, "It is declared to be the policy of the State that the costs of solution of seepage and erosion problems which arise or will arise by reason of construction and operation of water projects should be borne by the project." (Wat.Code, § 12627.3, Stats.1959, ch. 2128, § 1, p. 5030.) This policy weighs in Paterno's favor. (Van Alstyne, supra, 20 Hastings L.J. at pp. 464-465 ["statutory policy supports the view that seepage damage should be treated as a costs of the water project"].) The State relegates the seepage statute and policy to a footnote deriding Paterno's assertion as "having no application here. Moreover, the [trial court] did not find that the damage was caused by reason of Linda Levee." In light of our interpretation of the factual findings in the statement of decision, we reject the State's view.

Here, the seepage directly led to the levee collapse, and we see no reason why the flood damages should not be attributed to the project as a whole, rather than fall on the hapless property owners behind the levee. The State in effect gambled that the location and construction of the levee would prove adequate. A grossly disproportionate burden would fall on Paterno were his damages not spread out as part of the deferred costs of the project for flood control.

As stated, the trial court credited expert testimony that the levee "was poorly constructed and didn't meet any engineering standards that existed any time during its life[,]" which shows it was fortunate it did not break in 1955 or in 1964. Like a corroding pipe buried under Santa Monica or San Diego, the Linda levee was destined to fail. Therefore, Paterno has borne the deferred costs of maintenance of the system, which costs should instead be spread to the public at large, which benefited from that system. (Pacific Bell, supra, 81 Cal.App.4th at pp. 607-608, 96 Cal. Rptr. 2d 897; McMahan's, supra, 146 Cal.App.3d at pp. 697-698, 194 Cal. Rptr. 582 [plan shifted "the deferred risks of the project ... and it is proper to require the City to bear the loss when the damage occurs"]; see Belair, supra, 47 Cal.3d at p. 566, 253 Cal. Rptr. 693, 764 P.2d 1070 ["reimbursement ... of their contribution of more than their ``proper share [to] the public undertaking'"].)

III. The Acceptance Doctrine

The trial court held defendants could not be liable for the alignment or basic construction of the levee because those plans were crafted by Yuba County, a former defendant which got out of this case over a decade ago, after an appellate settlement conference. (Abbott v. County of Yuba (May 9, 1991, C009262).)

When a public entity accepts responsibility for an improvement, it becomes that entity's public improvement regardless of who built it. (Heimann v. City of Los Angeles (1947) 30 Cal. 2d 746, 756-757, 185 P.2d 597; Tyler v. Tehama County (1895) 109 Cal. 618, 626, 42 P. 240; Souza v. Silver Development Co. (1985) 164 Cal. App. 3d 165, 170, 210 Cal. Rptr. 146; Marin v. City of San Rafael (1980) 111 Cal. App. 3d 591, 595-596, 168 Cal. Rptr. 750; Sheffet v. County of Los Angeles (1970) 3 Cal. App. 3d 720, 734-735, 84 Cal. Rptr. 11; Stoney Creek Orchards v. State of California (1970) 12 Cal. App. 3d 903, 906-907, 91 Cal. Rptr. 139.) Some cases speak of "substantial participation," but the concept is the same: "A public entity is a proper defendant in an action for inverse condemnation if the entity substantially participated in the planning, approval, construction, or operation of a public project or improvement that proximately caused injury to private property. [Citation.] So long as the plaintiffs can show substantial participation, it is immaterial ``which sovereign holds title or has the responsibility for operation of the project.'" (Arreola, supra, 99 Cal.App.4th at p. 761, 122 Cal. Rptr. 2d 38.)

"Approval and acceptance by the public agency may be implied by official acts of dominion or control of the property and by continued use of the improvement by that agency for many years." (2 Condemnation Practice in Cal. (Cont.Ed.Bar 2d ed. 2003) General Background, § 13.3, p. 657.) "Streets, utilities, and drainage systems, when accepted and approved by a municipality, become a public improvement and part of its system of public works." (Id. at § 13.7, p. 670; cf. DiMartino v. City of Orinda (2000) 80 Cal. App. 4th 329, 337-340, 95 Cal. Rptr. 2d 16 [no evidence entity approved of or accepted drainpipe under private property].)

In this case the State plans (the Jackson Report, modified by the Grant Report, resulting in the SRFCP) call for the State "to exercise control [and] to incorporate the [levee] into a unified public [flood control] system." (Locklin, supra, 7 Cal.4th at p. 370, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) The State benefited by the cost savings of accepting the Morrison Grade and improving it slightly, rather than rerouting the levee or correcting its structural flaws. The State should not be relieved of liability, if otherwise applicable, because of the fortuity that Yuba County built the levee. In short, the State accepted the levee as a State levee.

The State argues that the acceptance doctrine does not apply in this case, asserting the doctrine applies in and only in strict liability cases. But other cases applying Belair's reasonableness standard have used the doctrine. One claim in Locklin itself was that a creek had become a work of public improvement, but Locklin held the evidence did not show "intent to exercise control or to incorporate the creek into a unified public drainage system." (Locklin, supra, 7 Cal.4th at p. 370, 27 Cal. Rptr. 2d 613, 867 P.2d 724; see also id. at p. 338, 27 Cal. Rptr. 2d 613, 867 P.2d 724 ["if it has incorporated the watercourse into a public drainage system or otherwise converted the watercourse itself into a public work"].) Locklin then recited a traditional formulation of the acceptance doctrine: "A governmental entity must exert control over and assume responsibility for maintenance of the watercourse if it is to be liable for damage caused by the streamflow on a theory that the watercourse has become a public work." (Id. at p. 370, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) None of this discussion would have been necessary if the doctrine applies in and only in strict liability cases. Another case not involving strict liability concluded "a public entity is a proper defendant in a claim for inverse condemnation if it has the power to control or direct the aspect of the public improvement that is alleged to have caused the injury." (Arreola, supra, 99 Cal.App.4th at pp. 762-763, 122 Cal. Rptr. 2d 38; cf. also Bunch, supra, 15 Cal.4th at p. 437, 63 Cal. Rptr. 2d 89, 935 P.2d 796 [private developer built flood control system taken over by public entity, no suggestion this made any difference; finding of no inverse liability based on other grounds].) Based on these authorities we reject the attempt to cabin the doctrine to strict liability cases.

The State had the "power to control or direct the aspect of the public improvement that is alleged to have caused the injury" (Arreola, supra, 99 Cal.App.4th at pp. 762-763, 122 Cal. Rptr. 2d 38), and "incorporated the [levee] into a public [flood control] system." (Locklin, supra, 7 Cal.4th at p. 338, 27 Cal. Rptr. 2d 613, 867 P.2d 724.) No more need be shown.

In the trial court the State argued that adoption of the Jackson Report "shifted no risks of damage," to Paterno's predecessors because it was Yuba County which "caused Linda levee to have the siting, alignment, foundation and core composition that it had at the time of the 1986 flood.... [T]hose risks were shifted to private property in 1904 before the Jackson Report, Grant Report and their adoption by the State of California." But if, as we have said, we determine disproportionate burdens by determining "``whether the system, as designed, constructed, operated, and maintained,'" exposed an owner to an unreasonable risk of harm, the inquiry cannot be frozen as of the date of construction, which necessarily predates operation and maintenance. (Akins, supra, 61 Cal.App.4th at p. 28, fn. 20, 71 Cal. Rptr. 2d 314.) This negates the State's view that the only relevant shift of risk took place in 1904. The many years the State operated the levee is also relevant.

As we explain below, the District did not control the levee in such a way that it could cure its defects.

IV. Liability for Failure to Upgrade

Although parts I through III of this opinion demonstrate that an unreasonable State plan caused Paterno's damages and that he is entitled to recover therefor, it is important to clarify what we meant in Paterno I by an impermissible "upgrade" liability theory, to avoid confusion in future cases.

The SRFCP plan called for the levee to carry about 120,000 cubic feet per second (c.f.s) of water. Paterno does not now argue it should have been able to carry more water, or for a longer time. Nor does he claim that simply because the levee broke while carrying about 63,000 c.f.s. that he is entitled to recover. He complains the levee did not carry the planned water because of poor construction and alignment, causing severe seepage problems leading to its collapse.

In Paterno I, we rejected Paterno's claim that liability could be predicated on the failure to increase the flow capacity of the levee (Paterno I, supra, 74 Cal.App.4th at pp. 96-97, 87 Cal. Rptr. 2d 754):

"The trial court ... found the levee should have been upgraded in light of increased urbanization below the levee.... [T]he government need not provide any level of flood protection.... It would be an unwarranted usurpation of power for a judge to impose liability for failure to upgrade a project, rather than for a defect in the project planned by the executive and legislative branches.

"... [¶] ... [¶]

"The Attorney General properly observes liability based on a failure to upgrade ``places the determination of a project's scope in the hands of those who have caused the protected area to be more extensively used, rather than in the hands of the public entities and elected officials charged with that determination.' He also observes flood concerns invoke competing public interests ... and continues, ... ``Whether [resolution of such concerns] should be achieved by building more dams, bigger levees, restricting development in high risk areas, or some other means, however, has not been assumed by the courts as within their province to decide.' We agree. Judges do not decide where to build dams and levees, nor how high.

"Paterno's argument also ignores the passage of Van Alstyne, approved in Bunch, to the effect the reasonableness calculus must be made as of the time the public entity is making the decision, for example, to erect an 80-foot levee, instead of a 90-foot levee."

Thus, in Paterno I, we emphasized that the State was not an insurer against flood risks, and rejected a claim of liability based on the idea that the State has to increase flood protection simply because the value of property to be protected has increased. (See also Bunch, supra, 15 Cal.4th at p. 454, 63 Cal. Rptr. 2d 89, 935 P.2d 796.) Imposing liability for the failure to redesign levees and dams to provide greater levels of protection would in effect allow the courts to usurp executive functions and would ultimately deter the construction of flood control projects. We simply applied the evolving rules of inverse condemnation to the claim raised by Paterno in that case, viz., that "the levee should have been upgraded in light of increased urbanization below the levee." (Paterno I, supra, 74 Cal.App.4th at p. 96, 87 Cal. Rptr. 2d 754.)

However, the trial court derived a different rule from Paterno I. After finding that the 1934 and 1940 projects met their limited objectives, the trial court stated "the provision of seepage control features and the reconstruction of the levee segments or their realignments would constitute upgrades in the condition of the levee which the State was not required to provide and which the court may not consider as the basis for imposition of inverse condemnation liability." As the District put it, the trial court found defendants had no duty to "beef up existing Linda levee structures by installing seepage controls or reconstructing or relocating the levee."

Taken to its end, this would mean that once a public work was built, no inverse liability could be predicated on a claim that it was poorly designed or built, and any curative measure would be an upgrade. That would contravene precedent. (Belair, supra, 47 Cal.3d at p. 565, 253 Cal. Rptr. 693, 764 P.2d 1070 [where "design, construction or maintenance of a flood control project is shown to have posed an unreasonable risk of harm to the plaintiffs, and such unreasonable design, construction or maintenance constituted a substantial cause of the damages, plaintiffs may recover"].)

In Paterno I, we said, "the reasonableness calculus must be made as of the time the public entity is making the decision," (74 Cal.App.4th at p. 97, 87 Cal. Rptr. 2d 754, citing Van Alstyne and Bunch) but we were speaking of a specific hypothetical decision ("for example, to erect an 80-foot levee, instead of a 90-foot levee"), and we did not hold that the reasonableness calculus is frozen at the time an official plan is adopted, as the State implies. We agree with Paterno that the State reads this passage out of context. We did not mean an entity can ignore evidence the improvement does not actually meet design standards and poses a risk of failure, then seek refuge in the defense that any cures after the date of construction would be upgrades. The State's view has been rejected: "Counties also contend that the reasonableness calculus must be made as of the time the public entity is making the decision to approve the project, and that the trial court incorrectly focused on conduct that took place after adoption of the federal maintenance regulations. This contention confuses the purpose of the balancing analysis. The balancing analysis required by Locklin applies to the public entities' action that results in the injury. In Belair, supra, 47 Cal. 3d 550, 253 Cal. Rptr. 693, 764 P.2d 1070, it was the design of the levee system that resulted in the injury so that the reasonableness of the design would have been the proper consideration. Here, the trial court applied the analysis to the Counties' long-standing policy of allowing the Project channel to deteriorate.... [I]t was that long-standing policy that caused the damage." (Arreola, supra, 99 Cal.App.4th at p. 741, 122 Cal. Rptr. 2d 38.)

We largely agree with Paterno that "Paterno I made clear that increasing the level of flood protection of a project is what constitutes an upgrade. [Citation.] Nothing in Paterno I — or any other authority — suggests that measures required so that a project provides the planned level of protection are somehow an upgrade. Work that restores a levee's design level of protection is maintenance, not an upgrade."

Our conclusion does not punish the State for failing to upgrade the project, nor does it interfere with the State's executive prerogative to choose where and how to build levees. It simply implements the California Constitution's command that the State must pay for damaging property, as refined by California Supreme Court precedent applicable to flood control cases, to the effect that it is unfair to saddle Paterno with a disproportionate share of the damages caused directly by the SRFCP when the SRFCP plans deferred the costs of curing the defects not called for by the designers.

V. Apportionment of Responsibility

As we have explained, liability is based on the State's plan which incorporated the Linda levee into what is now the SRFCP. Paterno argues the District is also liable. Not so.

"A plaintiff in inverse condemnation must establish the proportion of damage attributable to the public entity from which recovery is sought." (Jordan v. City of Santa Barbara (1996) 46 Cal. App. 4th 1245, 1274, 54 Cal. Rptr. 2d 340, citing Locklin, supra, at p. 372, 27 Cal. Rptr. 2d 613, 867 P.2d 724; see Mehl v. People ex rel. Dept. Pub. Wks. (1975) 13 Cal. 3d 710, 718, 119 Cal. Rptr. 625, 532 P.2d 489.) In theory, two public entities might be equally liable if they were "in joint charge of the public works." (Akins, supra, 61 Cal.App.4th at p. 48, fn. 41, 71 Cal. Rptr. 2d 314.)

The District does routine maintenance for the State. (See Clement, supra, 35 Cal.2d at p. 645, 220 P.2d 897; Riley, supra, 192 Cal. at p. 150, 218 P. 762.) It collects assessments from local landowners to control weeds and rodents, and patrol for boils during high water. (See Paterno I, supra, 74 Cal.App.4th at p. 104, 87 Cal. Rptr. 2d 754.) It has no authority to reconstruct the levee, even if it had the resources to do so. (See Van Alstyne, supra, 20 Hastings L.J. at p. 494, fn. 288 [discussing financial problems of small entities and difficulty of insuring against inverse liability].) As the District puts it, Paterno "failed to establish that the siting, design, or construction of the levee was attributable to any plan adopted by RD 784 [.]" (Italics added.)

We reject Paterno's claim that the State's relationship with the District mandates a joint liability finding. Such liability extends to acts arising "in the performance of" an agreement between public entities. (Gov.Code, § 895.2.) Nothing in the State's relationship with the District gave the District the ability to change the levee, and the liability we find did not occur during the performance of an agreement inter sese. (See 1 Van Alstyne, Cal. Gov. Tort Liability (Cont.Ed.Bar 4th ed. 2003) Defenses and Indemnification, § 4.32, p. 140.)

DISPOSITION

The judgment in favor of the District is affirmed. Paterno shall pay the District's costs on appeal. (Cal. Rules of Court, rule 27(a).) The District also will be entitled to its costs of suit from Paterno. (Locklin, supra, 7 Cal.4th at pp. 375-377, 27 Cal. Rptr. 2d 613, 867 P.2d 724.)

The judgment in favor of the State is reversed and the cause is remanded with directions to enter judgment for Paterno and conduct such further proceedings as are necessary to determine the damages of nonsample plaintiffs. The State shall pay Paterno's costs on appeal. (Cal. Rules of Court, rule 27(a).) Paterno also will be entitled to his costs of suit from the State, including "reasonable attorney, appraisal, and engineering fees" actually incurred. (Code Civ. Proc., § 1036.)

The trial court is directed to give this case priority over all civil cases except as statutes otherwise require, and to take all feasible steps to expedite this case.

We concur: SIMS, Acting P.J., and NICHOLSON, J.