Rose Acre Farms, Inc. ("Rose Acre") filed the present action in the United States Court of Federal Claims in 1992, claiming that United States Department of Agriculture ("USDA") regulations that restricted egg sales from and imposed other requirements on farms that tested positive for the presence of salmonella bacteria effected a taking requiring compensation under the Fifth Amendment. The trial court held that Rose Acre was entitled to compensation for a taking of the eggs affected by the regulations, Rose Acre Farms, Inc. v. United States, 55 Fed. Cl. 643, 660 (2003), as well as for hens seized for testing. Id. at 662. The court misapplied, however, the standards governing regulatory takings claims under Penn Central Transportation Co. v. New York City, 438 U.S. 104, 98 S. Ct. 2646, 57 L. Ed. 2d 631 (1978). In particular, the court incorrectly analyzed the severity of the economic impact of the regulations and erroneously concluded that the Penn Central factor pertaining to the character of the government's actions favored Rose Acre. The court further erred in concluding that the regulations effected a per se taking of Rose Acre's hens. Accordingly, we vacate and remand for appropriate reconsideration.

Rose Acre is a family-owned business based in Seymour, Indiana. It is primarily engaged in the production of table eggs, which are raw poultry eggs sold in their shells. Between 1955 and 1990, Rose Acre grew from a single layer-hen farm with 1,800 hens to a highly integrated table-egg production business consisting of eight layer-hen farms with millions of hens. Three of Rose Acre's Indiana farms are at issue in this case, namely, Cort Acres (in Cortland), White Acres (in White County), and Jen Acres (in Jennings County).

The production units on each farm are individual layer houses having varying capacities. In 1990, Cort Acres had thirty-six layer houses, each of which contained approximately 70,000 hens, White Acres had twelve layer houses, each containing approximately 125,000 hens, and Jen Acres had twenty-two houses, twenty-one of which were in production with capacities ranging from 67,320 to 112,000 hens.

The details of Rose Acre's vertically integrated production system are set forth in the trial court's opinion. Rose Acre, 55 Fed.Cl. at 647. We note here, though, that all of the layer hens in a given layer house at any one time are, as a result of Rose Acre's production system, approximately the same age. Once young hens capable of laying eggs are placed in a layer house, production in that house normally continues uninterrupted for a period of about fifty-seven to sixty weeks, until the hens therein reach the end of their productive lives. When that cycle has ended, the hens are removed and destroyed, and the house is cleaned before new hens are introduced.

To maximize its production and provide a consistent supply of table eggs to the market, Rose Acre must carefully manage its layer house population and depopulation schedules. The trial court found that "[s]cheduling and timing ... are key components of [Rose Acre's] business. An interruption in [Rose Acre's] scheduling system affects the entire organization, thus causing [Rose Acre] to be unable to supply eggs to its customers." Id.

In the late 1980s, the Centers for Disease Control ("CDC") determined that the incidence and geographic spread of human illness resulting from exposure to Salmonella enteritidis serotype enteritidis ("SE") bacteria was increasing.1 In response to the increase, the Animal Plant Health and Inspection Service ("APHIS"), a USDA division responsible for preventing the spread of communicable diseases, determined that emergency regulations were necessary to control the spread of SE in poultry flocks. On February 16, 1990, USDA published interim regulations that restricted the interstate sale and transportation of eggs and poultry from flocks determined under the regulations to be SE-contaminated. Poultry Affected by Salmonella Enteritidis, 55 Fed.Reg. 5576 et seq. (1990) (codified at 9 C.F.R. §§ 82.30-82.36 (1991)). The interim regulations were effective immediately upon publication, USDA having "determined that there is good cause for publishing this rule without prior opportunity for public comment," namely, the need for "[i]mmediate action ... to prevent harm to the egg-type chicken industry and the public." Id. at 5580.

The interim regulations applied to "flocks," defined as "[a]ll the poultry on one premises," 9 C.F.R. § 82.30 (1991), and operated as follows. If "a Federal or State representative determine[d] through epidemiologic investigation that [a] flock [was] the probable source of disease in an outbreak of [SE-caused] disease in humans or poultry," USDA designated the flock as a "study flock." Id. § 82.32. A study flock was subsequently designated a "test flock" if either (1) "one or more" environmental test samples, i.e., "manure samples and egg transport machinery samples ... collected and tested in accordance with" procedures set forth in the interim regulations tested positive for SE, or (2) "the person in control of the flock" refused to allow or interfered with the collection of such samples. Id. § 82.32(b). At the time the interim regulations were published, USDA believed that evidence of SE in layer hens' environment meant that the hens were infected and would, therefore, be more likely to produce SE-contaminated eggs. See 55 Fed.Reg. at 5576 (describing the "vertical" (hen to egg) and "horizontal" (environment to hen) modes of SE transmission).

"Test flock" status triggered restrictions on the interstate movement of eggs. Specifically, eggs from a test flock could be moved interstate only for uses requiring pasteurization,2 and then only if the shipper obtained a permit and met other conditions. 9 C.F.R. § 82.33(a) (1991). Thus, the interim regulations prohibited the interstate shipment of test flock eggs for sale as table eggs.

Specified numbers of the hens in test flocks were also required to undergo blood and internal-organ testing. Id. § 82.32(c). A test flock was designated an "infected flock" if the organs of one or more hens tested positive for SE. Id. Infected flocks were subject to the same interstate transportation restrictions as test flocks. Id. § 82.33(a). An infected flock retained its "infected" designation until either (1) the flock was retested in accordance with the regulations and no internal organ tested positive for SE or (2) the houses that contained the infected flock were depopulated, subjected to specified wet cleaning and disinfecting procedures, and repopulated with a new flock. Id. § 82.32(c).

After USDA reviewed comments received from interested parties following the publication of the interim regulations, it published final SE regulations on January 30, 1991. Chickens Affected by Salmonella enteritidis, 56 Fed.Reg. 3730 (1991) (codified at 9 C.F.R. §§ 82.30-82.38 (1992)). The final regulations incorporated all of the above requirements, but authorized the imposition of restrictions on individual layer houses as opposed to whole flocks. 9 C.F.R. § 82.33(a). A provision conditioning release from "infected" status on a successful post-cleaning inspection of a depopulated infected house by a federal or state official was added. Id. § 82.37. Additional testing and retesting requirements were imposed on all houses on the same premises as any infected house. Id. § 82.38.

APHIS administered these SE regulations until mid-1995. A total of thirty-eight flocks were restricted between 1990 and 1994, resulting in over 1.3 billion eggs being diverted from the United States table egg market to breaker plants.

In 1990, after the interim regulations took effect, SE illness outbreaks were traced to each of Cort Acres, White Acres and Jen Acres. As a result of testing carried out in accordance with the interim regulations, USDA first restricted the interstate transportation of eggs from these three farms on October 5, 1990, November 27, 1990, and January 15, 1991, respectively. In each case, Indiana officials similarly restricted the intrastate transportation of eggs (except for uses requiring pasteurization) shortly after receiving notice of the federal restrictions.

After "test flock" restrictions were imposed as a result of environmental testing at each affected Rose Acre farm, USDA conducted blood and organ testing as set forth in the regulations. For organ testing, USDA employees physically removed 60 hens (whose blood had tested positive) from each house, killed them, and transported their carcasses to a USDA laboratory in Ames, Iowa. As described above, a single positive organ result in a given house resulted in an "infected house" designation.3 No additional transportation restrictions were imposed as a result of an "infected" designation; obtaining release from restricted status, however, became more difficult. At first, Rose Acre tried to obtain release through continued organ testing of the hens in infected houses. For the most part, however, Rose Acre had to depopulate, clean, and disinfect infected houses, and then have those houses pass USDA inspection. The trial court noted that in some cases, houses were empty for long periods while awaiting inspection. Rose Acre, 55 Fed.Cl. at 651. It also noted that USDA inspection officials did no more than visually examine the interior of depopulated houses (after cleaning) with the aid of flashlights. Id.

Rose Acre finally succeeded in obtaining release from the restrictions imposed on Cort Acres, White Acres, and Jen Acres on July 16, 1992, May 8, 1992, and October 30, 1992, respectively. Thus, for a period of twenty-five months, Rose Acre was unable to sell eggs as table eggs from one or more of the three farms.

Shortly after its operations became subject to the federal and state restrictions, Rose Acre filed an action in the United States District Court for the Southern District of Indiana seeking a declaration that the interim regulations were invalid. In that action, Rose Acre contended that (1) the interim and final regulations deprived Rose Acre of due process, (2) the interim regulations were not promulgated in accordance with the Administrative Procedure Act, (3) both sets of regulations exceeded USDA's statutory authority, (4) the final regulations could not be applied retroactively, (5) both sets of regulations unlawfully delegated authority to state officials, (6) the application of certain monitoring provisions was invalid, and (7) it was entitled to compensation for eggs diverted to breaker plants. Rose Acre Farms, Inc. v. Madigan, No. NA 90-175-C, 1991 U.S. Dist. LEXIS 8691, at *3-4 (S.D. Ind. June 5, 1991). Ultimately, the United States Court of Appeals for the Seventh Circuit held that the regulations were neither arbitrary nor capricious and were promulgated within the authority of the Secretary of Agriculture. Rose Acre Farms, Inc. v. Madigan, 956 F.2d 670, 675-77 (7th Cir.1992) ("Rose Acre I"). It further held that "[i]t is to the claims court that Rose Acre must go" to pursue any claim for compensation. Id. at 674.

Rose Acre filed the present action in the Court of Federal Claims on October 13, 1992, alleging violations of 21 U.S.C. §§ 114a4 and 134a5 (2000) and an uncompensated taking of its eggs and hens. The trial court granted the government's motion to dismiss Rose Acre's section 114a claim for failure to state a claim, Rose Acre, 55 Fed.Cl. at 653, and held, after a two-week trial, that section 134a provides Rose Acre no relief beyond that available under the Fifth Amendment, id. at 662. The trial court awarded Rose Acre compensation in the amount of $6,165,297.72 for what it held was a regulatory taking of its eggs diverted to breaker plants and a categorical taking of the hens confiscated for internal-organ testing. Id. at 664. The court also awarded Rose Acre $2,414,744.81 in attorney fees and expenses. Id. at 670.

The government appeals, challenging the trial court's holding that the government actions at issue here constituted a regulatory taking and a categorical taking and the award of fees and expenses (as based on an erroneous judgment that takings occurred). We have jurisdiction pursuant to 28 U.S.C. § 1295(a)(3).

A determination of whether a taking compensable under the Fifth Amendment has occurred is a question of law based on factual underpinnings. Alves v. United States, 133 F.3d 1454, 1456 (Fed.Cir.1998). Thus, we review the trial court's legal analysis and conclusions de novo and its findings of fact for clear error. Bass Enters. Prod. Co. v. United States, 133 F.3d 893, 895 (Fed.Cir.1998).

The trial court analyzed Rose Acres claim for compensation for its diverted eggs under regulatory takings law and applied the law of categorical takings to the claim for compensation for the hens.

Rose Acre acknowledged and the trial court recognized that the government may regulate private property to some extent without subjecting itself to takings liability. Rose Acre, 55 Fed. Cl. at 656. Otherwise, "[g]overnment could hardly go on...." Penn. Coal Co. v. Mahon, 260 U.S. 393, 413, 43 S. Ct. 158, 67 L. Ed. 322 (1922). However, "if regulation goes too far it will be recognized as a taking." Id. at 415, 43 S. Ct. 158.

The challenge, of course, is determining how far is "too far." A regulatory takings claim "`arises from some public program adjusting the benefits and burdens of economic life to promote the common good,'" as opposed to a government appropriation of private property for its own use. Tahoe-Sierra Pres. Council, Inc. v. Tahoe Reg'l Planning Agency, 535 U.S. 302, 324-35, 122 S. Ct. 1465, 152 L. Ed. 2d 517 (2002) (quoting Penn Central, 438 U.S. at 124, 98 S. Ct. 2646). When a takings claim arises from the former, it is necessary to determine whether justice and fairness require that economic injuries caused by public action be compensated by the government, rather than remain disproportionately concentrated on a few persons. Penn Central, 438 U.S. at 124-25, 98 S. Ct. 2646. No per se rules or "set formula[s]" govern such determinations; instead, courts "engag[e] in ... essentially ad hoc, factual inquiries." Id. The Supreme Court has, however, identified several factors having particular significance, namely, the "economic impact of the regulation on the claimant," "the extent to which the regulation has interfered with distinct investment-backed expectations," and "the character of the governmental action." Id. Application of these Penn Central criteria is required where, as here, "less than a `complete elimination of value'" resulted from the regulation at issue. Tahoe-Sierra, 535 U.S. at 330, 122 S. Ct. 1465 (quoting Lucas v. S.C. Coastal Council, 505 U.S. 1003, 1019-20 n. 8, 112 S. Ct. 2886, 120 L. Ed. 2d 798 (2002)).

The trial court held that each of these factors favored Rose Acre, and, as a result, Rose Acre was entitled to compensation for the revenue it lost as a consequence of the diversion of its eggs to the breaker egg market. As to two of the three factors, however, the trial courts analysis was incorrect. Remand for reconsideration of the evidence under the applicable legal standards, and for a weighing of those factors in accordance with the standards set forth in Penn Central and its progeny to determine whether compensation is required under the Fifth Amendment, is, therefore, necessary.

The government did not dispute that Rose Acre sold over 97% of the eggs it produced at the three farms as table eggs before its sales were restricted. It also does not challenge the trial court's findings that (1) after the restrictions were lifted, Rose Acre immediately returned to selling over 97% of its eggs as table eggs, (2) both before and during the period of restriction, Rose Acre sold 90% of its eggs in interstate commerce and 10% in intrastate commerce, (3) Rose Acre diverted over 57.5 million dozen eggs (nearly 700 million eggs) to breaker plants during the restricted period, (4) the average price Rose Acre and other sellers received for table eggs during the restricted period was 59 cents per dozen, (5) the average total cost for Rose Acre to produce a dozen eggs during the period of restriction was 54.96 cents, (6) during the restricted period, Rose Acre processed 33,753,843 dozen restricted eggs in breaker plants it owned and sold 24,006,780 dozen restricted eggs to breaker plants owned by others, and (7) Rose Acre received an average of 46.64 cents per dozen for eggs processed in its own breaker plants and an average of 41.46 cents per dozen for eggs sold to outside breaker plants during the restricted period. The government does, however, challenge the trial court's finding that "[t]he economic impact of the diversion was ... severe."

The trial court's analysis of the economic impact was limited to (1) making the above-noted findings, (2) noting that government witnesses testified that "the restrictions could mean financial ruin for table egg producers," "`[b]reaker eggs are not rewarding,'" and "producers faced considerable revenue losses from the required diversion of eggs," and that the government's expert estimated Rose Acre's loss on breaker eggs to exceed $9.2 million, and (3) favorably comparing Rose Acre's situation to that of the plaintiff in Yancey v. United States, 915 F.2d 1534 (Fed.Cir.1990), where we held that a USDA-imposed quarantine had effected a taking of healthy breeder turkeys. Rose Acre, 55 Fed.Cl. at 657-58. This analysis was insufficient.

The trial court noted that a large number (57.5 million dozen) of Rose Acre's eggs were, as a result of the restrictions, diverted for sale at less than Rose Acre's average total cost of production. However, neither the testimony nor the economic data cited by the trial court appropriately gauge the severity of the economic impact of the regulations on Rose Acre. The cited testimony is not specific to Rose Acre, and the data — divorced from any economic context — represents only the first step in the required analysis. Simply put, it is not possible to determine the economic impact of a regulatory scheme applied to a private actor without casting the appropriate absolute measures of the effect of the regulation against the backdrop of relevant indicators of the economic vitality of the actor. In the present case, some of the eggs on some of Rose Acre's farms suffered a reduction in value. In addition, while the reduction in value of each affected egg was permanent, the period during which the regulations had an impact on Rose Acre's operations was relatively brief — approximately two years — after which Rose Acre reverted to its pre-regulation table-egg sales levels. The trial court's opinion reflects no consideration of these factors. Yet, an assessment of the severity of the economic impact of the regulations on Rose Acre, in accordance with Penn Central and its progeny, must take them into account.

In regulatory takings cases concerning diminished real estate values, this concept is known as the "parcel as a whole." Penn Central, 438 U.S. at 130-31, 98 S. Ct. 2646. The Supreme Court has noted:

"Taking" jurisprudence does not divide a single parcel into discrete segments and attempt to determine whether rights in a particular segment have been entirely abrogated. In deciding whether a particular governmental action has effected a taking, this Court focuses rather both on the character of the action and on the nature and extent of the interference with rights in the parcel as a whole — here, the city tax block designated as the "landmark site."

Id. The Court recently elaborated on this concept, stating:

This requirement that "the aggregate must be viewed in its entirety" explains why, for example, a regulation that prohibited commercial transactions in eagle feathers, but did not bar other uses or impose any physical invasion or restraint upon them, was not a taking. Andrus v. Allard, 444 U.S. 51, 66, 100 S. Ct. 318, 62 L. Ed. 2d 210 (1979). It also clarifies why restrictions on the use of only limited portions of the parcel, such as setback ordinances, Gorieb v. Fox, 274 U.S. 603, 47 S. Ct. 675, 71 L. Ed. 1228 (1927), or a requirement that coal pillars be left in place to prevent mine subsidence, Keystone Bituminous Coal Assn. v. DeBenedictis, 480 U.S. [470,] 498, 107 S. Ct. 1232, 94 L. Ed. 2d 472 [(1987)], were not considered regulatory takings.

Tahoe-Sierra, 535 U.S. at 327, 122 S. Ct. 1465.

The methodologies used to determine how much of an affected owners property constitutes the relevant "parcel" in real estate takings cases are not, of course, directly applicable to a case such as the present. Nonetheless, a determination of the relative economic effect of the regulations is required. See Penn Central, 438 U.S. at 130, 98 S. Ct. 2646 ("[T]he submission that [the claimants] may establish a taking simply by showing that they have been denied the ability to exploit a property interest that they heretofore had believed was available is quite simply untenable.").

Rose Acre recognizes this, and thus defends the trial courts finding that the economic impact was severe by arguing that the relevant "parcel," or "denominator in the takings fraction," Palazzolo v. Rhode Island, 533 U.S. 606, 631, 121 S. Ct. 2448, 150 L. Ed. 2d 592 (2001), in this case was each individual restricted egg, or, since all of the eggs in a given affected house were restricted, each affected house. Based on the difference between the revenue that would have been obtained for those eggs but for the restrictions (found by the district court to be 59 cents per dozen) versus that actually obtained for those eggs in the breaker market during the period of restriction (which the district court found to be 46.64 and 41.46 cents per dozen for the eggs processed in Rose Acres own breaker plants and those sold to outside breaker plants, respectively), Rose Acre argues that the relevant diminution in value was 26.1%.

The government, in contrast, argues that the relevant parcel is the total combined production of the three affected farms. It then argues that the 59 cent-per-dozen price found by the district court to be the prevailing price of table eggs during the period of restriction cannot properly be applied to either (1) the total egg production of the three affected farms during the period of restriction (if the governments denominator is used), or (2) the egg production of the affected houses (if Rose Acres "denominator" is used), because Rose Acre argued (and the district court found) that even before the regulations took effect nearly three percent of Rose Acres eggs were sent to breaker plants. The government bases its economic analysis, instead, on a table-egg price of 57.62 cents per dozen, which takes into account Rose Acre's three percent "baseline" breaker egg rate. Given that 42.6 percent of the eggs produced by the three affected farms were diverted to breaker plants during the period of restriction, it argues that the appropriate diminution-in-value figure for purposes of evaluating the severity of the economic effect on Rose Acre is 10.6. Moreover, argues the government, neither the 10.6 diminution it asserts nor the 26.1 diminution asserted by Rose Acre comes close to the 77.2 diminution in value suffered by the plaintiffs in Yancey, the case the trial court regarded as analogous.

In Yancey, USDA imposed an emergency quarantine on poultry in an effort to contain an outbreak of pathogenic Avian Influenza, a highly contagious viral disease. 915 F.2d at 1536. The Yanceys maintained a flock of turkey breeder hens and toms for purposes of selling turkey hatching eggs in interstate commerce. Id. Despite the fact that testing showed the turkeys were not diseased, the Yanceys decided to sell their turkey flock for meat after the quarantine was imposed, because of the quarantine's indefinite nature. Id. This court affirmed a decision that the Yanceys were entitled to compensation under the Fifth Amendment, id. at 1542, based on the 77% reduction (from $91,616 to $20,887, the amount received from the slaughter) in the value of their turkey breeder flock, id. at 1539.

We recognized in Yancey that comparable or even larger diminutions in value had been held insufficient for takings purposes in particular cases. See id. at 1541 (citing Euclid v. Ambler Realty Co., 272 U.S. 365, 47 S. Ct. 114, 71 L. Ed. 303 (1926) (holding that a 75% value diminution caused by a zoning law did not constitute a taking), and Hadacheck v. Sebastian, 239 U.S. 394, 36 S. Ct. 143, 60 L. Ed. 348 (1915) (holding that no taking resulted from an 87½% diminution in value)). We rejected the notion that such cases set a minimum value diminution that, as a matter of law, had to be demonstrated for liability to lie. Id. In so doing, we recognized that "the modern Penn Central approach" requires a balancing of "all the relevant considerations." Id.

Nonetheless, in apparent recognition of the significant disparity between the diminution in value in this case, even under Rose Acre's view of the appropriate denominator, and that found in Yancey, Rose Acre argues in the alternative that the trial court's finding as to economic effect should be sustained based upon an analysis of the loss it says it suffered on sales of the restricted eggs. Under this alternative diminution-in-return (profit-based) approach, using the same denominator and table-egg and breaker-egg prices on which it relies in its above-described diminution-in-value calculation, Rose Acre argues that it earned a return of -15.14% on breaker eggs processed by its own breaker plants and a return of -24.56% on eggs sold to outside breaker plants, compared with the +7.35% return it would have received for those eggs had they been sold as table eggs. These results are based on Rose Acre's average total cost (fixed plus marginal) of producing an egg destined for the table egg market across all of its farms, which total cost, apparently undisputed by the government, is 54.96 cents per dozen.6

Rose Acre notes that resulting diminution in return (306% and 434%, respectively, for breaker eggs it processed and those sold to outside breaker plants, respectively), is much larger than the 96% diminution we held sufficient to support a takings holding in Cienega Gardens v. United States, 331 F.3d 1319 (Fed.Cir.2003). Perhaps anticipating the government's argument that a returns-based approach is better suited to situations involving property, such as real property, that has the ability to generate income over a long term, Rose Acre notes that the Claims Court in Yancey cited the Yanceys' investment loss in support of its conclusion that a compensable taking had occurred:

Even considering the receipt of $21,496 from the Commonwealth of Virginia, plaintiffs incurred a substantial loss on their investment. Although plaintiffs were able to mitigate their loss by slaughtering the flock, there was no other alternative, economically viable use for the flock while the quarantine was in effect.

Yancey v. United States, No. 413-85C, slip op. at 17 (Cl.Ct. Dec.12, 1988).

Again, however, whether the alternative (breaker) use here was "economically viable" depends on the variables factored into the analysis. The government, in addition to noting that the 59-cents-per-dozen table-egg price fails to account for the three percent of eggs Rose Acre usually sends to breaker plants and advocating use of a three-affected-farms "parcel," disputes the propriety of relying on total cost. It argues, instead, that only Rose Acre's marginal cost — the cost of producing each egg over and above Rose Acre's fixed, operation-wide costs — is the appropriate basis for determining the return Rose Acre earned on the restricted eggs. The government's values, it notes, result in a 64.44% return for those eggs versus the 83.91% return Rose Acre would otherwise have enjoyed (i.e., on sales as table eggs). Even using Rose Acre's "parcel," application of the marginal cost basis yields a profit of 42.03% — a lower return than that achieved for the three affected farms as a whole, but a positive one, nonetheless — and a diminution in return of 52%. Still another potential measure of the impact of the restrictions on Rose Acre is obtained using the 57.72 cent-per-dozen table-egg price (to take into account the three percent of eggs Rose Acre usually sent to breaker plants), the government's "denominator" (the three affected farms), and Rose Acre's total cost figure. These figures result in a -6.3% return on the breaker eggs versus a +4.8% return Rose Acre would have enjoyed but for the restrictions. The difference is a 231% diminution in return.

All of which is to note that there are a number of different ways to measure the severity of the impact of the restrictions on Rose Acre. As noted above, the trial court's opinion does not reflect consideration of these various alternatives, or explicitly rest its conclusion that the impact was severe on any appraisal of the effect of the restrictions relative to Rose Acre's relevant unaffected property interests.

The trial court's discussion of the economic impact factor is not entirely unrevealing, however. It suggests that the trial court may have applied a profits-based analysis, as it quotes our reference in Yancey to the observation of the trial court there that "`[a]lthough plaintiffs were able to mitigate their loss by slaughtering the flock, there was no other alternative, economically viable use for the flock while the quarantine was in effect.'" Rose Acre, 55 Fed. Cl. at 658 (quoting Yancey, 915 F.2d at 1539 (quoting Yancey, slip op. at 17 (emphasis added))). By way of favorable comparison to Yancey, the trial court noted that diversion of the restricted eggs to the breaker market "was not an economically viable option for [Rose Acre] because it was not able to recoup its investment in these diverted eggs." Id. This language suggests that the trial court accepted Rose Acre's contention that its total average cost provided the appropriate basis for a returns-based calculation, for, as noted above, only when the 54.96 cent-per-dozen cost figure is used do the calculations yield negative returns: -15.14% and -24.56%, respectively, for the Rose Acre-breaker-plant- and outside-breaker-plant-processed eggs, if the restricted houses constitute the "denominator," versus a -6.3% return if the "denominator" is the three farms (taken together).

We reject the government's contention that a returns-based analysis is per se less suitable than one based on diminution in value in the present case. If anything, it appears that the latter is less appropriate where, as here, the issue concerns the economic impact, albeit temporary, of government regulations on a going business concern. We need not choose, however, between these two analytical modes. Several observations regarding their proper use, though, are required.

The government challenges Rose Acre's returns-based analysis, arguing that the relevant denominator is the combined total egg sales from the three farms during the period of restriction, but that profit should be figured using only the marginal cost to Rose Acre of producing each individual egg in the restricted houses. Rose Acre argues precisely the reverse, asking that we look only at the revenue derived from the sale of (breaker) eggs from the restricted houses, but determine its profit using its total cost, including the (allocated) fixed costs it incurs in producing eggs in all of the houses on all its farms. We believe neither is appropriate, as the inconsistency built into each inaccurately prejudices the result.

Moreover, whether the economic impact is judged by value decline or profitability decrease, an evaluation based on Rose Acre's asserted "denominator" — the restricted houses — runs afoul of the "parcel as a whole" principle announced in Penn Central and illustratively applied in Keystone Bituminous. In the latter, a Pennsylvania statute restricted (for subsidence purposes) the amount of coal that could be removed from a given mine. Keystone Bituminous, 480 U.S. at 476-77, 107 S. Ct. 1232. The Supreme Court specifically rejected the plaintiffs' contention that the restricted coal — less than two percent of the total coal in the thirteen mines they operated — was the relevant "denominator," noting:

[P]etitioners have sought to narrowly define certain segments of their property and assert that, when so defined, the Subsidence Act denies them economically viable use....

Because our test for regulatory taking requires us to compare the value that has been taken from the property with the value that remains in the property, one of the critical questions is determining how to define the unit of property "whose value is to furnish the denominator of the fraction." ...

. . .

The 27 million tons of coal [required to be left in place] do not constitute a separate segment of property for takings law purposes. Many zoning ordinances place limits on the property owner's right to make profitable use of some segments of his property. A requirement that a building occupy no more than a specified percentage of the lot on which it is located could be characterized as a taking of the vacant area as readily as the requirement that coal pillars be left in place. Similarly, under petitioners' theory one could always argue that a setback ordinance requiring that no structure be built within a certain distance from the property line constitutes a taking because the footage represents a distinct segment of property for takings law purposes.... There is no basis for treating the less than 2% of petitioners' coal as a separate parcel of property.

. . .

When the coal that must remain beneath the ground is viewed in the context of any reasonable unit of petitioners' coal mining operations and financial-backed expectations, it is plain that petitioners have not come close to satisfying their burden of proving that they have been denied the economically viable use of that property....

Id. at 495-99, 107 S. Ct. 1232. The analogies to the present set of facts are apparent. As in Keystone Bituminous (and as discussed more fully below), the regulation at issue was an exercise of the government's police powers, designed to protect the health and safety of the populace. There, as here, the regulation reduced the value of a portion of Rose Acre's property,7 even if only the three farms implicated by the 1990 tracebacks are considered.8 And, significantly, the restricted portion — the eggs here, the coal in Keystone Bituminous — comprised fungible units of the owner's stock of property.

Thus it would seem that the "parcel as a whole" rule applies with even greater force here, where no unique attribute inherent in the restricted portion provides a basis for distinguishing it from the unrestricted portion, than might be the case with real property. Rose Acre tries to overcome this fact by arguing that the regulations were applied on a house-by-house basis. It is true that the interstate and intrastate transport restrictions were applied ultimately to individual houses, but the tracebacks that resulted in the "study flock" designation at each of Cort Acres, White Acres, and Jen Acres (triggering the environmental testing that ultimately led to the identification of the restricted "test houses") were, in accordance with the interim and final regulations, directed to each farm as a whole. 9 C.F.R. § 82.32(a) (1991) (interim regulations); 9 C.F.R. § 82.32(a) (final regulations). Additionally, as long as any one house on any farm was designated as an "infected house," all other houses on that farm were required to undergo testing for purposes of monitoring. 9 C.F.R. § 82.38. Thus, as a matter of law, because the regulations at issue applied to each farm as a whole, their economic impact cannot be measured by considering the restricted houses alone. Severance of the restricted property for purposes of the economic analysis was held improper in Keystone Bituminous, where the regulation at issue did not include such premise-wide features. Here, then, such severance is even less appropriate.

On remand, then, using the three farms (combined) as the relevant "denominator," the trial court must determine whether the economic impact in this case is best measured by the value decline (a 10.6% diminution) or profitability decrease (at most, a reduction from a 4.8% profit to a 6.3% loss) caused by the restrictions. Either conclusion marks only the first step in the takings analysis, however, because, as discussed below, the court must weigh the private and public interests reflected in the application of the three Penn Central criteria to the circumstances of this case.

B. Reasonable Investment-Backed Expectations

The government assigns error to the trial court's conclusion that this factor favored Rose Acre. As the trial court noted, "[t]his factor limit[s] takings recoveries to owners who [can] demonstrate that they bought their property in reliance on a state of affairs that did not include the challenged regulatory regime." Rose Acre, 55 Fed.Cl. at 658 (quoting Loveladies Harbor, Inc. v. United States, 28 F.3d 1171, 1177 (Fed.Cir.1994)).

The trial court noted that, although the poultry industry in general is highly regulated, government experts previously believed that salmonella could contaminate the interior of a shell egg only via a crack or break in the shell. Id. at 659 (citing a government expert's testimony regarding the 1970s-era belief held by the Food and Drug Administration and the CDC that shell eggs were not associated with food-borne diseases). Accordingly, prior to 1990, eggs were subject only to inspection and restriction for evidence of potential environmental contamination. See id. (citing 21 U.S.C. §§ 1031-1056).

The government seeks to define the field of relevant regulation more broadly, citing to long-standing regulations aimed at preventing the spread of communicable diseases in birds and poultry. See 21 U.S.C. §§ 111-135 (West 1972 & Supp.1994); 9 C.F.R. Part 82. It argues that a new regulation aimed at a specific, recently-recognized disease threat was not unforeseeable in such an environment. But the SE regulations were more than an extension of comparable regulations to a new disease. They were grounded in new scientific understanding (i.e., that salmonella could be transmitted from hen to egg) and were unprecedented in their reliance on environmental and hen testing. Accordingly, even accounting for the history of regulation in the poultry and egg industries, we cannot agree that the trial court erred in concluding that this factor favors Rose Acre.

The Supreme Court explained the need for inquiry into the character of the government's action as follows: "[a] `taking' may more readily be found when the interference with property can be characterized as a physical invasion by government, ... than when interference arises from some public program adjusting the benefits and burdens of economic life to promote the common good." Penn Central, 438 U.S. at 124, 98 S. Ct. 2646 (citation omitted). The Penn Central majority cited with approval in this regard a decision holding that no compensation under the Fifth Amendment was due orchard owners ordered to cut down their ornamental cedar trees infected with cedar rust to protect neighboring apple orchards. Id. at 125-26, 98 S. Ct. 2646 (citing Miller v. Schoene, 276 U.S. 272, 48 S. Ct. 246, 72 L. Ed. 568 (1928)). This decision so held based solely on the power of the state to prevent impending harm to a valuable public resource — the apple industry. Miller, 276 U.S. at 279-80, 48 S. Ct. 246 ("[W]here the public interest is involved preferment of that interest over the property interest of the individual, to the extent even of its destruction, is one of the distinguishing characteristics of every exercise of the police power which affects property."); see also Goldblatt v. Hempstead, 369 U.S. 590, 594, 82 S. Ct. 987, 8 L. Ed. 2d 130 (1962) ("Although a comparison of values before and after" a regulatory imposition "is relevant,... it is by no means conclusive...." (citation omitted)).

As noted above, Keystone Bituminous also sprang from a police power regulation — action on the part of the Commonwealth of Pennsylvania "to protect the public interest in health, the environment, and the fiscal integrity of the area." Keystone Bituminous, 480 U.S. at 488, 107 S. Ct. 1232. The Court there noted that "the nature of the State's interest in the regulation is a critical factor in determining whether a taking has occurred." Id. It reaffirmed the principle that "no individual has a right to use his property so as to create a nuisance or otherwise harm others," and noted that "the Takings Clause did not transform that principle to one that requires compensation whenever the State asserts its power to enforce it." Id. at 492, 107 S. Ct. 1232.9

The trial court acknowledged these principles and the government's reliance thereon. Rose Acre, 55 Fed.Cl. at 659-60 (quoting Creppel v. United States, 41 F.3d 627, 631 (Fed.Cir.1994) ("If the regulation prevents what would or legally could have been a nuisance, then no taking occurred. The state merely acted to protect the public under its inherent police powers.")).10 The court recognized that "[s]almonella may be considered a nuisance," and that "the public has a strong interest in eating safe food." Id. at 660. The court held, however, that this "character" factor also favored Rose Acre because the means chosen by the government were inappropriate. Id. ("A regulation that burdens private property may `constitute a "taking" if [the burden is] not reasonably necessary to the effectuation of a substantial public purpose.'" (quoting Penn Central, 438 U.S. at 127, 98 S. Ct. 2646)). Specifically, the trial court concluded that "the SE regulations were misguided because they relied on ineffective testing methods.... [T]he regulations at issue went too far in protecting [the public] interest by prohibiting the sale of [Rose Acre's] healthy eggs as table eggs." Id.

In support of its conclusion that the regulations were "misguided," the trial court noted (1) the government "never sought to have the eggs tested, despite the fact that they were the alleged sources of the SE outbreaks," (2) after Rose Acre completed the wet cleaning procedures required in lieu of achieving consecutive negative hen tests, "the government never retested the environment or the hens," confining its inspection to "a simple walk through of the houses with a flashlight," (3) SE is ubiquitous in the environment, and is impossible to eradicate, and (4) SE-infected hens "shed" the bacteria intermittently; a hen with SE, therefore, may never lay an infected egg. Id. at 655 & n. 29. It further found that

[The government] was aware in the late 1980's that SE could exist inside an egg. In addition, the scientific technology for testing the inside of eggs existed at the time the regulations were enacted. Such testing was used in the United Kingdom and during [the government's] Pennsylvania Pilot Project. The Project discovered that 99.9725% of eggs were SE-free. It was possible, therefore, for [the government] to test [Rose Acre's] eggs for the bacterium.

Id.

It is apparent from the foregoing, and from the trial court's repeated contextual references to Rose Acre's "healthy eggs,"11 that the court's misgivings about the regulations are primarily based on its finding that a less-burdensome, alternative regulatory scheme — egg testing — was feasible. See id. at 662 ("[The government] never actually tested the eggs for SE, despite the fact that [egg] testing was feasible and would have been directed at the alleged source of the SE outbreaks"). The government vigorously disputes this proposition. It cites to the testimony of Dr. John Mason, director of the "SE Task Force" established by USDA in 1990 to respond to the SE-related public health emergency of the late 1980s, who noted that "although some people had already done work to show that SE could be found in eggs" by 1990, "eggs were not considered as a practical basis for testing," Tr. Trans. at 712, and "there hadn't been any experience with testing eggs as part of a regulatory effort." Id. at 759. Dr. Mason outlined the rationale underlying the regulations, as follows:

It was, I think, considered that first testing the environment, to find out if it was an environment, was a pretty good indication that you would find it in the birds. And if you tested in the birds and tested the tissues and it was found in the tissues, which this would [sic] imply there was a good chance it was getting into the ovaries and ovary ducts, and you could be pretty sure that at some point, it was going to get into the eggs.

Id. at 712. He described how USDA continued to investigate how best to discern the extent of SE contamination in eggs after the regulations were instituted, including testing the eggs of "three or four flocks that had been implicated in an outbreak" to determine whether USDA "could ... culture eggs in a practical way and find SE in them." Id. at 759. In fact, USDA found that "in each case in flocks where the environment was positive and the birds were positive, [it] could also isolate [SE] from the eggs." Id. As the government observes, Dr. Mason's description makes clear that USDA learned after the regulations were issued that egg testing was feasible. Rose Acre's own expert, Dr. Patricia Curtis, Director of the Poultry Products Safety and Quality Peak of Excellence Program at Auburn University, agreed:

Q Back in the 1990s, in the early 1990s, '90, '91, there wasn't a good test for eggs; is that true?

[Dr. Curtis] Yes.

Q In the early 1990s, it wasn't feasible to test a large number of eggs; is that true?

[Dr. Curtis] Yes.

Q It only became feasible to test eggs in the mid-1990s; isn't that correct?

[Dr. Curtis] Somewhere along in there. I don't know the exact date, yes.

Q Sometime after 1993, it became feasible to test eggs; is that right?

[Dr. Curtis] Yes.

Id. at 550. Finally, the other research to which the trial court cited — testing in the United Kingdom and the government's "SE Pilot Project" — took place during and after the egg testing experiments Dr. Mason described.12 Accordingly, the trial court clearly erred in finding that egg testing was feasible at the time the government imposed the restrictions at issue on Rose Acre.

Rose Acre correctly notes that the government does not contend that "some new technology was discovered between the time USDA promulgated the SE Regulations in 1990 and when it actually conducted egg testing for the first time in 1991." Therefore, it argues, the trial court's finding that "[i]t was possible... for [the government] to test [Rose Acre's] eggs for the bacterium" was not clearly erroneous. But the issue is not whether a less restrictive alternative to the government action existed or was "possible." It is whether there is a nexus between the regulation and its underlying public purpose. Nollan v. Cal. Coastal Comm'n, 483 U.S. 825, 837, 107 S. Ct. 3141, 97 L. Ed. 2d 677 (1987) (noting that if the regulation at issue "utterly fails to further the end advanced as the justification," the "purpose then becomes the obtaining of an easement to serve some valid governmental purpose, but without payment of compensation").

Similarly, the trial court's finding that Rose Acre "shared a disproportionate amount of the burden of the SE regulations," even if correct, is not relevant. That the source of the contaminated eggs could not be determined in 70% of the reported SE outbreaks during the relevant period does not undermine the rationality of investigating identifiable sources. As the court noted, Rose Acre is one of the largest egg producers in the United States, Rose Acre, 55 Fed. Cl. at 647, and the three farms determined to have been sources of SE-illness related contamination13 produced, according to Rose Acre's own laying rate and hen capacity figures, millions of eggs per day, id. at 647, 650-51 nn. 9, 11, and 12. Thus it is not surprising or significant that more than one-half (700 million) of the total (1.3 billion) eggs restricted pursuant to the regulations were Rose Acre's.

Rose Acre repeatedly emphasizes the end result of the regulatory scheme at issue — millions of its healthy eggs were restricted. However, this assertion itself relies on scientific understanding developed after the regulations were imposed (via the SE Pilot Project)14 and lifted (as reported in the 1998 "SE Risk Assessment"). Nowhere does Rose Acre argue (or did it show) that the regulatory means were inconsistent with knowledge the government possessed at the time they were adopted or applied against Rose Acre. Nor does Rose Acre contend that there was no nexus between those means and the substantial public purpose underlying the regulations — protecting the public against exposure to a potentially serious, even fatal, food-borne illness. Accordingly, the trial court's conclusion that the "character of the governmental action" factor favors Rose Acre, whether treated as a question of fact subject to review only for clear error or a question of law considered de novo, cannot be sustained.15

The Penn Central test was "designed to allow `careful examination and weighing of all the relevant circumstances.'" Tahoe-Sierra, 535 U.S. at 322, 122 S. Ct. 1465 (quoting Palazzolo, 533 U.S. at 636, 121 S. Ct. 2448 (O'Connor, J., concurring)). Whether the regulations at issue in this case went too far and, therefore, constituted a taking, Pa. Coal, 260 U.S. at 415, 43 S. Ct. 158, is thus determined by balancing their interference with Rose Acres right to use its property in accordance with its reasonable economic expectations against the substantiality of the governments purpose and the nexus between that purpose and the means undertaken to achieve it. We leave that task to the trial court on remand, but note several considerations that should feature in the analysis.

First, as noted above, courts have traditionally rejected takings claims in the absence of severe economic deprivation. This hesitation stems from the very nature of a regulatory takings claim. Such a claim, lacking the "typically obvious and undisputed" predicate of a physical invasion or appropriation of private property by the government, is in essence, a claim that "a taking has occurred because a law or regulation imposes restrictions so severe that they are tantamount to a condemnation or appropriation." Tahoe-Sierra, 535 U.S. at 322 n. 17, 122 S. Ct. 1465. In considering whether the economic deprivation here was of the requisite severity, the court should also consider the significance of the fact that the regulations restricted Rose Acres operations temporarily — for a period of about two years — after which Rose Acre returned to pre-restriction table egg sales levels.

Second, as of the time the restrictions were imposed, Rose Acre made 10 of its table egg sales within the state of Indiana, a fact noted but not further considered by the trial court. Rose Acre, 55 Fed.Cl. at 652. As noted above, state, not federal, officials were responsible for restricting intrastate sales. However, the trial court appears to have attributed the state-imposed restrictions to the federal government for purposes of the takings analysis. Such attribution is proper, however, only if the state officials were acting as agents of the federal government or pursuant to federal authority. B G Enters., Ltd. v. United States, 220 F.3d 1318, 1323-24 (Fed.Cir.2000). The regulations at issue provide no such authority, however, as they authorize limitations only on interstate movement of eggs from test and infected houses. 55 Fed.Reg. 5584; 56 Fed.Reg. 3741.16 Thus, unless a finding of agency or authority is otherwise appropriate, the trial court must consider only the actions of the federal government (and their effects) in evaluating Rose Acres takings claim.

Only when all of the relevant criteria and circumstances are considered, and considered together, can a conclusion be reached as to whether compensation is required in this case. On remand, then, the court must reevaluate the severity of the economic impact in accordance with the discussion above, and weigh that against the other Penn Central factors, taking into account all of the above-noted considerations.

II. Destroyed Hens: Categorical Takings Claim

The compensation the trial court awarded Rose Acre for the hens taken for necropsy represents a small portion ($15,671.99) of the total award. Rose Acre, 55 Fed.Cl. at 664. The government nonetheless also appeals that aspect of the trial court's judgment, arguing that (1) a categorical takings analysis is inapplicable and (2) the trial court's reliance on its finding that the "testing procedures were flawed" in its treatment of the "merits" of Rose Acre's claim as to its hens reveals that the court's erroneous finding in this regard contaminated its treatment of this claim, despite the court's invocation of the "categorical" label. The trial court correctly rejected the government's contention that a "per se" takings analysis is never applicable when personal property is at issue.17 Whether Rose Acre's claim for compensation for its hens should be so resolved is not so clear, however.

What is clear is that had the regulations required Rose Acre itself to kill and test the hens, no per se taking could be found. See Seiber v. United States, 364 F.3d 1356, 1370 (Fed.Cir.2004) (rejecting the applicability of per se treatment to a logging permit denial that entirely (albeit temporarily) destroyed the value of each affected tree even assuming each individual tree represents a separate property interest before cut from the land). The claimants in Seiber had invoked the holding in Lucas that a categorical taking results where a regulation prohibits all productive use of an entire parcel. Id. at 1368 (citing Lucas, 505 U.S. at 1015, 112 S. Ct. 2886). Setting aside (for purposes of decision) the issue of whether Tahoe-Sierra completely bars per se treatment for a temporary interference with property rights, we held that "the impact of an alleged taking must be considered in terms of the `parcel as a whole,' whether analyzed by categorical or ad hoc standards." Id. (citing the holding in Tahoe-Sierra, 535 U.S. at 329-32, 122 S. Ct. 1465, "that the Lucas per se rule only applies in the `extraordinary case' where three prerequisites are met: the regulation must (1) permanently deprive, (2) the whole property, (3) of all its value" (emphasis in original)). Accordingly, because the permit denial in Seiber did not affect "all of the timber on the two hundred-acre parcel," we rejected the claimants' categorical takings claim. Id. at 1370.

The question, then, is whether the mere fact that government officials carried out the testing (and the prerequisite seizure and destruction of the hens) is enough to transform what could otherwise qualify for takings compensation only if the Penn Central standard were met into a categorical taking. We think not. The Supreme Court has analyzed a takings claim that arose in an analogous context as a regulatory takings claim. See Ruckelshaus v. Monsanto Co., 467 U.S. 986, 104 S. Ct. 2862, 81 L. Ed. 2d 815 (1984). Monsanto concerned a federal statute that required pesticide manufacturers to register their products with the government prior to their sale in interstate or foreign commerce. Id. at 991, 104 S. Ct. 2862. Provisions of the statute authorized the Environmental Protection Agency ("EPA") to publicly disclose certain data submitted by registration applicants and, in some instances, to consider data submitted by one applicant in support of an application submitted by another party for registration of a similar chemical. Id. at 992-96, 104 S. Ct. 2862. Monsanto challenged these data-disclosure and data-consideration provisions on the ground that they effected a taking of those of its submitted trade secrets the government disclosed (or considered for the benefit of others) without just compensation. Id. at 998-99. Thus, as in the present case, the claim was that government officials, acting pursuant to one aspect of an integrated regulatory scheme enacted for the benefit of public health and safety, deprived the claimant of its interest in a portion of its relevant personal property as a condition for permission to sell its product in interstate commerce. After recognizing that Monsanto's health, safety, and environmental data was cognizable as a property right under Missouri law, id. at 1003-04, the Court "confront[ed] the difficult question whether a `taking' will occur when EPA discloses those data or considers the data in evaluating another application for registration" in accordance with the Penn Central analytical framework. Id. at 1004-14.

Here then, the same analysis must be employed to determine whether Rose Acre is entitled to compensation under the Fifth Amendment for its hens. The trial court, on remand, must evaluate the severity of the economic impact of the destroyed hens on Rose Acre in accordance with the standards and requirements discussed above in connection with its claim for compensation for the eggs affected by the regulations. Further inquiry into whether the regulations interfered with Rose Acre's reasonable investment-backed expectations and furthered a substantial public interest need not be undertaken in light of our above conclusions as to those issues, which apply with equal force to the claim for compensation for the hens. However, the court must ultimately weigh all the Penn Central factors, taking into account all the relevant circumstances, as noted above, to determine whether the Fifth Amendment requires compensation for the hens.

In accordance with the foregoing, we remand for reconsideration of the severity of the economic impact wrought by the relevant (i.e., federal) restrictions on Rose Acre, and for consideration of the significance of that impact in light of the other relevant factors, namely the regulations' interference with Rose Acre's reasonable investment-backed expectations and their furtherance of the substantial government health and safety interest.

VACATED and REMANDED.

No costs.