Rose Acre Farms, Inc., Cross-Appellant v. Edward Madigan, Secretary of Agriculture, Cross-Appellees

956 F.2d 670

• Court: Court of Appeals for the Seventh Circuit
• Decided: February 26, 1992
• Docket: 91-2358, 91-2514
• Status: Published

Opinion

EASTERBROOK, Circuit Judge.

The bacterium Salmonella enteritidis serotype enteritidis produces a violent illness, accompanied by fever, diarrhea, and vomiting, often requiring hospitalization. In the young, the old, and those weak for other reasons, salmonella can be fatal. (For convenience we call both the bacterium and the disease “salmonella.”) Milk and poultry products provide nutrients for the bacterium, and improperly handled food is its principal vector. An upsurge in reported instances of salmonella transmitted through food led federal officials to inquire why. One distressing possibility: a mutated form of the bacterium passes from chicken to egg before the shell forms, lying in wait to cause disease whenever the eggs are not cooked. As some persons use raw eggs to make hollandaise sauce and Caesar salads, and others do not store or cook eggs properly, the presence of bacteria poses a danger to public health. Hotels, nursing homes, and restaurants may use hundreds of eggs to produce batches of food. If even one egg contains salmonella, the batch will be contaminated; if the mixture sits at room temperature between steps in the preparation, the bacteria multiply rapidly.

The Department of Agriculture issued regulations that it believes will reduce the risk of salmonella transmitted in eggs. 56 Fed.Reg. 3730 (Jan. 30,1991), to be codified at 9 C.F.R. §§ 82.30-82.38. (Amendments at 57 Fed.Reg. 776 (Jan. 9, 1992), do not affect our case.) After salmonella strikes, federal officials try to find the source of the food that the victims ate. When eggs are involved, the Department tests both the chickens and their surroundings for the bacterium. While the testing is underway — and afterward, if salmonella is found in either the birds or their environment— the producer may not sell eggs from that flock for consumption as table eggs. Until the flock has been certified salmonella-free, the owner may sell the birds for meat or may sell the eggs to “breakers” (firms that incorporate the eggs into other products, such as cake mixes, after pasteurization that eliminates all risk of salmonella), but may not sell whole eggs in cartons.

Under this approach risk to consumers falls close to zero. The producer pays for the consumers’ gain. Whole table eggs fetch a higher price than eggs sold to breakers, and layer hens are worth much more when producing eggs than when slaughtered. A ban on the sale of whole eggs can turn a profitable operation into a losing one — and the loss may be substantial when the producer is specialized to the production of table eggs, as Rose Acre Farms is. See A.A. Poultry Farms, Inc. v. Rose Acre Farms, Inc., 881 F.2d 1396, 1407-08 (7th Cir.1989). Salmonella has been traced to three of Rose Acre’s flocks, and the firm filed this suit under the Administrative Procedure Act seeking an end to interference with its sale of whole eggs.

Rose Acre advanced three principal challenges to the regulations: first that they are beyond the power of the Department of Agriculture because they protect consumers rather than animals; second that any belief that salmonella in chickens poses a risk to consumers (or that these regulations would alleviate the problem) is arbitrary; third that the Department’s unwillingness to offer compensation for the loss created by the ban invalidates the rules. The district court rejected the first two lines of argument but accepted the third. 1991 U.S. Dist. Lexis 8691 (S.D.Ind.). The court went on to hold that the testing provisions in one rule are independently invalid. The Department of Agriculture has appealed. So has Rose Acre — a puzzling step, as it won in the district court. A prevailing party is entitled to advance in support of its judgment all arguments it presented to the district court. It need not and should not file a cross-appeal just because the district judge rejected one of its arguments on the way to deciding in its favor. Massachusetts Mutual Life Insurance Co. v. Ludwig, 426 U.S. 479, 96 S.Ct. 2158, 48 L.Ed.2d 784 (1976); Jordan v. Duff and Phelps, Inc., 815 F.2d 429, 439 (7th Cir.1987).

I

We start with the reason the district judge gave for annulling the regulations— that Rose Acre is entitled to compensation. There are three potential sources of obligation to compensate: the takings clause of the fifth amendment and two statutes, 21 U.S.C. §§ 114a and 134a(d). The Secretary of Agriculture believes that none of these requires compensation, not only because he has not ordered the destruction of any animal but also because the owners may make productive use of their birds by selling the eggs to breakers. To this Rose Acre replies that it is impossible to eliminate the bacterium from the environment without “depopulating” the chicken coops (a euphemism for killing the hens). Allowing farmers to sell the eggs or hens diminishes the loss but does not remove the directive from the category of takings.

Although the Secretary draws a bright line between ordering the destruction of the birds and ordering their owners to take other actions that lead to loss, we suppose that even the Department of Agriculture would concede that giving the owner a choice between killing the birds and lofting them into orbit (in pressure suits, so they do not die) is the equivalent of an order to destroy the birds. An “alternative” that is less attractive financially than slaughter is the functional equivalent of a command to destroy the animals. And after United States v. Causby, 328 U.S. 256, 66 S.Ct. 1062, 90 L.Ed. 1206 (1946), it is hardly possible to say that only the transfer or physical destruction of property is a “taking.” Airplanes flying low over a chicken farm caused some birds to die from fright; surviving birds laid fewer eggs. The diminution in the value of the farm was a “taking,” the Court held. Perhaps an order justified by the need to prevent the spread of disease does not call for compensation under the Constitution, see Miller v. Schoene, 276 U.S. 272, 48 S.Ct. 246, 72 L.Ed. 568 (1928), but Rose Acre observes that § 134a requires payment even when the destruction is justified by contagion.

Before issuing the regulations, the Secretary concluded that neither the Constitution nor a statute requires compensation. The preamble to the regulations states that the Department “does not intend to authorize payments of claims for chickens voluntarily destroyed, to order mandatory depopulation of infected chicken flocks and pay indemnities to the owners of destroyed chickens, or to make indemnity payments for eggs sold for pasteurization.” 56 Fed. Reg. at 3731-32. The district judge thought otherwise. Holding that compensation is due under § 134a if not directly under the fifth amendment, the district court set aside the regulations themselves. No compensation, no regulation.

If either a statute or the Constitution requires compensation, one would suppose, the right way to proceed is to order payment, not to permit the sale of eggs that may kill people. What induced the district court to scratch out the regulations rather than to order payments is the Tucker Act’s allocation of jurisdiction between district courts and the claims court. Only the claims court may award more than $10,000 against the United States on account of claims under the Constitution or statutes such as § 134a. Compare 28 U.S.C. § 1346(a)(2) with 28 U.S.C. § 1491(a)(1). Because the district court could not give Rose Acre the money to which that court believed it entitled, the court instead ordered the Secretary to cease the acts that gave rise to the need for compensation.

The Secretary contends that the district court lacks such power because the United States has not surrendered its immunity from suit, but this argument is unavailing. The APA allows judicial review of agency actions, and Congress has terminated sovereign immunity when the suit does not involve “money damages.” 5 U.S.C. § 702. Rose Acre did not obtain money damages, did not even obtain an order that will lead to damages. Instead the judge issued an order that eliminates the need for damages. Such an order is within the court’s subject-matter jurisdiction. See Bowen v. Massachusetts, 487 U.S. 879, 108 S.Ct. 2722, 101 L.Ed.2d 749 (1988).

Jurisdiction to enjoin enforcement of the regulations is not the problem. Instead the difficulty lies in the choice of remedy. If indeed the Constitution or a statute calls for compensation — a question on which we express no view — then setting aside the regulation is the wrong remedy. Compensation is the right one. The district court held the regulations invalid only because it could not award compensation. Yet Rose Acre’s decision to file this suit in a court that cannot award the proper remedy hardly entitles it to some substitute and inappropriate remedy.

Rose Acre asked the district court to forbid acts that it believes “take” its eggs, and the court obliged. But the takings clause does not forbid takings; it requires compensation for takings. It reads: “nor shall private property be taken for public use, without just compensation.” No one doubts that if these regulations “take” Rose Acre’s property, the taking is for “public use”. Cf. Berman v. Parker, 348 U.S. 26, 75 S.Ct. 98, 99 L.Ed. 27 (1954). If compensation is forthcoming, the Constitution allows the taking. The Tucker Act offers whatever compensation the Constitution requires. Rose Acre does not contend that the Tucker Act has been repealed by any of the agricultural statutes. Access to payment under the Tucker Act means that the taking may proceed. That a district court cannot order payment is irrelevant; the question is whether the United States will supply “just compensation.” Through the claims court it will (if any is due), so there is no justification for interference with the taking. Regional Rail Reorganization Act Cases, 419 U.S. 102, 124-25, 95 S.Ct. 335, 349, 42 L.Ed.2d 320 (1974); Ruckelshaus v. Monsanto Co., 467 U.S. 986, 1016, 104 S.Ct. 2862, 2879, 81 L.Ed.2d 815 (1984); Preseault v. ICC, 494 U.S. 1, 11-12, 110 S.Ct. 914, 921-922, 108 L.Ed.2d 1 (1990).

Congress could of course create an entitlement to be free o/takings in lieu of the constitutional requirement of compensation for takings. Rose Acre does not contend that any statute does this. The only statutes pertinent to the subject, 21 U.S.C. §§ 114a & 134a(d), are, like the fifth amendment itself, compensation rules rather than obstacles to the taking. Section 134a(d) provides: “Except as provided in subsection (e) of this section, the Secretary shall compensate the owner of any animal, carcass, product, or article destroyed pursuant to the provisions of this section.” Section 114a allows the Secretary to provide compensation in additional circumstances. Neither statute offers the slightest foundation for interdicting the order to destroy the animal or agricultural product.

As the district judge saw things, the warning that the Secretary “does not intend to authorize payments” is equivalent to forbidding payments. We do not read the language so. The Department has announced a litigating position. It believes that neither the fifth amendment nor a statute requires compensation, so it does not plan to pay. Someone who disagrees may repair to the claims court. If, as the district court believed, the Secretary is wrong, then the claims court will award compensation. If, however, the Secretary is right, then no compensation is due. Either way, it is inappropriate to halt the regulatory program. A court may not insist that the Department trumpet the producers’ view of the law; capitulation is not a precondition to regulation. Even if this is wrong, and a preamble to a regulation has the force of law, the appropriate remedy is to excise the offending text, and the appropriate surgeon is the claims court.

Rose Acre submits that the claims court lacks “jurisdiction” to “invalidate” the “rule” barring compensation. True it is that the claims court lacks general equitable powers, Richardson v. Morris, 409 U.S. 464, 93 S.Ct. 629, 34 L.Ed.2d 647 (1973), although it may award some equitable relief ancillary to an award of money. 28 U.S.C. § 1491(a)(2), (3). See Bowen v. Massachusetts, 487 U.S. at 905-08, 108 S.Ct. at 2737-39. Yet whether the claims court can pluck the preamble from the page is no concern of Rose Acre's. Notwithstanding anything in the Federal Register, the claims court will implement the Constitution and federal statutes. If Rose Acre gets just compensation, it has no further interest in the contents of musty journals. Almost the entire business of the claims court is awarding money over the opposition of the executive branch; that the executive announces an intention to oppose a given demand is no reason to think the claims court an inadequate forum.

Rose Acre must believe that the claims court will decide wrongly — that although compensation is due, the claims court and federal circuit will say that payment is unnecessary or will compute the compensation incorrectly. It should go without saying that the possibility of error by the right forum is no reason for a different forum to award an inappropriate remedy. The chance that the claims court will deny compensation even though money is due is no greater (and may well be less) than the risk that the district court will think compensation due even though it is not. It is to the claims court that Congress has entrusted the task of determining whether a statute or the Constitution requires compensation. It is to the claims court that Rose Acre must go.

II

The Secretary of Agriculture must “prepare such rules and regulations as he may deem necessary for the speedy and effectual suppression and extirpation of pleuropneumonia and other dangerous, contagious, infectious, and communicable diseases”. 21 U.S.C. § 114. Sections 111, 114a, and 120 confer similar powers. That salmonella is dangerous and infectious is common ground. Rose Acre observes that salmonella is not dangerous to the animals; it is contagious among animals but does not injure them. According to Rose Acre, only diseases dangerous to animals fall within the jurisdiction of the Secretary of Agriculture. When the disease is dangerous exclusively to humans, Rose Acre insists, the Food and Drug Administration is the appropriate regulator. The Secretary of Agriculture has neither statutory mandate nor expertise in dealing with human diseases, it submits. (Rose Acre assured us at oral argument that if the FDA had adopted the identical rules, it would not have argued that only the Department of Agriculture has a statutory mandate and expertise in dealing with farm animals. We are not comforted, but our suspicions are neither here nor there.)

The district judge rejected Rose Acre’s contention, and properly so. As counsel for the Secretary cracked at oral argument, Title 21 is not animal rights legislation. (The principal means for dealing with diseases are quarantining or killing the animals, not comforting or curing them.) Control of illness among farm animals is for the welfare of humans: to protect our health from diseases animals carry, and to protect our wallets from the costs of sacrificing additional animals should the infection spread. Salmonella spreads from animals to people; it spreads among animals, potentially increasing the financial cost to farmers; and fear of salmonella depresses the demand for dairy and poultry products, again injuring agriculture. Nothing in the text of §§ 111, 114, 114a, or 120 confines the Secretary to addressing diseases fatal to animals.

Any ambiguity in these statutes is for the Secretary to resolve. Section 114 and similar provisions are express delegations of power to make rules. Delegation includes the power to interpret. Contrast Pauley v. BethEnergy Mines, Inc., — U.S.-, 111 S.Ct. 2524, 2534, 115 L.Ed.2d 604 (1991), with Adams Fruit Co. v. Barrett, 494 U.S. 638, 649-50, 110 S.Ct. 1384, 1390-91, 108 L.Ed.2d 585 (1990). Congress called for the judgment of the Secretary, not the judgment of a court, on the appropriate response to diseases carried by animals. Nothing in these statutes forbids a judgment that danger to humans justifies regulation. Trichinosis may be harmless to pigs, but its suppression is within the Secretary’s purview. So too with salmonella.

Rose Acre protests that the regulatory scheme is silly. Properly cooked eggs are safe. Bacteria in the eggs therefore are harmless, and it is absurd to expend large sums keeping harmless substances out of eggs. To this the Secretary replies: eggs are not always stored and cooked properly. Kitchen workers may spread salmonella to hundreds of persons by oversight. These negligent workers cannot compensate the victims; even their employers may lack sufficient assets. The costs of carelessness in the preparation of food can be reduced by more care elsewhere in the food chain, with net savings — or so the Secretary is entitled to conclude. Evidence may be adduced on both sides, but the administrative record supports the Secretary’s conclusion that “scientific evidence suggests that SE [salmonella] is passed along to eggs before shell formation occurs if the hen is infected systemically with SE bacteria, and suggests vertical passage of SE from hens to chicks.” 56 Fed.Reg. at 3730.

The district court reached the same conclusion independently after receiving evidence, which is not an appropriate procedure. Courts review administrative decisions on administrative records; if the record or explanation is inadequate, the court should remand the proceedings rather than make its own decision. Cronin v. Department of Agriculture, 919 F.2d 439, 444 (7th Cir.1990) (“only in an emergency should a reviewing court ... conduct its own evidentiary hearing.”). Informal rule-making under the APA creates both a record and an explanation; the district court should have limited its attention to these. It is the Secretary’s assessment of the data, not a judge’s, that the statute makes controlling. We accept the Secretary’s judgment on the Secretary’s record. Cf. FCC v. National Citizens Committee for Broadcasting, 436 U.S. 775, 796-97, 98 S.Ct. 2096, 2112-13, 56 L.Ed.2d 697 (1978).

Even if the Secretary is entitled to curtail the volume of salmonella in eggs, Rose Acre submits, he is not entitled to do this in an irrational way. These regulations are irrational for two reasons, Rose Acre tells us: first, it would be cheaper to test the eggs for bacteria than to eliminate shipments from whole flocks; second, it is whimsical to limit testing and quarantine to flocks identified by tracing back from outbreaks of salmonella. Neither of these possibilities need detain us.

Testing for salmonella is expensive and difficult. See General Foods Corp. v. Valley Lea Dairies, Inc., 771 F.2d 1093 (7th Cir.1985). Testing destroys the eggs; testing even a large sample of eggs from a given flock will not be sufficient if the objective is to assure a very low probability of salmonella in any shipment. The lower the acceptable rate of infection, the greater the proportion of eggs that must be tested. Rose Acre has not suggested a testing regimen that both achieves the Secretary’s objective and costs less than the current regulations.

As for tracing: the Secretary is entitled to conclude that working backwards from an outbreak identifies the flocks that pose the greatest risk to humans. Selectivity is the essence of judgment. If grandfather clauses are legitimate, see FCC v. NCCB, 436 U.S. at 802-09, 98 S.Ct. at 2115-19, tracing rules are not troubling. Of course the match is not perfect. Maybe 1% of all eggs are infected, and tracing catches only luckless suppliers whose eggs fell into the hands of careless food preparers. But if this is so, tracing is equivalent to random testing. Often the best way to enforce a rule is by spot checks — testing, say, one of every hundred flocks each year and acting sternly when the test reveals salmonella. A farmer selected randomly for testing could not protest that the Secretary should have tested 2% or 10% of all flocks rather than 1%. Cf. FTC v. Universal-Rundle Corp., 387 U.S. 244, 87 S.Ct. 1622, 18 L.Ed.2d 749 (1967); Moog Industries, Inc. v. FTC, 355 U.S. 411, 78 S.Ct. 377, 2 L.Ed.2d 370 (1958). If the Secretary is right, tracing helps find and deal with the most dangerous sources of salmonella at the least cost; if Rose Acre is right, tracing is the equivalent of random checks; either way, tracing is a lawful strategy.

Ill

Although the district court concluded that the rules as a whole are within the Secretary’s power (questions about compensation to one side), it held 9 C.F.R. § 82.38, 56 Fed.Reg. at 3742-43, invalid as arbitrary.

Once an outbreak of salmonella has been traced to a source, the flock is tested and, if any test is positive, sales are restricted. If however the producer can show that it maintains “biosecurity” between poultry houses, only the birds in coops where salmonella has been found are deemed infected; the producer can freely sell the eggs from other layers. “Biosecurity means that flock management procedures are in place to ensure that there is no contact between poultry houses through exposure to chickens, feed, water, manure, equipment, or personnel from other poultry houses.” 9 C.F.R. § 82.30 (definition of “separate poultry house”), 56 Fed.Reg. at 3739. Section 82.38(b) then provides that if any poultry house on a farm is infected, “[a]ll other poultry houses ... shall undergo monitoring tests ... until 120 days after the date infected house status is removed from all poultry houses on the premises.” Monitoring means that a federal or state employee “shall collect manure and egg transport machinery samples from each house ... at intervals of not less than 45 days and not more than 60 days.” Any house from which a sample is positive for salmonella is designated a “test house,” with shipments restricted until further testing shows it free from salmonella or the house has been “depopulated” and disinfected.

It is this repetitious testing on farms containing one or more infected poultry houses that the district court deemed arbitrary. Given “biosecurity,” the court thought, there is no reason to believe that a poultry house is at special risk just because some other house on the farm is infected, and therefore no reason for special testing. Grant the premise, and there is also no less reason to suppose that such a house is infected than is any randomly selected house at a randomly selected farm. Random testing is appropriate — indeed, Rose Acre wants the Secretary to do less tracing and more comprehensive or at least random testing — so testing of houses in this fashion also would be appropriate.

What is more, the Secretary does not grant the premise. “Biosecurity” under these regulations is not all-or-nothing. It is not as if a “separate poultry house” is inside a pressurized bubble, with food, water, air, clothing and the farm workers contained therein sterilized as they enter. Not even biological laboratories achieve such isolation. It brings a bit of mirth to our grey profession to compare a poultry farm with a center for research into recombinant DNA or a plant where vaccine is manufactured — places where greater efforts in the direction of biosecurity cost hundreds of dollars per pound of product. Farms are biologically leaky. Salmonella from the infected house may spread to nearby houses despite the best efforts. Testing to find out whether this has happened is a precaution. It may be that the probability of spreading does not justify the costs, but such arguments are for the Secretary and not the court. We repeat, it is the Secretary’s view of the costs and benefits of regulation that Congress has made controlling.

Many thoughtful persons believe that the Code of Federal Regulations overflows with unjustified, even perverse, rules; every thoughtful person believes that some of the rules are unjustified. Under the APA, the judicial task is limited to pruning the outliers. If the administrative record does not reveal a compelling need to adopt these regulations in the form the Secretary chose, it assuredly does not reveal that the regulations are muddleheaded meddlesomeness. How far, and how, to regulate poultry production to curb the risk of salmonella are economic, social, and political rather than legal questions. How far compensation may be necessary is a legal question, but one for another court.

Reversed