FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
RANCHERS CATTLEMEN ACTION
LEGAL FUND UNITED
STOCKGROWERS OF AMERICA,
Plaintiff-Appellee,
No. 05-35264
v.
UNITED STATES DEPARTMENT OF D.C. No.
CV 05-006 RFC
AGRICULTURE, Animal and Plant
OPINION
Health Inspection Service; MIKE
JOHANNS, in his capacity as the
Secretary of Agriculture,
Defendants-Appellants.
Appeal from the United States District Court
for the District of Montana
Richard F. Cebull, District Judge, Presiding
Argued and Submitted
July 13, 2005—Seattle, Washington
Filed July 25, 2005
Before: A. Wallace Tashima, Richard A. Paez, and
Consuelo M. Callahan, Circuit Judges.
Opinion by Judge Tashima
8731
RANCHERS CATTLEMEN v. USDA 8735
COUNSEL
Mark B. Stern, Civil Division, U.S. Department of Justice,
Washington, DC, for the defendants-appellants.
Russell S. Frye, Frye Law PLLC, Washington, DC, for the
plaintiff-appellee.
Maureen E. Mahoney, Latham & Watkins LLP, Washington,
DC, for amicus curiae Government of Canada.
Gregory G. Garre, Hogan & Hartson LLP, Washington, DC,
for amici curiae National Cattlemen’s Beef Association,
American Farm Bureau Federation, National Pork Producers
Council, 29 State Cattlemen’s Associations, 18 State Farm
Bureaus, and 9 Individual Cattle Producers.
Sarah Weinstein, Mayer, Brown, Rowe & Maw LLP, Palo
Alto, California, for amicus curiae Alberta Beef Producers.
Michael B. Gillett, McElroy Law Firm, PLLC, Seattle, Wash-
ington, for amicus curiae Easterday Ranches, Inc.
Joseph O. Click, Blank Rome LLP, Washington, DC, for
amici curiae Canadian Cattlemen’s Association and Its Affili-
ated Organizations.
Jonathan L. Abram, Hogan & Hartson, Washington, DC, for
amici curiae American Meat Institute, North American Meat
Processors, Southwestern Meat Association, Eastern Meat
8736 RANCHERS CATTLEMEN v. USDA
Packers Association, American Association of Meat Proces-
sors, National Restaurant Association, and United Food and
Commercial Workers.
John O’Brien, Kerr Brosseau Bartlett O’Brien, LLC, Denver,
Colorado, for amicus curiae Pioneer, Inc.
Gregg Spyridon, Spyridon, Koch, Palermo, & Dornan, LLC,
Metairie, Louisiana, for amici curiae the Camelid Alliance, et
al. Alan Charles Raul, Sidley Austin Brown & Wood LLP,
Washington, DC, for amicus curiae Tyson Foods, Inc.
David A. Domina, Domina Law pc llo, Omaha, Nebraska, for
amici curiae 67 National, State, and Local Consumer and
Research Groups, Public Interest Organizations, Farm and
Ranch Organizations, and Local and Private Organizations.
Christian D. Tweeten, Chief Civil Counsel, Montana Attorney
General, Helena, Montana, for amici curiae States of Mon-
tana, Connecticut, Nevada, New Mexico, North Dakota, and
South Dakota.
OPINION
TASHIMA, Circuit Judge:
We must decide whether the district court erred in issuing
a preliminary injunction prohibiting the implementation of a
regulation of the United States Department of Agriculture
(“USDA”) permitting the resumption of the importation of
Canadian cattle into the United States. We conclude that it did
and therefore reverse the district court.
At the heart of this case lies a relatively new cattle disease
caused by the practice of feeding cows, herbivores by nature,
the brains and other central nervous system tissues of other
RANCHERS CATTLEMEN v. USDA 8737
cows. Technically known as Bovine Spongiform Encephalop-
athy (“BSE”), this disease, popularly known as mad cow dis-
ease, has spread from farms in England to 25 countries around
the world since its discovery in 1986.
As BSE spread throughout the globe during the past 20
years, USDA instituted a policy of barring the importation of
ruminants1 and ruminant products from countries where BSE
was known to exist. In a final rule entitled Bovine Spongiform
Encephalopathy: Minimal Risk Regions and Importation of
Commodities; Final Rule and Notice, 70 Fed. Reg. 460 (Jan.
4, 2005) (the “Final Rule”), USDA relaxed this longstanding
practice, allowing limited ruminant imports from Canada,
despite the fact that two cases of BSE had been found in Can-
ada at the time.
Plaintiff-Appellee, Ranchers Cattlemen Action Legal Fund
United Stockgrowers of America (“R-CALF”), successfully
blocked the implementation of the Final Rule, convincing the
court below to find the rule arbitrary and capricious under the
Administrative Procedure Act (“APA”), 5 U.S.C. § 706(2),
and to issue a preliminary injunction prohibiting its enforce-
ment. See Ranchers Cattlemen Action Legal Fund United
Stockgrowers of Am. v. United States Dep’t of Agric., 359 F.
Supp. 2d 1058 (D. Mont. 2005) (“R-CALF I”). Because we
conclude that the district court applied an incorrect legal stan-
dard, we reverse.2
1
Ruminants are hoofed mammals generally defined by their four-
chambered stomachs and their practice of chewing a cud consisting of
regurgitated, partially digested food. Ruminants include cattle, sheep,
goats, deer, giraffes, camels, llamas, and okapi, among others.
2
On July 14, 2005, after the completion of briefing and oral argument
we issued a stay of the preliminary injunction pending the resolution of
this appeal. See Fed. R. App. P. 8(a).
8738 RANCHERS CATTLEMEN v. USDA
I. FACTUAL AND PROCEDURAL BACKGROUND
A. Bovine Spongiform Encephalopathy
BSE was first diagnosed in England in the late 1980s. This
new disease spread rapidly, infecting thousands of English
cattle and eventually reaching countries all over the globe.
Although the disease has since been largely contained, it con-
tinues to persist, and it resides at the center of the current law-
suit.
BSE is a species of Transmissible Spongiform Encephalop-
athy (“TSE”), a family of degenerative neurological diseases
that affects a wide range of animals, including sheep, goats,
and deer, as well as humans. Although there remains some
dispute, it is widely believed that BSE and other TSEs are
caused by prions, abnormally shaped and extremely hardy
proteins that were only recently discovered.
TSEs have a debilitating neurological impact on their vic-
tims. After an incubation period of months or years, the dis-
eases create myriad tiny holes in the brain, slowly
deteriorating their victims’ mental and physical abilities until
death eventually results. In cattle, BSE has an incubation
period of two to eight years, during which time the infected
animal shows no outward sign of the illness. Once the disease
progresses, however, infected cattle begin showing symptoms
within two to three months. These symptoms can include ner-
vousness or aggression, abnormal posture, impaired coordina-
tion, decreased milk production, and loss of body condition
despite continued appetite.
At the height of the BSE epidemic in the United Kingdom,
tens of thousands of cattle were confirmed to have the dis-
ease, and by some estimates the number of infected cattle in
the United Kingdom may have reached into the millions. All
told, there have been more than 187,000 confirmed cases of
RANCHERS CATTLEMEN v. USDA 8739
BSE in cattle worldwide, over 95 percent of which have
occurred in the United Kingdom.
Epidemiological investigations in England quickly deter-
mined that BSE was likely spread through cattle feed that was
infected with the BSE agent. The blame for the contaminated
feed fell squarely on the practice, common in Europe at the
time, of creating high-protein cattle feed through the “recy-
cling” of otherwise unusable cattle parts. This process is
known as “rendering,” and involves placing animal protein in
large tanks and cooking at temperatures high enough to kill
most microorganisms.3 Although the rendering process is able
to eliminate most bacterial and viral diseases, the BSE agent
is resistant enough to heat and other sterilization processes to
withstand the conversion into feed. Infected tissue from a sin-
gle infected cow, when rendered into cattle feed, could there-
fore be fed to hundreds of cattle, exposing them all to the
possibility of infection.
Several years after the discovery of BSE, the disease
became a matter of much more serious concern. In 1996, the
British government announced that a new form of TSE in
humans, variant Creutzfeldt-Jakob Disease (“vCJD”), was
likely caused by human consumption of cattle products that
were contaminated with the BSE agent. To date, only approx-
3
Rendering continues to this day in the United States, where approxi-
mately 50 billion pounds of tissue from dead animals are converted into
animal feed each year. The breadth of the practice at one Baltimore ren-
dering facility has been reported to include:
Bozeman, the Baltimore City Police Department quarter horse
who died last summer in the line of duty. . . . A baby circus ele-
phant who died while in Baltimore this summer. Millions of tons
of waste meat and inedible animal parts from the region’s super-
markets and slaughterhouses. Carcasses from the Baltimore zoo.
The thousands of dead dogs, cats, raccoons, possums, deer, foxes,
snakes, and the rest that local animal shelters and road-kill patrols
must dispose of each month.
Van Smith, What’s Cookin?, Baltimore City Paper, Sept. 27, 1995.
8740 RANCHERS CATTLEMEN v. USDA
imately 150 cases of vCJD have been identified worldwide,
the vast majority of which occurred in England during the
height of its BSE epidemic. Although vCJD has been diag-
nosed in two people in North America, in both cases the dis-
ease is believed to have been contracted in England; no case
of vCJD has ever been linked to North American beef.4
Because BSE is a relatively new disease, and because
prions are a relatively recent scientific discovery, the state of
knowledge surrounding BSE is somewhat incomplete. Efforts
to understand the disease fully have been hampered because
current testing methodology is not particularly effective in
identifying it. No live animal test for BSE exists, meaning that
cows must be slaughtered before they can be tested. In addi-
tion, the tests that do exist are unable to detect the disease
during the vast majority of the time a cow is infected. The
earliest point at which current tests can detect the disease is
two to three months before an animal starts showing clinical
signs of infection. BSE has an incubation period that lasts for
four to five years on average, however, during which the ani-
mal carries the disease but shows no outward symptoms.
Given these testing limitations, there remain a number of
open public health questions surrounding BSE, in particular
concerning the means through which the disease can be trans-
mitted. The only documented method of BSE transmission is
through the consumption of feed contaminated with the BSE
agent. Some research involving both BSE and other TSEs,
however, suggests that BSE may be transmitted through
means other than contaminated feed. For example, in experi-
ments on sheep, mice, and hamsters, both BSE and scrapie, a
TSE disease that affects sheep, were transmitted through
4
The single known case of vCJD in the United States occurred in a Flor-
ida woman who was born in England in 1979. It is believed she was
exposed to BSE before she moved to the United States in 1992. Similarly,
the single case of vCJD in Canada occurred in a man who had stayed in
the United Kingdom on multiple trips.
RANCHERS CATTLEMEN v. USDA 8741
whole blood transfusion. At least one case of vCJD is also
believed to have been transmitted through human blood trans-
fusion. Other studies have suggested that prions can be
exchanged through saliva, while still others suggest that BSE
may be transmitted maternally.
Despite the highly infectious nature of the BSE agent, evi-
dence suggests that meat from cows infected with BSE may
be safely consumed by humans because BSE does not occur
in all parts of its host. Specifically, the BSE agent appears not
to exist in muscle tissue of cattle. Rather, the disease is gener-
ally confined to the central nervous system — the brain, spi-
nal cord, eyes, dorsal root ganglia, and trigeminal ganglia5 —
although it has also been found in the tonsils and distal ilium,
a part of the small intestine, of cattle. Research on other TSEs,
however, calls into question whether the BSE agent is truly
limited to these tissues. Specifically, some research has sug-
gested that sheep infected with scrapie may have prions in
their muscle tissue.
Despite the fact that it has only been known to exist for 20
years, the geographic range of BSE is substantial. From
England, it has spread to cattle in most of Europe, as well as
in the Middle East, Japan, and Canada. 9 C.F.R. § 94.18(a)(1)
(2003). As of the date of the district court’s opinion, however,
BSE had never occurred in a cow native to the United States.
That changed on June 24, 2005, when the Secretary of Agri-
culture announced that a cow in Texas had tested positive for
BSE. Statement by Dr. John Clifford Regarding the Epidemi-
ological Investigation into the recently confirmed BSE case
(June 29, 2005), available at http://www.aphis.usda.gov/lpa/
issues/bse/bse.html. A subsequent investigation revealed that
the cow was born in the United States approximately 12 years
ago.
5
Trigeminal ganglia are clusters of nerve cells connected to the brain
that lie close to the exterior of the skull. Dorsal root ganglia are clusters
of nerve cells attached to the spinal cord and contained within the bones
of the vertebral column.
8742 RANCHERS CATTLEMEN v. USDA
B. United States Regulation of BSE
The federal government has implemented a number of
safety measures to minimize the threat of BSE to U.S. citizens
and livestock. These precautions consist of an interlocking
regulatory framework overseen by three different federal
agencies. First and foremost, since 1997, the Food and Drug
Administration (“FDA”) has overseen a feed ban that prohib-
its the feeding of ruminant protein to other ruminants. See 21
C.F.R. § 589.2000 (2005). Such feed bans are generally the
first line of defense against the spread of BSE, and they have
been highly effective in other countries. The prevalence of
BSE in the United Kingdom, for example, dropped drastically
after it implemented its feed ban.
Critics, however, question whether the FDA feed ban is
truly effective. See, e.g., Thomas O. McGarity, Federal Regu-
lation of Mad Cow Disease Risks, 57 Admin. L. Rev. 289,
307 (2005). Given the highly infectious and resilient nature of
the BSE agent, these critics argue that the FDA feed ban has
“gaps” that could result in the use of feed derived from ren-
dered cattle protein as feed for cattle. For example, cattle are
allowed to be fed human “plate waste” from establishments
such as amusement parks, despite the fact that this plate waste
may contain beef products. In addition, the feed ban allows
rendered cattle protein to be fed to non-ruminants, such as
pigs and chickens. Thus, BSE could be spread through misla-
beled feed or through misfeeding on a farm. Finally, waste
from the floor of chicken coops is commonly scooped up and
fed to cattle; uneaten chicken feed or chicken droppings that
contain the BSE agent could therefore be fed to cattle via this
procedure.
An agency within USDA, Food Safety and Inspection Ser-
vices (“FSIS”), oversees a second line of defense against
BSE. FSIS promulgates regulations to ensure that the nation’s
food supply of meat, eggs, and poultry is safe. See http://
www.fsis.usda.gov/About_FSIS/index.asp. These regulations
RANCHERS CATTLEMEN v. USDA 8743
restrict certain cattle parts from being incorporated into the
human food supply. For example, FSIS regulations prohibit
the use of “downer” cattle6 as human food because inability
to stand is a common BSE symptom. 9 C.F.R. § 309.2 (2005).
FSIS regulations also prohibit those cattle parts that have
demonstrated BSE infectivity, known as specified risk materi-
als (“SRMs”), from being used in human food.7 9 C.F.R.
§ 310.22 (2005). Finally, FSIS regulations prohibit certain
methods of slaughter and butchering thought to increase the
risk of contaminating meat with central nervous system tissues.8
6
Non-ambulatory or “downer” cattle are cattle that “cannot rise from a
recumbent position or that cannot walk.” 9 C.F.R. § 309.2(b) (2005). FSIS
banned these cattle from the human food supply because “surveillance
data from European countries in which BSE has been detected indicate
that non-ambulatory cattle are among the animals that have a greater inci-
dence of BSE than other cattle.” Prohibition of the Use of Specified Risk
Materials and Requirements for the Disposition of Non-Ambulatory Dis-
abled Cattle, 69 Fed. Reg. 1862, 1862 (January 12, 2004) (“FSIS SRM
Rule”).
7
Because BSE infectivity spreads as a cow ages, current regulations
define only the distal ilium and tonsils of all cattle to be SRMs. 9 C.F.R.
§ 310.22(a) (2005). The brain, spinal cord, and other central nervous sys-
tem components are only considered to be SRMs in cattle of 30 months
of age and older. Id.
8
Specifically, FSIS regulations prohibit the use of “air-injection captive
bolt stunning,” a process through which a metal bolt and compressed air
are driven into the cranium of cattle, because the practice poses a risk of
contaminating edible meat with central nervous system tissue. See 9
C.F.R. § 310.13(a)(2)(iv)(C) (2005). The regulations also prohibit the use
of “Advanced Meat Recovery” systems and the labeling of “mechanically
separated beef” as meat. See FSIS SRM Rule, 69 Fed. Reg. at 1866. The
former “is a technology that enables processors to remove the attached
skeletal muscle tissue from livestock bones without incorporating a signif-
icant amount of bone or bone product into the final meat product.” Id.; see
also Meat Produced by Advanced Meat/Bone Separation Machinery and
Meat Recovery (AMR) Systems, 69 Fed. Reg. 1874, 1876 (Jan. 12, 2004).
The latter “is a paste-like and batter-like meat product produced by forcing
[beef] bones with attached edible meat under high pressure through a
sieve.” See http://www.fsis.usda.gov/oa/pubs/lablterm.htm.
8744 RANCHERS CATTLEMEN v. USDA
Another branch of USDA, Animal and Plant Health Inspec-
tion Services (“APHIS”), provides the final link in the regula-
tory framework. APHIS promulgates regulations designed to
protect the United States from the introduction of BSE from
other countries. To achieve this goal, until the Final Rule was
promulgated, APHIS banned the importation of all ruminants
and ruminant products from countries where BSE was known
to exist. See 9 C.F.R. §§ 93.401, 94.18 (2003).
APHIS has also been actively involved in the development
of international guidelines to fight the spread of BSE. In this
role, APHIS works with the Office International des Epizoo-
ties (“OIE”), the organization recognized by the World Trade
Organization as responsible for the development and periodic
review of standards, guidelines, and recommendations with
respect to animal health and “zoonoses” (diseases that are
transmissible from animals to humans).
C. Factual Background
Early this year, APHIS announced its decision to relax its
ban on the importation of ruminants and ruminant products
from countries where BSE was known to exist. The genesis
of this policy change occurred on May 20, 2003, when a cow
in Alberta was diagnosed with BSE. This represented not only
the first case of BSE native to North America, but it wreaked
havoc on the highly integrated beef market that exists
between the United States and Canada. Shortly after the
infected cow was announced, then Secretary of Agriculture
Veneman issued an Emergency Order adding Canada to the
list of regions where BSE was known to exist. Change in Dis-
ease Status of Canada Because of BSE, 68 Fed. Reg. 31,939
(May 29, 2003). Under the regulations then in effect, all
imports of live ruminants or ruminant meat products from
Canada were prohibited. See 9 C.F.R. §§ 93.401, 94.18
(2003).
Beginning in August 2003, the Secretary incrementally
began moving to reopen the border to Canadian ruminants
RANCHERS CATTLEMEN v. USDA 8745
and ruminant products and to reestablish the voluminous
North American beef trade. On August 8, 2003, the Secretary
announced that she would begin allowing certain “low-risk”
ruminant products to be imported into the United States from
Canada, the most significant of which was “[b]oneless bovine
meat from cattle under 30 months of age.” See Final Rule, 70
Fed. Reg. at 536; USDA News Release No. 0281.03 (Aug. 8,
2003), available at http://www.usda.gov/wps/portal.
On November 4, 2003, the Secretary published notice of a
proposed rule, seeking to amend the regulations governing the
importation of ruminants from countries where BSE is known
to exist. Bovine Spongiform Encephalopathy; Minimal Risk
Regions and Importation of Commodities, 68 Fed. Reg.
62,386 (Nov. 4, 2003). The proposed rule would have allowed
the importation of ruminants from countries in a newly cre-
ated category — “regions that present a minimal risk of intro-
ducing [BSE] into the United States via live ruminants and
ruminant products.” Id. The new regulation proposed to desig-
nate only Canada as a minimal-risk region. Id. The comment
period for the proposed rule was set to expire on January 5,
2004. Id.
A month and a half after the Secretary published the notice
of proposed rule, on December 23, 2003, a cow in Washing-
ton State was diagnosed with BSE. Bovine Spongiform
Encephalopathy; Minimal Risk Regions and Importation of
Commodities, 69 Fed. Reg. 10,633 (Mar. 8, 2004). An investi-
gation revealed that the cow was born in Canada and was
imported into the United States in 2001. Id. at 10,634. Given
that the cow was born before Canada’s feed ban went into
effect in 1997, USDA determined that the likeliest cause of its
BSE infection was contaminated feed. Id. Nevertheless, in
response to this discovery USDA reopened the comment
period for its proposed rule for an additional 30 days, extend-
ing it until April 7, 2004. Id. at 10,633.
8746 RANCHERS CATTLEMEN v. USDA
On April 19, 2004, USDA moved, without public notice, to
expand the types of ruminant products eligible to be imported
from Canada.9 Ranchers Cattlemen Action Legal Fund United
Stockgrowers of Am. v. United States Dep’t of Agric., 2004
WL 1047837 (D. Mont. 2004) (“R-CALF TRO”). R-CALF
sued to prevent this move, and the district court granted a
temporary restraining order on April 26, 2004, barring the
Secretary from proceeding with that plan. Id.
On January 4, 2005, USDA published its Final Rule. The
agency, after having considered 3,379 comments from inter-
ested parties, proceeded with its plan to reopen the border to
Canadian ruminants and ruminant products. Final Rule, 70
Fed. Reg. at 460, 469. Among other provisions, the Final Rule
allowed the importation of Canadian cattle under 30 months
of age provided the cattle were immediately slaughtered or
fed and then slaughtered.10 Id. at 548. The Final Rule also per-
mitted the importation of beef products from Canadian cattle
of all ages. Id. at 461, 465. The rule was scheduled to go into
effect on March 7, 2005. Id. at 460.
9
Specifically, USDA issued a memorandum stating that, effective April
19, 2004, all existing permits to import meat from Canada “will be
deemed to cover all edible bovine meat products (bone-in, boneless,
ground meat, further processed),” provided each shipment is accompanied
by a statement that the meat was processed in “establishments that are cer-
tified to FSIS as eligible for export to the United States.” R-CALF TRO,
2004 WL 1047837 at *2. USDA also published a table identifying “Low
Risk Canadian Products.” That table included “boneless, bone-in, ground
meat, and further processed bovine meat products,” bovine tongue, bovine
hearts, kidneys, and tripe, and bovine lips.” Id.
10
According to amicus Pioneer, Inc., a family-owned feedlot, the cattle
industry is generally comprised of three parts: ranchers, who breed cattle
and grow them until they reach approximately 650 pounds; feedlots,
which purchase cattle from ranchers and feed them high protein feed until
they reach approximately 1,150 pounds; and meat packers, which pur-
chase cattle from feedlots and process them for human consumption. Thus,
the Final Rule allowed Canadian cattle either to be sold to a feedlot for
feeding or to be sold directly to a meat packing company for slaughter.
RANCHERS CATTLEMEN v. USDA 8747
At roughly the same time that USDA published its Final
Rule, two additional cases of BSE were confirmed in Alberta
— one on January 2, 2005, and another on January 11. Bovine
Spongiform Encephalopathy; Minimal Risk Regions and
Importation of Commodities; Finding of No Significant
Impact and Affirmation of Final Rule, 70 Fed. Reg. 18,252,
18,254 (Apr. 8, 2005). One of these cows, like the two previ-
ous Canadian cattle diagnosed with BSE, was born before
Canada’s feed ban; the other, however, was born shortly
thereafter. Id. at 18,258. Once again, USDA attributed the
infections in both cows to contaminated feed manufactured
before Canada’s feed ban went into effect. Id. at 18,255.
Nonetheless, USDA indefinitely suspended the implementa-
tion of the portion of its Final Rule that permitted the importa-
tion of beef products from cattle over 30 months of age.11
Bovine Spongiform Encephalopathy; Minimal Risk Regions
and Importation of Commodities; Partial Delay of Applicabil-
ity, 70 Fed. Reg. 12,112 (Mar. 11, 2005).
D. Procedural History
Six days after USDA published the Final Rule, R-CALF
filed this action, seeking to enjoin the rule’s implementation.12
In its complaint, R-CALF alleged that USDA’s rulemaking
violated the Administrative Procedure Act (“APA”), the Reg-
ulatory Flexibility Act (“RFA”), and the National Environ-
mental Policy Act (“NEPA”). On February 1, 2005, three
weeks after filing its complaint, R-CALF filed its application
for a preliminary injunction to enjoin the Final Rule pendente
lite.
11
As mentioned above, an additional cow in the United States tested
positive for BSE on June 24, 2005. Because this cow was approximately
12 years old, USDA has attributed its infection to contaminated feed it
was exposed to before the U.S. feed ban came into effect.
12
R-CALF describes itself as a non-profit cattle association that repre-
sents U.S. “cattle producers, cattle backgrounders, and independent feedlot
owners” on issues concerning international trade and marketing.
8748 RANCHERS CATTLEMEN v. USDA
On March 2, 2005, the district court issued a preliminary
injunction, barring USDA from implementing its Final Rule.
See R-CALF I, 359 F. Supp. 2d at 1074. The district court’s
primary reason for enjoining the Final Rule was its finding
that the rule was arbitrary and capricious in violation of the
APA. Id. at 1063-69; see also 5 U.S.C. § 706(2). The district
court’s overarching concern was that USDA, “ignoring its
statutory mandate to protect the health and welfare of the peo-
ple of the United States, established its goal of re-opening the
border to the importation of live beef from Canada and there-
after attempted to work backwards to support and justify this
goal.” R-CALF I, 359 F. Supp. 2d at 1066. Given the agency’s
“preconceived intention, based upon inappropriate consider-
ations, to rush to reopen the border regardless of uncertainties
in the agency’s knowledge,” the district court found the Final
Rule to be arbitrary and capricious. Id. at 1074.
The district court specifically based its determination that
the Final Rule was arbitrary and capricious under the APA on
six independent grounds. First, the court found that USDA
failed adequately to quantify the risk of Canadian cattle to
humans, instead relying on a qualitative statement that the risk
was “low” or “very low.” Id. at 1064-65. Without a quantita-
tive assessment, the district court felt that it “ha[d] no way of
assessing the merits of the USDA’s actions.” Id. at 1065.
Second, the district court held that USDA had erroneously
calculated the prevalence of BSE in the Canadian herd. Id. at
1065-66. USDA had divided the number of cases in the last
12 months (two) by the total size of the Canadian herd over
24 months of age (5.5 million) to arrive at a prevalence rate
of approximately 0.4 cases per million head of adult cattle.
Final Rule, 70 Fed. Reg. at 464. The district court rejected
this calculation, however, and instead adopted R-CALF’s
measure of 5.5 cases per million head.13 R-CALF I, 359 F.
Supp. 2d at 1066.
13
To achieve the rate of 5.5 cases per million head, R-CALF calculated
the prevalence of BSE among tested cattle in Alberta (one in 3,000) and
RANCHERS CATTLEMEN v. USDA 8749
Third, the district court found that USDA’s reliance on the
Canadian feed ban was unjustified. Id. at 1066-68. The court
found that the science was uncertain in this area and that
methods of BSE transmission other than consumption of con-
taminated feed may exist. Id. at 1066. It also found that the
feed ban had not been in place an adequate amount of time,
and that it was not fully effective because it allowed both
bovine blood and rendered animal fat in cattle feed. Id. at
1067-68.
Fourth, the court found that USDA’s reliance on the
removal of SRMs to protect human health was also unjusti-
fied. Id. at 1068. According to the district court, evidence
indicated that “it is no longer reasonable to presume that there
is no risk of exposure to BSE infectious agents once an SRM
removal requirement is in place.” Id.
Fifth, the district court found that USDA’s failure to ban
the importation of pregnant cows was arbitrary and capri-
cious. Id. at 1069. According to the district court, BSE may
be transmitted both maternally and through fetal bovine
blood. Id. Thus, because the Final Rule did not require heifers
to be pregnancy checked as a condition of entry into the
United States, calves born to imported cattle could become “a
vector for BSE infection in the U.S.” Id.
Finally, the district court found that USDA had failed to
respond adequately to comments recommending mandatory
BSE testing for Canadian cattle. Id. Because testing can iden-
tify a BSE infection up to three months before the cow shows
outward signs of the disease, the court found that testing
divided it by 60, the assumed amount by which tested cattle will have BSE
over untested cattle (because tested cattle, which show outward signs of
the disease, are more likely to have BSE than the population at large). The
result is one infected cow per 180,000 head of cattle, or approximately
5.56 per million.”
8750 RANCHERS CATTLEMEN v. USDA
would be useful because it would “detect some cases of BSE
that would otherwise go undetected.” Id. In light of the “irrep-
arable injury” that it believed a case of BSE would cause, the
court viewed USDA’s actions as arbitrary and capricious. Id.
In addition to finding the Final Rule arbitrary and capri-
cious under the APA, the district court also relied on two
other bases for enjoining its implementation. First, the court
held that USDA had failed to satisfy NEPA’s procedural
requirements, both by failing to make its environmental
assessment available for public review and comment before
the Final Rule was published, and by failing to prepare an
environmental impact statement. Id. at 1069-71. Second, the
court concluded that USDA had violated the RFA by failing
to consider whether product labeling or voluntary BSE testing
would have mitigated the Final Rule’s impact on small busi-
nesses. Id. at 1071-73.
Based on the above, the district court found that R-CALF
had raised “very serious questions on the merits.” Id. at 1074.
The district court also found that R-CALF, and the American
public, would be irreparably harmed by allowing the importa-
tion of Canadian beef. Id. at 1073-74. The court specifically
found that the introduction of BSE into the United States
would cause irreparable harm to the American public because
of the increased risk of vCJD to consumers of beef. Id. at
1073. Further, it found that the association with Canadian
beef would stigmatize all U.S. meat, causing a “serious, irrep-
arable impact on ranchers in the U.S. and the U.S. economy.”
Id. Finally, the district court found that the NEPA violation,
in and of itself, would cause irreparable harm and warranted
preliminary injunctive relief. Id.
In light of its determination that R-CALF was likely to suc-
ceed on the merits, and that the balance of hardships tipped
in R-CALF’s favor, the district court issued a preliminary
injunction barring implementation of the Final Rule. Id. at
1074. Two weeks later, USDA filed this timely appeal.
RANCHERS CATTLEMEN v. USDA 8751
II. ANALYSIS
“A district court’s order granting a preliminary injunction
is subject to limited review.” Price v. City of Stockton, 390
F.3d 1105, 1109 (9th Cir. 2004). We will reverse “only where
the district court abused its discretion or based its decision on
an erroneous legal standard or on clearly erroneous findings
of fact.” Id. A reviewing court should generally refrain from
reviewing “the underlying merits of the case.” Southwest
Voter Registration Educ. Project v. Shelley, 344 F.3d 914,
918 (9th Cir. 2003) (en banc). Rather, “[a]s long as the district
court got the law right, it will not be reversed simply because
the appellate court would have arrived at a different result if
it had applied the law to the facts of the case.” Earth Island
Inst. v. United States Forest Serv., 351 F.3d 1291, 1298 (9th
Cir. 2003).
“The standard for granting a preliminary injunction bal-
ances the plaintiff’s likelihood of success against the relative
hardship to the parties.” Clear Channel Outdoor, Inc. v. City
of Los Angeles, 340 F.3d 810, 813 (9th Cir. 2003). This circuit
has recognized two different sets of criteria for preliminary
injunctive relief. Under the traditional test, a plaintiff must
show: “(1) a strong likelihood of success on the merits, (2) the
possibility of irreparable injury to plaintiff if preliminary
relief is not granted, (3) a balance of hardships favoring the
plaintiff, and (4) advancement of the public interest (in certain
cases).” Save Our Sonoran, Inc. v. Flowers, 408 F.3d 1113,
1120 (9th Cir. 2005). The alternative test requires that a plain-
tiff demonstrate “either a combination of probable success on
the merits and the possibility of irreparable injury or that seri-
ous questions are raised and the balance of hardships tips
sharply in his favor.” Id. (emphasis in original). “These two
formulations represent two points on a sliding scale in which
the required degree of irreparable harm increases as the prob-
ability of success decreases. They are not separate tests but
rather outer reaches of a single continuum.” Id.
8752 RANCHERS CATTLEMEN v. USDA
As we conclude below, the district court’s finding that R-
CALF had a strong likelihood of success on the merits was
premised on legal error. Further, we disagree with the district
court’s assessment of the irreparable harm threatened by the
Final Rule. Thus, we hold that a preliminary injunction was
unwarranted in this case.
A. Likelihood of Success on the Merits
The district court identified three distinct grounds for its
finding that R-CALF had a strong likelihood of success on the
merits: (1) that the Final Rule was arbitrary and capricious
under the APA; (2) that USDA had failed to satisfy NEPA’s
procedural requirements; and (3) that USDA had failed ade-
quately to consider the Final Rule’s effect on small busi-
nesses, as required by the RFA. None of these grounds
withstands scrutiny.
1. Administrative Procedure Act
The APA provides that a court, when reviewing agency
action, shall “hold unlawful and set aside agency action, find-
ings, and conclusions found to be . . . arbitrary, capricious, an
abuse of discretion, or otherwise not in accordance with law.”
5 U.S.C. § 706. An agency’s action violates this standard if
the agency has relied on factors which Congress has
not intended it to consider, entirely failed to consider
an important aspect of the problem, offered an expla-
nation for its decision that runs counter to the evi-
dence before the agency, or is so implausible that it
could not be ascribed to a difference in view or the
product of agency expertise.
City of Sausalito v. O’Neill, 386 F.3d 1186, 1206 (9th Cir.
2004) (quoting Motor Vehicles Mfrs. Ass’n v. State Farm Mut.
Auto Ins. Co., 463 U.S. 29, 43 (1983)).
RANCHERS CATTLEMEN v. USDA 8753
[1] Regulations are presumed to be valid, and therefore
review is deferential to the agency. Nat’l Ass’n of Home
Builders v. Norton, 340 F.3d 835, 841 (9th Cir. 2003). All that
is required is that the agency have “considered the relevant
factors and articulated a rational connection between the facts
found and the choices made.” Id. Further, “[t]he court is not
empowered to substitute its judgment for that of the agency.”
Ariz. Cattle Growers’ Ass’n v. United States Fish & Wildlife
Serv., 273 F.3d 1229, 1236 (9th Cir. 2001). Deference to the
informed discretion of the responsible federal agencies is
especially appropriate, where, as here, the agency’s decision
involves a high level of technical expertise. Id.
While review is therefore deferential, it is not toothless;
courts must conduct a “thorough, probing, in-depth” inquiry
into the validity of regulations. Nat’l Ass’n of Homebuilders,
340 F.3d at 841. This inquiry must be “searching and careful”
to ensure that the agency decision does not contain a clear
error of judgment. City of Sausalito, 386 F.3d at 1206; Nat’l
Ass’n of Homebuilders, 340 F.3d at 841. In performing this
inquiry, the court is not allowed to uphold a regulation on
grounds other than those relied on by the agency. Ariz. Cattle
Growers’ Ass’n, 273 F.3d at 1236 (“The reviewing court may
not substitute reasons for agency action that are not in the
record.”).
[2] The district court failed to abide by this deferential stan-
dard. Instead, the district court committed legal error by fail-
ing to respect the agency’s judgment and expertise. Rather
than evaluating the Final Rule to determine if USDA had a
basis for its conclusions, the district court repeatedly substi-
tuted its judgment for the agency’s, disagreeing with USDA’s
determinations even though they had a sound basis in the
administrative record, and accepting the scientific judgments
of R-CALF’s experts over those of the agency. For example,
in assessing the prevalence of BSE in the Canadian herd, the
district court rejected USDA’s calculation and accepted the
8754 RANCHERS CATTLEMEN v. USDA
prevalence rate provided by R-CALF’s expert, completely
without explanation. R-CALF I, 359 F. Supp. 2d at 1066.
The district court’s lack of deference may be attributable to
its misreading of the Animal Health Protection Act
(“AHPA”), 7 U.S.C. § 8301 et seq., the statute under which
the Final Rule was promulgated. Based on the AHPA’s state-
ment of congressional findings, 7 U.S.C. § 8301, the district
court appears to have imposed a requirement on USDA that
its Final Rule present no additional risk to human or animal
health.14 See R-CALF I, 359 F. Supp. 2d at 1065 (“The
[AHPA] directs the Secretary of the USDA to protect the
health and welfare of the people of the United States.”). The
AHPA is, in fact, based upon congressional findings that “the
prevention, detection, control, and eradication of diseases and
pests of animals are essential to protect . . . animal health
[and] the health and welfare of the people of the United
States.” 7 U.S.C. § 8301(1). The provision of the Act under
which the Final Rule was promulgated, however, states only
that “the Secretary [of Agriculture] may prohibit or restrict
. . . the importation or entry of any animal, article, or means
of conveyance . . . if the Secretary determines that the prohibi-
tion or restriction is necessary to prevent the introduction into
14
While the district court never explicitly stated that it was imposing
such a “zero-risk” requirement, its reasoning suggests that it did. For
example, the court faulted the agency for “presum[ing] that there is no risk
of exposure to BSE infective agents once an SRM removal requirement
is in place.” R-CALF I, 359 F. Supp. 2d at 1068. Similarly, the court found
the Final Rule arbitrary and capricious because the agency refused to act
to remove the “small probability” that BSE could be transmitted from a
pregnant Canadian cow to its offspring. Id. at 1069.
Indeed, the district court appears to have required USDA to disprove all
scientific uncertainty associated with BSE. It noted, for example, that
there is no “conclusive scientific proof” that cattle feed is the only method
of BSE transmission. Id. at 1066. In other areas of the opinion, any level
of scientific uncertainty surrounding a USDA decision rendered that deci-
sion an “assumption.” E.g., id. at 1066, 1067, 1068; see also id. at 1074
(criticizing USDA for acting despite “uncertainties in the agency’s knowl-
edge of the possible impacts on human and animal health”).
RANCHERS CATTLEMEN v. USDA 8755
or dissemination within the United States of any pest or dis-
ease of livestock.” 7 U.S.C. § 8303(a)(1).
[3] The AHPA was only recently enacted, in 2002, and, as
of yet, there are few reported cases interpreting its provisions.
Nonetheless, the statute’s terms indicate a congressional
intent to give the Secretary wide discretion in dealing with the
importation of plant and animal products. More to the point,
the AHPA does not impose any requirement on USDA that all
of its actions carry no associated increased risk of disease.
Indeed, the statute’s use of the word “may” suggests that the
Secretary is given discretion over such decisions as whether
to close the borders. See, e.g., United States v. George, 85
F.3d 1433, 1437 (9th Cir. 1996) (statute’s use of term “may”
“indicates that we should review a district court’s decision . . .
for abuse of discretion”). Although sparse, the AHPA’s legis-
lative history also supports this view. See H.R. Conf. Rep.
107-424, reprinted in 2002 U.S.C.C.A.N. 141, 388 (in order
to best protect against animal disease, “a regulatory definition
of disease should be left to the discretion of the Secretary,”
which will allow “the agency to have maximum flexibility to
focus its resources and respond to new or emerging disease
threats”). It is also notable that open borders are a default
under the AHPA, and the Secretary can close them only if
“necessary” to prevent livestock disease. See 7 U.S.C. § 8303.
[4] The structure of the AHPA is therefore inconsistent
with the district court’s strict requirement that the USDA reg-
ulation remove all risk of BSE entering the United States.
Because the district court interpreted the statute to contain
such a requirement, its analysis of the Final Rule’s compli-
ance with the APA was fundamentally flawed.15
15
We are in no way, of course, implying that the Secretary has unlimited
powers to open and close the borders as he sees fit. As the AHPA’s struc-
ture indicates, however, the Secretary has considerable discretion to
decide when an open or closed border is appropriate. Absent a strong
showing that the Secretary is not exercising that discretion consistent with
the statutory requirements, his judgment should not be overturned.
8756 RANCHERS CATTLEMEN v. USDA
Our own review of the Final Rule leads us to conclude that
the Secretary had a firm basis for determining that the
resumption of ruminant imports from Canada would not sig-
nificantly increase the risk of BSE to the American popula-
tion. In conducting this review, we believe it is appropriate to
view the BSE prevention measures currently in place as part
of a comprehensive system. Thus, rather than follow the “di-
vide and conquer” strategy of analyzing each protective com-
ponent of the regulatory system in isolation, we evaluate the
cumulative effects of the multiple, interlocking safeguards.
USDA’s comprehensive protections begin, first, with the
low incidence of BSE in Canadian cattle. This assures that if
any infected cattle are imported, the number will be relatively
small.16 Next, Canada’s feed ban, which USDA considers
effective, and its import restrictions on cattle from areas with
high BSE rates, ensure that Canada’s prevalence rate will not
rise dramatically. Canada also takes other measures, such as
BSE testing and epidemiological investigations, that help it
find and understand the source of BSE in its cattle population,
which helps it further minimize the prevalence of BSE in its
herd. These steps ensure, as USDA found, that Canada’s
already low rate of BSE is decreasing. Final Rule, 70 Fed.
Reg. at 464.
From the already low prevalence rate in the Canadian herd
as a whole, USDA permits the importation of only a subset of
those animals that are extremely unlikely to have BSE —
those under 30 months of age. In England, only 0.01 percent
of those animals diagnosed with BSE were under 30 months
of age. Meat Produced by Advanced Meat/Bone Separation
Machinery and Meat Recovery (AMR) Systems, 69 Fed. Reg.
1874, 1875 (Jan. 12, 2004). In addition, USDA’s scientific
evidence suggests that Canadian cattle under 30 months of
16
For example, assuming two million cattle enter the United States from
Canada per year, less than one would be expected to have BSE based on
Canada’s prevalence rate of 0.4 cases per million head of adult cattle.
RANCHERS CATTLEMEN v. USDA 8757
age will be far less likely to be in the advanced stages of BSE,
given that the incubation period of BSE depends on the
amount of BSE agent to which an animal has been exposed.
Based on Canada’s low BSE rate and its feed ban, Canadian
cattle should have a much lower exposure than English cattle,
resulting in a correspondingly greater incubation period.
Thus, the age restriction further reduces the risk of introduc-
tion of BSE from Canada’s herd.
Inside the United States, the risk of dissemination of BSE
is addressed by the requirement that Canadian cattle be imme-
diately slaughtered or fed and then slaughtered before they
reach the age of 30 months. Again, because of BSE’s lengthy
incubation period, this age limit helps to ensure that BSE will
not progress in any infected animals before they are slaugh-
tered. Once they are slaughtered, the FDA’s feed ban ensures
that they will not be fed to other cattle, preventing further dis-
semination of the disease if, in fact, an imported cow were
infected.
As for human health, cattle slaughtered in the United States
are subject to FSIS regulations designed to minimize the risk
that any infectious material will enter the human food supply.
These regulations largely prohibit parts of the central nervous
system and other cattle parts that have shown BSE infectivity
from contaminating human food. In addition, FSIS has placed
restrictions on the manner in which cattle may be slaughtered
— air compression devices are banned to protect against the
possibility that they might inject parts of the brain into the
bloodstream. FSIS regulations also require the removal of all
SRMs from slaughtered cattle, and they restrict the use in
human food of “mechanically separated beef” and meat
obtained from “Advanced Meat Recovery” systems.
The final defense against human BSE infection is biologi-
cal. The limited nature of the vCJD outbreak indicates that
there may be a substantial species barrier that prevents BSE
from easily infecting humans. Indeed, the fact that there have
8758 RANCHERS CATTLEMEN v. USDA
been only slightly over 150 confirmed cases of vCJD world-
wide — orders of magnitude less than the number of cases of
BSE in cattle — suggests that humans likely do not contract
the disease easily.
[5] This regulatory system, with its numerous overlapping
and complementary safeguards, is designed to minimize the
risk of BSE to American livestock and consumers. Thus, sub-
stantial evidence supports USDA’s conclusion that these pro-
tections will effectively achieve that goal. Further, a
comprehensive study commissioned by USDA, known as the
“Harvard-Tuskegee Study,” evaluated the likely effects of the
introduction of BSE into the United States. The study con-
cluded that, if 10 infected cows were imported into the United
States from Canada, on average only three new cases of BSE
would result and the disease was “virtually certain” to be
eradicated from the United States within 20 years.
Instead of evaluating the BSE safeguards as part of a larger
system, the district court parsed the regulations and faulted
USDA for any risk that a given step failed to remove. The dis-
trict court listed six specific grounds as the bases for its find-
ing that the Final Rule was arbitrary and capricious. We
examine each of them seriatim and conclude that none of
them supports its conclusion.
a. Lack of quantitative standards
The district court faulted USDA for “ma[king] assumptions
of qualitative judgments,” rather than performing “a quantita-
tive assessment of the risk of various options.” R-CALF I, 359
F. Supp. 2d at 1065. It concluded that, “[p]resented with the
USDA’s conclusions that the risks to U.S. cattle and consum-
ers are ‘low’ without any definition as to what that means and
why the risks presented by the Final Rule are acceptable, this
Court has no way of assessing the merits of the USDA’s
actions.” Id.
RANCHERS CATTLEMEN v. USDA 8759
[6] The district court’s imposition of such a bright-line pro-
hibition on qualitative standards was incorrect. The Supreme
Court has made clear that courts should not upset agency
decisions, even those announced with “less than ideal clarity,”
if “the agency’s path may reasonably be discerned.” Alaska
Dep’t of Envtl. Conservation v. EPA, 540 U.S. 461, 496
(2004) (internal quotation marks omitted); see also Vigil v.
Leavitt, 381 F.3d 826, 833 (9th Cir. 2004); Nat’l Ass’n of
Homebuilders, 340 F.3d at 846. Moreover, the AHPA does
not require the Secretary to quantify a permissible level of
risk or to conduct a risk assessment.
[7] Under this standard, the administrative record is an ade-
quate basis for discerning USDA’s conclusions. For example,
USDA’s conclusion that the prevalence of BSE in the Cana-
dian herd is “very low” is supported by its observation that
“Canada’s incidence rate of two infected cattle in 2003 out of
a population of 5.5 million cattle over 24 months of age [is
well below] OIE’s recommendation of less than two infected
cattle per million during each of the last four consecutive 12-
month periods within the cattle population over 24 months of
age.” Final Rule, 70 Fed. Reg. at 464. Similarly, the “very
low” risk of a consumer contracting vCJD is supported by its
finding that “the removal of SRMs effectively mitigates the
BSE risk to humans.” Id. at 465. Indeed, the Harvard-
Tuskegee Study, one of the centerpieces of USDA’s rulemak-
ing, concluded that SRM removal “would reduce . . . potential
human exposure to BSE by 95 percent.” Id. at 467.
The low risk of a human developing vCJD is also supported
by USDA’s observation that “the number of cases of vCJD
identified to date suggest a substantial species barrier that
may protect humans from widespread illness due to BSE.” Id.
at 462. It is also supported by anecdotal evidence of vCJD
outbreaks in other parts of the world. In Switzerland, for
example, the BSE rate in 1995 was 73.6 cases per million
head of cattle, and has been above 20 for most of the past 10
years, see http://www.oie.int/eng/info/en_esbincidence.htm,
8760 RANCHERS CATTLEMEN v. USDA
yet Switzerland has not identified a single case of vCJD.
Finally, no case of vCJD has ever been attributed to Canadian
beef or to the North American meat supply.
b. Prevalence of BSE in Canada
The district court concluded that “Canada has not con-
ducted sufficient testing for BSE to accurately assess the rate
of BSE infection in Canada.” R-CALF I, 359 F. Supp. 2d at
1065. It also concluded that the actual rate of BSE in Canada
was “greater than 5.5 cases per million head of cattle . . . [put-
ting] Canada on par with a number of European countries
with a BSE problem.” Id. at 1066. Based on this number, the
district judge found that the importation of “2-3 million head
of cattle from Canada during the remainder of 2005” pre-
sented a “potentially catastrophic risk of danger to the beef
consumers in the U.S.” Id.
[8] The district court, in this instance, impermissibly substi-
tuted its judgment for that of the agency. The USDA, in its
Final Rule, calculated Canada’s BSE prevalence rate to be
between 0.3 and 0.4 per million head of cattle. Final Rule, 70
Fed. Reg. at 464. The district court gave no reason for depart-
ing from this calculation and, instead, adopting the calculation
of R-CALF’s expert wholesale. The district court did so even
though R-CALF’s calculation contained the same type of
unexplained assumptions that the court found fatal to the
Final Rule. For example, R-CALF’s expert assumed that cat-
tle with outward signs of BSE are 60 times more likely to
have the disease than cattle with no symptoms, and assumed
that the prevalence rate of BSE in Alberta was representative
of the rate in Canada as a whole.
[9] USDA, on the other hand, based its calculation of Cana-
da’s BSE rate on OIE guidelines; indeed, the OIE website
lists Canada’s 2003 incidence rate as 0.33 and its 2004 rate as
0.149. See http://www.oie.int/eng/info/en_esbincidence.htm.
The district court erred by departing from USDA’s method of
RANCHERS CATTLEMEN v. USDA 8761
calculation, which was supported by the administrative
record, without providing any reason. Marsh v. Or. Nat. Res.
Council, 490 U.S. 360, 378 (1989) (“When specialists express
conflicting views, an agency must have discretion to rely on
the reasonable opinions of its own qualified experts even if,
as an original matter, a court might find contrary views more
persuasive.”); Nat’l Wildlife Fed’n v. United States Army
Corps of Eng’rs, 384 F.3d 1163, 1177 (9th Cir. 2004).
c. Effectiveness of Canadian feed ban
The district court also questioned USDA’s reliance on the
Canadian feed ban. First, it found that there was “no conclu-
sive scientific proof” that consumption of infected feed is the
only method of BSE transmission, commenting that transmis-
sion may occur through blood and saliva. R-CALF I, 359 F.
Supp. 2d at 1066-67. Second, the court found evidence that
the feed ban had not been effective, both because the ban had
only been in place for seven years and because the 4.2-year
average incubation period of BSE suggested that the infected
Canadian cows had contracted BSE well after the feed ban
was put in place. Id. at 1067. Finally, the court found gaps in
the ban, finding that both bovine blood and rendered animal
fat were allowed in animal feed and that both could transmit
BSE. Id. at 1067-68.
As to the first reason, the USDA explicitly considered sci-
entific evidence on alternative theories of transmission and
rejected them, finding that “oral ingestion of feed contami-
nated with the BSE is the only documented route of field
transmission of the disease.” Final Rule, 70 Fed. Reg. at 486;
see also id. at 491 (discussing infectivity of blood).
[10] The trial court’s criticisms of Canada’s feed ban are
also baseless. The district court’s main criticism is that Cana-
da’s feed ban had been in place for only seven and a half
years, not the eight years recommended by OIE guidelines.
Applying such a strict reading of OIE guidelines, however,
8762 RANCHERS CATTLEMEN v. USDA
was incorrect. According to a declaration submitted by the
Head of the International Trade Department of OIE, OIE rec-
ommends that an importing country evaluate the exporting
country’s risk mitigation measures as a whole, and “would not
consider it appropriate for the importing country to apply each
criterion as an item on a checklist.” Thus, “a deficiency in the
length of time a feed ban has been effectively applied could
be addressed through restrictions on the age of live cattle
imported.” The Final Rule reveals that this is precisely the
approach that USDA took. See, e.g., Final Rule, 70 Fed. Reg.
at 463 (discussing multiple criteria used to evaluate a poten-
tial minimal-risk region); id. at 548 (restricting imports of
Canadian cattle to those under 30 months of age).
Nor do we agree that the 4.2-year average incubation
period demonstrates the ineffectiveness of Canada’s feed ban.
USDA explained that the incubation period of BSE in cattle
depends upon the level of exposure the cattle have to the BSE
agent. The 4.2-year figure was obtained from analyzing cattle
during the BSE epidemic in England, which represents the
highest level of exposure to BSE in history. Cows in Canada
can be expected to have a longer incubation period because of
their significantly lower levels of BSE exposure.
[11] Finally, the district court also erred in criticizing the
Canadian feed ban based on its “gaps,” which allow blood
and rendered animal fat in cattle feed. As discussed above,
USDA considered BSE transmission through blood and deter-
mined that the science did not support ingestion of blood as
a means of transmission. Id. at 491. USDA also considered
transmission through fat and concluded that, provided the fat
is not impure, it poses no risk of transmission of BSE. Id. at
500-01 (discussing potential transmission of BSE through tal-
low). Again, the district court gave no reason for rejecting
USDA’s expert scientific opinion.
d. Effectiveness of SRM removal
[12] The district court also found that R-CALF had pres-
ented sufficient evidence to establish that “it is no longer rea-
RANCHERS CATTLEMEN v. USDA 8763
sonable to presume that there is no risk of exposure to BSE
infectious agents once an SRM removal requirement is in
place.” R-CALF I, 359 F. Supp. 2d at 1068. USDA’s conclu-
sion that SRM removal is effective, however, had support in
the administrative record. See Final Rule, 70 Fed. Reg. at 467
(discussing Harvard-Tuskegee Study, which concluded that
SRM removal would reduce human exposure to BSE by 95
percent).
e. Maternal transmission of BSE and fetal blood serum
The district court also found the Final Rule arbitrary and
capricious because it “does not prohibit cattle of breeding age
from being bred either before or after entering the U.S.,” and
“there is a small probability that BSE can be transmitted
maternally.” R-CALF I, 359 F. Supp. 2d at 1069. In addition,
the court found USDA’s prohibition of fetal blood serum to
be inconsistent with the possibility of allowing pregnant cows
to be imported into the United States. Id.
[13] Contrary to the district court’s findings, however,
USDA has made it abundantly clear that cattle may not be
imported for breeding under the new regulations. Instead, they
must be immediately slaughtered, or fed and slaughtered
before they reach 30 months of age. Final Rule, 70 Fed. Reg.
at 548-49. Furthermore, USDA discussed the concerns that
the district court raised, and found that they were not suffi-
cient to justify addressing. Id. at 515 (“Although some evi-
dence suggesting maternal transmission exists, such
transmission has not been proven, and, if it occurs at all, it
occurs at very low levels not sufficient to sustain an epidem-
ic.”).
We also find that there is a basis for USDA’s disparate
treatment of fetal blood serum. As the district court acknowl-
edged, fetal blood serum is used for “bovine vaccine produc-
tion” and “bovine embryo transfer.” R-CALF I, 359 F. Supp.
2d at 1069. Because the serum is injected directly into an ani-
8764 RANCHERS CATTLEMEN v. USDA
mal’s bloodstream, it carries a higher risk of transmitting
BSE, and “might pose a risk of livestock if used in” these
applications. Final Rule, 70 Fed. Reg. at 502. Thus, any
inconsistency in the USDA’s approach to offspring of
imported Canadian cattle and fetal blood serum has an ade-
quate explanation in the record.
f. Mandatory testing of Canadian cattle
Finally, the district court held that it was arbitrary and
capricious for the agency not to require all Canadian cattle to
be screened for BSE, because the screening test could identify
some animals with BSE that would not otherwise be identi-
fied. R-CALF I, 359 F. Supp. 2d at 1069. The Final Rule,
however, contains a lengthy comment in which USDA
responded to requests for testing of Canadian cattle. Final
Rule, 70 Fed. Reg. at 475-76. USDA explained that, because
testing can only detect the disease two to three months before
a cow starts demonstrating clinical signs of the disease, a cow
may be infected and thus produce a false negative on a test.
Id. Because of the long incubation period of BSE, and the rel-
atively short window in which non-targeted testing is effec-
tive, the USDA did not consider testing to be a “food safety”
measure. Id. Rather, testing was best used to determine if BSE
exists in a country and to determine its prevalence — goals
that can both be achieved by targeted testing of animals with
clinical signs of BSE. Id.
[14] Over the past few years, USDA’s policies regarding
BSE testing have been subject to a high degree of criticism.
See, e.g., Mad Beef Policy, Los Angeles Times, Jul. 1, 2005;
McGarity, supra, at 337-40. These criticisms have generally
focused on USDA’s refusal to allow voluntary testing of cat-
tle, rather than its refusal to require mandatory testing of
Canadian cattle. Although these criticisms are not without
their valid points, we do not believe that they are so powerful
as to render USDA’s testing policy invalid. USDA’s approach
to BSE testing — that, until better tests are developed, pro-
RANCHERS CATTLEMEN v. USDA 8765
phylactic measures such as the feed ban and SRM removal are
the best methods of protecting human and animal health — is
defensible. While its wisdom may be subject to debate on the
merits, its choices are not so lacking support in the adminis-
trative record as to be “arbitrary and capricious.”
g. Conclusion
In sum, USDA decided to reopen the border to Canadian
ruminants after making a reasoned determination that the
importation of a small number of BSE-infected cattle into this
country would not pose a serious risk to humans or livestock.
As part of its determination, USDA necessarily decided that
the risks inherent in the uncertainty surrounding the current
scientific understanding of BSE were insufficiently significant
to justify the continued exclusion of Canadian cattle. Rather
than criticizing USDA for allowing these risks as a part of its
policy, the district court should have evaluated whether there
was an adequate basis in the administrative record for
USDA’s conclusion that the risks were acceptable.
[15] Our review of the record leads us to conclude that the
risks inherent in the Final Rule are small, and that the rule
likely is supported by an adequate administrative record. We
therefore conclude that the district court erred in finding that
R-CALF has a strong likelihood of success on the merits of
its APA claim.
2. Regulatory Flexibility Act
We also conclude that the district court erred in concluding
that R-CALF has a strong likelihood of success on its claim
under the RFA. The RFA was passed in 1980 to “encourage
administrative agencies to consider the potential impact of
nascent federal regulations on small businesses.” Assoc. Fish-
eries of Maine, Inc. v. Daley, 127 F.3d 104, 111 (1st Cir.
1997). In certain cases, it requires agencies to publish an “ini-
tial regulatory flexibility analysis” at the time a proposed rule
8766 RANCHERS CATTLEMEN v. USDA
is published, and a “final regulatory flexibility analysis” at the
time a final rule is published. 5 U.S.C. §§ 603, 604. Judicial
review is available only of the final analysis. 5 U.S.C. § 611.
The RFA requires that a final analysis contain the follow-
ing:
(1) a succinct statement of the need for, and objec-
tives of, the rule;
(2) a summary of the significant issues raised by the
public comments in response to the initial regulatory
flexibility analysis, a summary of the assessment of
the agency of such issues, and a statement of any
changes made in the proposed rule as a result of such
comments;
(3) a description of and an estimate of the number
of small entities to which the rule will apply or an
explanation of why no such estimate is available;
(4) a description of the projected reporting, record-
keeping and other compliance requirements of the
rule, including an estimate of the classes of small
entities which will be subject to the requirement and
the type of professional skills necessary for prepara-
tion of the report or record; and
(5) a description of the steps the agency has taken to
minimize the significant economic impact on small
entities consistent with the stated objectives of appli-
cable statutes, including a statement of the factual,
policy, and legal reasons for selecting the alternative
adopted in the final rule and why each one of the
other significant alternatives to the rule considered
by the agency which affect the impact on small enti-
ties was rejected.
RANCHERS CATTLEMEN v. USDA 8767
5 U.S.C. § 604(a).
The RFA imposes no substantive requirements on an
agency; rather, its requirements are “purely procedural” in
nature. United States Cellular Corp. v. FCC, 254 F.3d 78, 88
(D.C. Cir. 2001); see also Envtl. Defense Ctr., Inc. v. United
States EPA, 344 F.3d 832, 879 (9th Cir. 2003), cert. denied,
541 U.S. 1085 (2004) (“Like the Notice and Comment pro-
cess required in administrative rulemaking by the APA, the
analyses required by the RFA are essentially procedural hur-
dles; after considering the relevant impacts and alternatives,
an administrative agency remains free to regulate as it sees
fit.”). To satisfy the RFA, an agency must only demonstrate
a “reasonable, good-faith effort” to fulfill its requirements.
United States Cellular, 254 F.3d at 88; Alenco Communica-
tions, Inc. v. FCC, 201 F.3d 608, 625 (5th Cir. 2000); Assoc.
Fisheries, 127 F.3d at 114.
The district court faulted USDA for considering only two
alternatives in its final regulatory flexibility analysis: “leaving
the regulations unchanged or modifying the import require-
ments by either requiring that imported beef come from cattle
slaughtered at less than 30 months of age or continuing to
prohibit the entry of live ruminants.” R-CALF I, 359 F. Supp.
2d at 1072; see also Final Rule, 70 Fed. Reg. at 543. The dis-
trict court held that the agency erroneously rejected the alter-
natives of a country-of-origin labeling program and voluntary
testing of slaughtered Canadian cattle. R-CALF I, 359 F.
Supp. 2d at 1072.
[16] The district court erred in concluding that USDA did
not meet the RFA’s requirements. The Final Regulatory Flex-
ibility Analysis, available at http://www.aphis.usda.gov/lpa/
issues/bse/risk_assessment/03-080-3_econ_analysis.pdf,
reveals that USDA conducted a detailed economic assessment
of the impact of its proposed rule on small businesses. It con-
cluded that the majority of businesses affected by the pro-
posed Final Rule would qualify as small businesses, and that
8768 RANCHERS CATTLEMEN v. USDA
the effect of the Final Rule was likely to vary depending upon
the sector of the cattle industry the business occupied, rather
than the size of the business. The negative economic effects
the rule would create would generally affect those on the sup-
ply side of the beef industry — primarily ranchers — while
those on the production side — feedlots and meat packers —
would tend to benefit from the rule. In this respect, the alter-
natives identified by the district court would not necessarily
ease the burden on small businesses; rather, they would real-
locate the rule’s burden to small businesses in different sec-
tors of the beef industry. Cf. Assoc. Fisheries, 127 F.3d at 115
(where the majority of businesses affected by a rule are small
businesses, “Congress’s desire to have agencies write rules
that distinguish . . . between big and small businesses has
diminished relevance.”).
More importantly, the specific concerns the district court
raised were considered by USDA in its response to comments
on the rule. USDA rejected the first alternative — the imple-
mentation of a country-of-origin labeling program — because
it did not consider such a program to concern food safety or
animal health. Final Rule, 70 Fed. Reg. at 533. USDA
rejected the second alternative, voluntary BSE testing,
because it does not consider such testing reliable enough to be
used as a food safety measure, as discussed above. See Part
II.A.1.f, supra. Given that USDA discussed and rejected these
alternatives in the body of its Final Rule, the agency did not
err in failing to consider them as alternatives in its final regu-
latory flexibility analysis. See Assoc. Fisheries, 127 F.3d at
115 (“[S]ection 604 does not require that [a final regulatory
flexibility analysis] address every alternative, but only that it
address significant ones.”).
3. National Environmental Policy Act
NEPA was enacted in 1970 to “promote efforts which will
prevent or eliminate damage to the environment and bio-
sphere.” 42 U.S.C. § 4321; see also Robertson v. Methow Val-
RANCHERS CATTLEMEN v. USDA 8769
ley Citizens Council, 490 U.S. 332, 348 (1989) (“Section 101
of NEPA declares a broad national commitment to protecting
and promoting environmental quality.”). Like the RFA,
NEPA does not impose any substantive requirements on an
agency’s decision; rather, it mandates only a process that the
agency must follow. Id. at 350 (“NEPA itself does not man-
date particular results, but simply prescribes the necessary
process.”).
Under NEPA’s procedural requirements, an agency must
prepare a “detailed statement” on the environmental impact of
a proposed rule when that rule is a “major Federal action[ ]
significantly affecting the quality of the human environment.”
42 U.S.C. § 4332. NEPA provides no private right of action
to enforce its requirements. Stratford v. FAA, 285 F.3d 84, 88
(D.C. Cir. 2002). Thus, to bring suit to vindicate NEPA’s
requirements, a plaintiff must rely on the provisions of the
APA that confer “standing to an ‘aggrieved party’ within the
meaning of the substantive statute upon which the claim is
based.” Id.; see also 5 U.S.C. § 702; Clarke v. Sec. Indus.
Ass’n, 479 U.S. 388, 394-96 (1987).
[17] To narrow the wide range of potential plaintiffs who
may assert a “procedural injury” under this section of the
APA, the Supreme Court has adopted a “zone of interests” test.17
See id. at 397 n.12 (stating that the purpose of the zone of
17
R-CALF incorrectly argues that the Supreme Court’s decision in Ben-
nett v. Spear, 520 U.S. 154 (1997), drastically narrowed the applicability
of the zone of interests test. In Bennett, the Court considered the specific
question of standing under the Endangered Species Act’s citizen-suit pro-
vision, not the APA. Id. at 161-62. It expressly found that the “ESA’s
citizen-suit provision . . . negates the zone-of-interests test” based on its
language and its purpose. Id. at 164-66. Thus, Bennett simply does not
address actions under NEPA. Indeed, this court has continued to use the
zone of interests test to evaluate the standing of NEPA plaintiffs after Ben-
nett. See Save Our Sonoran, 408 F.3d at 1119; Cantrell v. City of Long
Beach, 241 F.3d 674, 679 (9th Cir. 2001); see also Stratford, 285 F.3d at
88 (applying the zone of interest test in a NEPA action).
8770 RANCHERS CATTLEMEN v. USDA
interests test is “to exclude those plaintiffs whose suits are
more likely to frustrate than to further statutory objectives”).
This test imposes the requirement, beyond constitutional
standing requirements, that a plaintiff assert an interest “argu-
ably within the zone of interests to be protected or regulated
by the statute or constitutional guarantee in question.” Nev.
Land Action Ass’n v. United States Forest Serv., 8 F.3d 713,
715-16 (9th Cir. 1993). Thus, to assert a claim under NEPA,
a plaintiff must allege injury to the environment; economic
injury will not suffice. Id. at 716 (“[A] plaintiff who asserts
purely economic injuries does not have standing to challenge
an agency action under NEPA.”); Stratford, 285 F.3d at 88
(“[A] NEPA claim may not be raised by a party with no
claimed or apparent environmental interest.”); W. Radio Serv.
Co. v. Espy, 79 F.3d 896, 902-03 (9th Cir. 1996) (“NEPA’s
purpose is to protect the environment, not the economic inter-
ests of those adversely affected by agency decisions.”) (inter-
nal quotation marks omitted). A plaintiff can, however, have
standing under NEPA even if his or her interest is primarily
economic, as long as he or she also alleges an environmental
interest or economic injuries that are “causally related to an
act within NEPA’s embrace.” Port of Astoria, Or. v. Hodel,
595 F.2d 467, 476 (9th Cir. 1979).
The injuries alleged in R-CALF’s complaint do not fall
within NEPA’s zone of interests. R-CALF points to only one
paragraph in its complaint to justify its standing under NEPA.
Every allegation in this paragraph, however, concerns the eco-
nomic interest of R-CALF members except the following: “R-
CALF USA members will also be adversely affected by the
increased risk of disease they face when Canadian beef enters
the U.S. meat supply.”
[18] We conclude that this alleged harm is insufficient to
fall within NEPA’s zone of interests. As mentioned above,
“NEPA’s purpose is to protect the environment.” W. Radio
Serv. Co., 79 F.3d at 902; see also Stratford, 285 F.3d at 88
(“[A] NEPA claim may not be raised by a party with no
RANCHERS CATTLEMEN v. USDA 8771
claimed or apparent environmental interest.”). More specifi-
cally, NEPA is concerned with harm to the physical environ-
ment: “If a harm does not have a sufficiently close connection
to the physical environment, NEPA does not apply.” Metro.
Edison Co. v. People Against Nuclear Energy, 460 U.S. 766,
778 (1983); cf. Cantrell, 241 F.3d at 679 (“In NEPA cases,
we have described this ‘concrete interest’ test as requiring a
‘geographic nexus’ between the individual asserting the claim
and the location suffering an environmental impact.”). R-
CALF’s claimed interest, however, has no connection to the
physical environment; rather, it is solely a matter of human
health. While it is true that NEPA contains references to
human health in its statement of policy, see 42 U.S.C. § 4321,
as the Supreme Court has explained, those references are to
the statute’s goals, not its means. Metro. Edison Co., 460 U.S.
at 773 (“[A]lthough NEPA states its goals in sweeping terms
of human health and welfare, those goals are the ends that
Congress has chosen to pursue by means of protecting the
physical environment.”). Here, R-CALF has failed to show
any relationship between risks to human health and environ-
mental harms. Cf. Port of Astoria, 595 F.2d at 476.
[19] Because R-CALF has failed to allege any connection
to injury to the physical environment, its injury falls outside
of NEPA’s zone of interests. Even assuming R-CALF’s
alleged injury could satisfy the zone of interests test, however,
its NEPA claim must fail for the additional reason that R-
CALF lacks organizational standing to assert a NEPA chal-
lenge.
An association has standing to bring suit on behalf
of its members when its members would otherwise
have standing to sue in their own right, the interests
at stake are germane to the organization’s purpose,
and neither the claim asserted nor the relief
requested requires the participation of individual
members in the lawsuit.
8772 RANCHERS CATTLEMEN v. USDA
Friends of the Earth, Inc. v. Laidlaw Envtl. Serv. (TOC), Inc.,
528 U.S. 167, 181 (2000) (citing Hunt v. Wash. State Apple
Advertising Comm’n, 432 U.S. 333, 343 (1977)). R-CALF
fails the second of these three elements.
As mentioned above, R-CALF is a “non-profit cattle asso-
ciation representing over 12,000 U.S. cattle producers on
issues concerning international trade and marketing.” As is
evident from the paragraph in its complaint that discusses
standing, economic issues are highly relevant to its purpose.
We do not see the connection, however, between the pur-
ported environmental interest that R-CALF attempts to raise
here and the “trade and marketing” interests it is organized to
protect.
[20] We therefore hold that R-CALF lacks standing to
bring a NEPA challenge to the Final Rule. Thus, the district
court erred in permitting R-CALF to proceed with its NEPA
claim and in concluding that it had a likelihood of success on
that claim.18
B. Balance of Hardships
After finding that R-CALF had demonstrated a strong like-
lihood of success on the merits, the district court found that
the Final Rule carried a definitive risk of causing “significant
irreparable harm.” R-CALF I, 359 F. Supp. 2d at 1073. The
district court identified three ways in which the Final Rule
would cause such harm: the increased risk of vCJD to Ameri-
can beef consumers, unspecified environmental injury stem-
ming from USDA’s failure to comply with NEPA, and injury
to the U.S. beef industry and the U.S. economy that would
result from a “stigma” that tainted Canadian beef would
inflict upon the U.S. meat supply. Id. We believe the district
18
Given our holding that R-CALF lacks standing to bring a NEPA
claim, we need not address the district court’s conclusion that the possibil-
ity of environmental harm justifies its preliminary injunction.
RANCHERS CATTLEMEN v. USDA 8773
court’s calculus overstated the harm that would result from
the rule.
[21] If the Canadian herd were to have a higher infection
rate than the U.S. herd, the importation of Canadian cattle
might pose some increased risk to the health of the U.S. popu-
lation, however slight. Even assuming, however, that the
introduction of a fatal disease into the United States would
constitute irreparable harm, cf. Harris v. Board of Supervi-
sors, 366 F.3d 754, 766 (9th Cir. 2004) (accepting as irrepara-
ble harm “pain, infection, amputation, medical complications,
and death”), the record does not justify the conclusion that the
Final Rule makes such harm likely, or even probable. Rather,
based on the low incidence of BSE in the Canadian herd, the
numerous safeguards against BSE in this country, the lack of
any Canadian cattle under 30 months of age found with BSE,
and the lack of any case of vCJD attributable to Canadian
beef, any increased risk to human and animal health created
by the Final Rule is negligible.
[22] In retrospect, the district court’s concern over the pos-
sibility of “stigma” harming the American beef industry
appears to be overstated. The record does not support the dis-
trict court’s alarmist findings that the “irreparable economic
harm” the district court foresaw from the stigma of Canadian
beef will actually befall the American beef industry. Follow-
ing the case of BSE diagnosed in a Washington State cow in
2003, consumer demand for, and confidence in, American
beef remained strong. Final Rule, 70 Fed. Reg. at 522.
According to USDA, American demand for beef in 2004 is
estimated to have increased seven to eight percent over 2003
levels. Yet, Canadian beef was flowing into this country
throughout 2004 under permits issued by USDA.19 This evi-
19
The district court’s April 26, 2004, temporary restraining order pre-
vented USDA only from expanding the categories of Canadian beef that
could be imported under existing importation permits. The court explicitly
limited its order to “all edible bovine meat products beyond those autho-
rized by USDA’s action of August 8, 2003 (boneless bovine meat, bone-
less Veal (meat), and bovine liver) from cattle under the age of 30
months.” R-CALF TRO, 2004 WL 1047837 at *9.
8774 RANCHERS CATTLEMEN v. USDA
dence belies the district court’s prediction of catastrophic
injury to the U.S. beef industry.20
C. Preliminary Injunction
[23] Contrary to the district court’s conclusion, we con-
clude that the Final Rule will likely survive judicial scrutiny
under the correct legal standard; thus, R-CALF has not shown
a likelihood of success on the merits of its action. We also
conclude that R-CALF has failed to make the requisite show-
ing of irreparable harm. For these reasons, we must reverse
the district court’s preliminary injunction. See Kootenai Tribe
v. Veneman, 313 F.3d 1094, 1125-26 (9th Cir. 2002).
III. CONCLUSION
For the foregoing reasons, the district court’s grant of a pre-
liminary injunction is
REVERSED.
20
Indeed, the district court’s finding of irreparable economic harm is
undermined by the industry itself. Numerous amici curiae briefs have
been filed in this case by organizations representing large sectors of the
American meat industry, all of whom seek reversal of the preliminary
injunction. If the Final Rule posed a true risk of exposing American beef
to an irreparable stigma one would not expect to see such a broad coalition
of industry members supporting its implementation.