The Free Trade Agreement Meets its First Challenge:
Dispute Settlement and the Pork Case
Andreas F. Lowenfeld*
The most
L’aspect le plus innovateur de l’Accord de
libre-ichange entre le Canada et les Etats-Unis
fut l’inclusion d’un processus de r6vision bila-
tdral des d6cisions des instances administrati-
yes des deux pays relativement aux pratiques
de commerce ill6gitimes. De fagon g6n6rale,
ce processus a bien fonctionn6; il aura entre
autres permis de d6politiser les conflits com-
merciaux, d’amoindrir les risques et les
craintes face aux <
rendre les organes administratifs plus respon-
sables et professionnels. Cependant, dans le
cas pr6cisdes exportations de porc du Canada
vers les Etats-Unis, le processus est devenu
tellement conflictuel que l’existence m~me de
l’Accord fut menac6e. L’auteur raconte le
d6roulement de l’affaire Pork, remontant, au
d6but de la proc6dure de contestation extraor-
dinaire, puis jusqu’au d6nouement subs6quent
de la controverse, qui ont permis Ia survie
du processus de rfglement des diff6rends et,
ultimement, peut-8tre de tout l’Accord de
libre-ichange.
innovative aspect of
the
Canada-U.S. Free Trade Agreement was the
provision for binational review of decisions of
each country’s administrative agencies on
issues of alleged unfair trade. On the whole
this process has worked well, depoliticizing
trade disputes, reducing the suspicion of
“home town justice,” and making the adminis-
trative agencies more accountable and profes-
sional. In one case, however, involving exports
of pork products from Canada to the United
States, the process became acrimonious, and
threatened to bring the whole Agreement
down. The author tells the story of the Pork
case, including initiation by the United States
of the Extraordinary Challenge Procedure, and
the subsequent d6nouement of the controversy,
making it possible for the dispute settlement
process –
and perhaps the Free Trade Agree-
ment as a whole –
to survive.
* Charles L. Dennison Professor of Law, School of Law, New York University. This paper is
adapted and updated from a much longer report prepared by the author for the Administrative Con-
ference of the United States on Binational Dispute Settlement under Chapters 18 and 19 of the
Canada-United States Free Trade Agreement (December 1990, April 1991). The opinions
expressed are solely those of the author.
McGill Law Journal 1992
Revue de droit de McGill
To be cited as: (1992) 37 McGill L.J. 597
Mode de citation: (1992) 37 R.D. McGill 597
McGILL LAW JOURNAL
[Vol. 37
Introduction
Synopsis
I.
H.
I.
Subsidy Phase: Round I
A. Upstream Subsidies and the “Pass-Through” Issue
B. Reading the Baucus Amendment
C. Measuring the Subsidy
D. Specificity
E. Deciding on the Evidence
F Raising New Arguments
Subsidy Phase: Round II
A. Specificity Again
B. From Corn to Hogs
C. Further Remand
Injury Phase: Round I
A. A Question of Numbers
B. A Question of Inferences
C. A Question of Logic
D. Summary of Round I
IV.
Injury Phase: Round H
A. Back before the Commission
B. Back before the Panel
1. Reopening the Record
2. Product Shifting
3. More on the Hog Cycle
V.
Injury Phase: Round M
VI. Some Reflections on the Clash between the Panel and the
Commission
A. Predictions, Predictions
B. The Pork Case and the FTA
VII. Extraordinary Challenge: Round IV
A. Raising the Challenge
B. The Decision of the Extraordinary Challenge Committee
Conclusion
1992]
DISPUTE SETTLEMENT & THE PORK CASE
Introduction
When the Canada-United States Free Trade Agreement’ entered into effect
in January 1989, there was much expectation and a good deal of apprehension
about its novel dispute settlement mechanism. Particularly Chapter 19 of the
agreement, concerning what Canadians call contingent protectionism and
Americans call defense against unfair trade, raised considerable doubts on both
sides of the border. The origin of Chapter 19 was, in essence, that the American
side, in agreeing to establish a binational area not separated by ordinary customs
duties, was not prepared to give up protection against goods made in Canada
and sold in the United States at unfairly low (“dumped”) prices and, even more
important, against goods receiving subsidies from the federal or provincial gov-
ernments of Canada. The Canadian side was not willing to permit determina-
tions of dumping and subsidies to be made unilaterally by American officials
and courts in whose priorities and predispositions they had less than complete
confidence. The unique solution reflected in Chapter 19 of the FTA was to find
a procedural way out of a controversy whose substantive elements remained –
appellate review
and continue to remain –
by binational panels of decisions of administrative agencies of each country –
has been widely described both in Canada and in the United States, and need
not be reviewed here. Overall, the process, in my judgment and in that of most
participants and observers on both sides of the border, has been quite successful.
Hard fought disputes have been resolved in much less time than would have
been taken if appeals had been submitted to the courts in each country; sloppy
findings by administrative agencies have been corrected; and the joint partici-
pation of public officials and private parties in a single international proceeding
has worked better than could have been expected. The opinions issued by the
panels have been thorough and intellectually rigorous. Most important, the par-
ticipants in the process, even when they did not prevail, were left with the
knowledge that their contentions had been fairly considered. Nearly all the
panel decisions were made unanimously: it did not matter whether a given panel
consisted of three American and two Canadian members or vice versa, and it did
not matter to the decision of the panels (putting aside questions of style)
whether the chairman was Canadian or American.
to be worked out. The solution –
To all of this there was one exception –
the case of Fresh, Chilled, and
Frozen Pork. The Pork case has received a great deal of publicity, especially in
Canada. But the issues were complicated, even for experts; the texts were
lengthy; the proceedings were confusing; and the mix of economics, politics,
and law are difficult to sort out. This paper makes the effort to tell the story step
by step.
The Pork case arose out of a claim by the National Pork Producers Council
of the United States that pork imported into the United States from Canada had
benefited from a variety of governmental assistance programs, that the United
122 December 1987, Can. T.S. 1989 No. 3, 27 I.L.M. 281 [hereinafter FTA] (Part A, Schedule
to the Canada-United States Free Trade Agreement Implementation Act, S.C. 1988, c. 65; entered
into force I January 1989).
REVUE DE DROIT DE McGILL
[Vol. 37
States pork industry had been injured as a result of these imports, and that a
countervailing duty should be imposed by the United States designed –
in the
modem clich6 –
to “level the playing field.” Under United States law, consistent
with the General Agreement on Tariffs and Trade,2 a countervailing duty pro-
ceeding has two separate phases. The existence and measure of subsidization are
determined by the International Trade Administration of the Department of Com-
merce, an agency of the executive branch; the existence of “material injury to an
industry” is determined by the U.S. International Trade Commission, a so-called
independent regulatory commission made up, when all positions are filled, of six
commissioners appointed for staggered six-year terms. For much of the time rel-
evant to the events here told, there were three vacancies on the Commission. The
Pork case, as detailed below, was fiercely contested both before the Commerce
Department and before the International Trade Commission, and two separate
panels under the FTA were convened to hear appeals from the decisions of the
respective agencies – more than once, as it turned out.
L
Subsidy Phase: Round 1V
In January 1989, the National Pork Producers Council, along with 13 state
pork producer associations, the National Pork Council Women, and 7 pork pro-
ducers, filed a petition with the International Trade Administration of the U.S.
Department of Commerce, alleging that producers and exporters of fresh,
chilled, and frozen pork in Canada received subsidies within the meaning of
U.S. trade law. The Commerce Department excluded some 33 programs recited
in the petition, but commenced an investigation of one federal, two joint federal/
provincial, and 36 provincial programs alleged to bestow countervailable sub-
sidies on Canadian pork producers.4 In fact there had been an earlier investiga-
tion into allegations of subsidies by Canadian and provincial governments with
respect to both live swine and pork, and the Department of Commerce had pre-
viously found that both producers of live swine and producers of pork in Canada
received countervailable subsidies.’ But in that case the bulk of the subsidies
found to be countervailable had been paid to producers of live swine, and on
appeal the U.S. Court of International Trade had found that the Department had
failed to apply the “upstream subsidy” provision of the U.S. law, Tariff Act of
1930 section 771A,6 and had remanded the proceeding to the Department.
Because in that case the International Trade Commission had not found injury,8
230 October 1947, Can. T.S. 1947 No. 27,55 U.N.T.S. 187, T.I.A.S. No. 1700, 1st supp. B.I.S.D.
(1953) 6 [hereinafter GATI].
3Fresh, Chilled and Frozen Pork from Canada (1990), 3 T.C.T. 8308 (Ch. 19 Panel), summa-
rized in 55 Fed. Reg. 41369 (1990) [cited to T.C.T.].
4Dept. of Commerce, Initiation of Countervailing Duty Investigation; Fresh, Chilled, and Fro-
5Dept. of Commerce, Live Swine and Fresh, Chilled, and Frozen Pork Products froin Canada,
zen Pork from Canada, 54 Fed. Reg. 5537 (1989).
50 Fed. Reg. 25097 (1985).
619 U.S.C. 1677-1, adopted in 626(a) of Pub. L. 98-573, eff. 30 October 1984.
7Canadian Meat Council v. United States, 661 F. Supp. 622 at 625-29 (Ct. Int’l Trade 1987).
8Live Swine and Pork from Canada, Inv. No. 701-TA-224 (Final), USITC Pub. 1733 (July 1985)
[hereinafter Live Swine], affirmed sub nom. National Pork Producers Council v. United States, 661
F. Supp. 633 (Ct. Int’l Trade 1987).
1992]
DISPUTE SETTLEMENT & THE PORK CASE
that investigation had become moot. Meanwhile Congress had adopted a new
provision, section 771B,9 and this provision provided the focus for the new
petition.
Once the International Trade Commission had made a preliminary determi-
nation that there was a “reasonable indication” that an industry in the United
States was materially injured or threatened with material injury,'” the Commerce
Department went through the required stages of determining whether the chal-
lenged products were benefiting from countervailable subsidies, including sub-
mission of a questionnaire to the government of Canada, receipt of preliminary
and supplemental responses from Ottawa as well as from the governments of
Alberta, Ontario, Manitoba, Quebec, and Saskatchewan, receipt of briefs by
petitioner and respondents, and conduct of a public hearing. On 24 July 1989,
the Department of Commerce published its final countervailing duty determina-
tion, holding 18 separate programs, some provincial, some federal, and some
mixed, to be countervailable subsidies, to a total amount of C$ 0.08 per kilo-
gram, or C$ 0.036 per pound.” The Canadian parties asked for review by a
panel under Chapter 19 of the FTA, challenging 7 of the determinations by the
Department. Not only the Canadian Meat Council and several Canadian meat
packers, but (for the first time) the government of Canada, as well as the gov-
ernments of Alberta, Quebec, and Ontario, appeared before the Panel. Since,
under article 1904 of the FTA, panel review is to be conducted under the law
of the importing country and according to the standards applicable in that coun-
try for judicial review of decisions of the agency in question, all the Canadian
parties, as well as the Department of Commerce and the U.S. petitioners,
appeared by American counsel. The principal issue was how to deal with sub-
sidies not given directly to the imported product – pork – but to a prior stage
(one can hardly say ingredient) of that product –
live swine. The Pork case rep-
resented the first opportunity to apply and interpret the 1988 amendment2 to a
1984 “upstream subsidy” amendment to the U.S. countervailing duty law,’3
which, as noted, the Department had not applied to the prior case involving the
same parties.’4
A. Upstream Subsidies and the “Pass-Through” Issue
“Upstream subsidies” is a relatively new term in international trade
designed to address the situation where the exporter of a processed or manufac-
tured product does not itself receive a subsidy but benefits from a subsidy to an
important input through cost savings not available to the domestic competitor.
For example, one of the complaints that led to the adoption of the 1984 amend-
ment was the claim that Mexican exporters of fertilizers were benefiting from
919 U.S.C. 1677-2, adopted in 1313(a) of Pub. L. 100-418, eff. 23 August 1988.
1019 U.S.C. 1671b(a)(1).
“Fresh, Chilled and Frozen Pork from Canada, 54 Fed. Reg. 30774 (1989). Eleven other
programs challenged by the National Pork Producers Council had been found not countervail-
able.
‘ 2Supra, note 9.
13Supra, note 6.
14See supra, note 7 and accompanying text.
McGILL LAW JOURNAL
[Vol. 37
below-market supplies of natural gas, and thus gaining an unfair advantage over
U.S. producers of fertilizers.’ 5
Section 771B, the 1988 amendment applied by the Commerce Department
in its finding of subsidy, reads as follows:
In the case of an agricultural product processed from a raw agricultural product in
which –
(1) the demand for the prior stage product is substantially dependent on the
demand for the latter stage product, and
(2) the processing operation adds only limited value to the raw commodity,
subsidies found to be provided to either producers or processors of the product
shall be deemed to be provided with respect to the manufacture, production,
or exportation of the processed product. 6
Senator Baucus, in introducing the amendment, had said that “pork is just
a very mature hog,”‘ 7 the point being that both of the conditions of the amend-
ment were easily satisfied. Before the Panel, however, the Canadian parties
argued that under the GATT and the overall U.S. countervailing duty law, coun-
tervailing duties may be enforced on a product only to the extent that it can be
shown that a subsidy had been received “on the production or export of such
product.”‘ 8 The Canadian parties argued that the record failed to show that pork
producers had received any portion of the subsidy found to be provided to hog
growers. Commerce argued that it was required to follow the statute, and denied
that it was acting contrary to the GA7T.
evidently for Canadian readers –
Thus Pork was the first case in which a panel established under the FTA
had to construe a U.S. statute that had not been construed previously. The Panel
explained –
that in the United States it is the
practice to consider legislative history, as well as to accord deference to the
interpretation given to a statute by the agency charged with its interpretation.
Moreover, while an act of Congress is supposed to be construed, when fairly
possible, so as not to conflict with international law, when a statute and an ear-
lier rule of international law or agreement cannot be fairly reconciled, the statute
prevails. 9 After considering the question of GATT compatibility, the Panel
declined to decide it, and focused exclusively on section 771B itself: it held that
15See J.Z. Barsy, “Upstream Subsidies and U.S. Countervailing Duty Law: The Mexican Ammo-
nia Decision and the Trade Remedies Reform Act of 1984” (1984) 16 Law & Pol’y Int’l Bus. 263.
‘6Supra, note 9.
17133 Cong. Rec. S8787 (daily ed. 26 June 1987).
‘8GAYTI, supra, note 2, art. VI(3); Agreement on Interpretation and Application of Articles VI,
XVI, and XXIII of the General Agreement on Tariffs and Trade, 26th supp. B.I.S.D. (1978-79) 56,
art. 4(2) [hereinafter GATT Subsidies Code].
19Supra, note 3 at 8319. See Restatement (Third) of the Foreign Relations Law of the United
States (St. Paul, Minn.: American Law Institute, 1987) 114, 115 quoted by the Panel. It may be
noted that while this is a correct statement of the later-in-time rule, it is not clear which instrument
is the latest to look at, since the FTA, which expressly incorporates the GAIT rules, was signed
on 2 January 1988, before passage of 771B, but entered into effect 1 January 1989, after passage
of the amendment. Both the United States and the Canadian acts implementing the FTA were
adopted after passage of the Omnibus Trade and Competitiveness Act of 1988, Pub. L. No.
100-418, which contained 771B.
1992]
DISPUTE SETTLEMENT & THE PORK CASE
the new section did not contain a “pass-through” test and indeed was designed
expressly to avoid such a requirement, which had been found to inhere in sec-
tion 771A, the 1984 amendment. Whether or not a pass-through test would be
necessary to make an upstream subsidies provision consistent with the GATT,
such consistency, the Panel said, is not a prerequisite to the application of a U.S.
statute.
20
B. Reading the Baucus Amendment
With the pass-through issue disposed of, it remained for the Panel to deter-
mine whether the two conditions expressly included in section 771B had been
met. The Panel agreed with Commerce that the demand for live swine, whose
producers received the challenged payments, was substantially driven by the
demand for pork, even if it could be shown that pork is to a considerable extent
an intermediate product and that the greater demand is for bacon, ham, and
other end-products.’ The other statutory requirement, that the processing oper-
ation adds only “limited value”‘2 to the raw commodity, was more difficult for
the Panel. The Canadian parties argued that the statute was not intended to apply
to substantial processing operations such as transformation of live swine to
pork, but only to minor finishing operations such as freezing of produce; and
further, that existing precedents (including the earlier Pork case)’ established
that 20 percent added value, as found in this case by the Department, would not
be considered “limited.” The Panel pointed out, however, that these precedents
all turned on different statutes; moreover, Commerce had found that most of the
added value of pork is attributable to marketing, whereas the actual cost of proc-
essing live swine into pork is substantially less than the added value and does
not change the essential character of the product. After elaborate discussion, the
Panel concluded that the Department’s interpretation of the new statute could
not be found to be impermissible; accordingly the finding that the pork produc-
ers had added only limited value was allowed to stand. 4
This left two more formidable problems for the Panel. First, assuming pay-
ments made to hog producers were countervailable, how should such payments
be converted in order to impose duties per pound of imported pork? Second,
which of the many kinds of payments were countervailable and which were not,
because they failed the requirement of specific or targeted benefits?
20Supra, note 3 at 8319-20. In a parallel proceeding initiated by Canada in GATTjust after the
United States imposed the countervailing duty in September 1989, a GATT panel did decide this
issue, and upheld the Canadian contention. The Panel recommended that the United States either
refund so much of the countervailing duties collected as represented subsidy paid to the swine pro-
ducers, or conduct an investigation to determine the pass-through on a basis that satisfies art. VI(3).
The Panel rejected a Canadian request that it recommend that 771B be withdrawn, on the ground
that such a request went beyond its mandate. United States – Countervailing Duties on Pork from
Canada, Recommended Decision of 3 August 1990, GATT Doc. 116721 (5 September 1990). As
of January 1992, this decision had not been adopted by the GATT Council, evidently because the
United States has withheld its consent.
21Supra, note 3 at 8318.
22lbid. at 8319.
23Supra, note 5.
24Supra, note 3 at 8320-21.
REVUE DE DROIT DE McGILL
[Vol. 37
C. Measuring the Subsidy
The Commerce Department had used a conversion factor based on the pork
yield of live swine, essentially assuming that the entire value of a hog was
expressed in pork products. The exporters, however, argued that this method
ignored the by-products of hog raising. The precedents, involving several lamb
and fish cases,’ were inconsistent, and, as the Panel found, yielded no clear
principle. Since there was no clear conflict on this issue between section 771B
and the GATT or the basic U.S. subsidies law, those sources, which look to “net
subsidies,”26 must be used. The Panel concluded that
it is unreasonable and not in accordance with law for Commerce to allocate the
entire subsidy conferred on hogs to pork products when other commercial prod-
ucts resulting from hogs are also benefiting from the subsidy.2 7
On this issue, the Panel remanded the case to the Department for reconsidera-
tion.2
D. Specificity
The question of “specificity” is a more general one. The idea is that if a
government benefit is generally available –
for instance a new highway, or an
advanced training institute, or an improved weather service –
it is not regarded
as a countervailable subsidy, even though a given producer or exporter may be
shown to have derived advantage from it. In Pork, the parties agreed that if a
given program of government assistance were generally available, it would not
be countervailable. They disagreed, however, about how to define or determine
whether or not a given program of assistance met the specificity test. The Cana-
dian parties in Pork contended that the so-called “tripartite” stabilization pro-
grams (federal government, provincial government, and producer councils)
were generally available and thus not countervailable. Commerce responded
that while the programs might in theory or de jure be open to all, de facto they
were used only by a few producers, including the producers of swine, and thus
met the test of specificity.29 It turned out that producers of nine agricultural
products had benefited from the tripartite programs, and producers of other eli-
gible products had for one or another reason not applied. The Panel, after sev-
eral pages of discussion, concluded that once a government program of assist-
ance is stated to be generally available, the burden was on Commerce to show
that the program was in fact targeted at the product sought to be made subject
of a countervailing duty, or at a discrete class including that product. On this
issue as well, the Panel remanded.”
1671(a)(2)(A)(ii).
5Cited in ibid. at 8322.
26GATT, supra, note 2, art. VI(3); GATT Subsidies Code, supra, note 18; 19 U.S.C.
27Supra, note 3 at 8323.
281bid.
29Ibid. at 8323-24.
30Ibid. at 8327.
19921
DISPUTE SETTILEMENT & THE PORK CASE
E. Deciding on the Evidence
The Panel proceeded one by one through all of the separate programs that
had been found by Commerce to be countervailable.3″ With respect to some of
the programs, Commerce itself asked for a remand because its lawyers had con-
cluded that the record did not contain substantial evidence to support the chal-
lenged finding. The Canadian parties requested that the Panel instruct the
Department to dismiss as to those programs; the Panel preferred to give the
Department a second chance, on the understanding that any action it would take
could again be challenged before the same Panel.32
With respect to other programs of government assistance, the evidence was
not clear: in some instances the best information available (BIA) was good
enough; in others the Panel remanded for more precise information. In an
instance where the Province of Alberta had declined to provide certain informa-
tion on the basis of a provincial confidentiality rule, the Panel upheld Com-
merce in using BIA, rejecting the contention that BIA had been used in an arbi-
trary or punitive way.33
F Raising New Arguments
One interesting issue of administrative law emerged when the government
of Quebec, which had participated in the proceeding before the Commerce
Department, sought to raise an issue before the Panel that it had not raised
before the Department. The Panel held that the doctrine of exhaustion of admin-
istrative remedies applied, but that Quebec’s appeal came within an exception,
because the issue that it sought to raise before the Panel had been decided
adversely to its position in earlier cases before the Department, and raising it in
Pork before the Department would have been futile. “To have required exhaus-
tion in this case,” the Panel wrote, “would have been insistence on a useless for-
mality.”‘ One may wonder whether the exception to the exhaustion doctrine
was justified in this case, given the right of agencies to change their mind and
a change of administrations between the first and second Pork cases; on the
other hand, granting an exception and declining to accept an unsatisfying ruling
on procedural grounds reflects the kind of consideration to foreign parties (par-
ticularly government parties) that one ought to expect from an international dis-
pute settlement mechanism.
311bid. at 8327-8336.
32Note that under art. 1904(8) of the FTA and under the corresponding section of the United
States-Canada Free Trade Implementation Act of 1988, Pub. L. 100-449, 102 Stat. 1851, 401(7),
a panel is not authorized to dismiss a proceeding, but it may instruct the national agency in question
to do so and the agency is required to follow the instruction.
33Supra, note 3 at 8333-34.
341bid. at 8335. The issue in question was whether Quebec’s Farm Income Stabilization Insur-
ance Program was or was not specific and therefore countervailable. The issue had been fully con-
sidered in the 1985 Live Swine case (supra, note 8). On the merits once it got over the exhaustion
point, the Panel remanded.
McGILL LAW JOURNAL
[Vol. 37
H. Subsidy Phase: Round H
On remand, the Commerce Department first asked for a 75-day extension;
when that was denied by the Panel, the Department did issue its Determination
on Remand on time. The effect of the reconsideration was to reduce the coun-
tervailing duty from C$0.08 to C$0.066 per kilogram.35 On the issue of whether
particular assistance programs were generally available or were, at least de
facto, specific enough to be countervailable, the Department omitted two of the
programs that had been included in its first determination, but found the other
three to meet the specificity test. The Canadian parties sought review with
respect to each of these programs and the Panel issued its decision 90 days
later.36
A. Specificity Again
The Panel had been struck in its first review by the small number of prod-
ucts benefiting from certain assistance programs such as the so-called Tripartite
Benefits Program, which seemed from the statute that established it37 to be open
to a much larger number of agricultural producers than actually participated in
the program. The Department, on remand, in effect said it could not give an
informed explanation of the discrepancy between the number of eligible farmers
and the number of actual participants, but it found that hog producers made up
approximately 35 percent of all farmers benefiting from the Tripartite Agree-
ments and received over half the total amount of payments –
enough to meet
the specificity test. On its second review, the Panel wrote that it
remains dissatisfied with Commerce’s efforts to set forth a rule of law which is
clear, principled and capable of distinguishing intentional … programs from those
that appear specific in a given year or two merely because of unpredictable eco-
nomic variations…. 38
Nevertheless the Panel concluded that Commerce had “not supplied sufficient
facts and rationale to justify its finding of specific subsidy in this matter.’ ’39
Further, the Panel in its first review had stated that Commerce did not pro-
vide adequate justification for its determination that a Quebec Farm Income Sta-
bilization Insurance Program (FISI) met the specificity test. The government of
Quebec, as we saw, had been permitted to raise an argument before the Panel
that it had not raised in the administrative proceeding, namely that 75 percent
of Quebec’s insurable agricultural products received benefits from FISI, show-
ing, Quebec argued, that the program was not specific to hog growers. 4 In its
on Fresh, Chilled and Frozen Pork (7 December 1990), USA-89-1904-06.
35Department of Commerce Remand Determination: Final Countervailing Duty Detennination
36Memorandum Opinion and Order Regarding Commerce Determination on Remand (8 March
1991), USA-89-1904-06, reported as Fresh, Chilled and Frozen Pork From Canada (1991), 4
T.C.T. 7026 (Ch. 19 Panel).
37Agricultural Stabilisation Act, R.S.C 1985, c. A-8, s. 13, amended R.S.C. 1985, c. 40 (1st
38Supra, note 36 at 7030.
391bid.
40Supra, note 3 at 8335-36.
Supp.), repealed S.C. 1991, c. 22, s. 27.
19921
DISPUTE SETTLEMENT & THE PORK CASE
decision on remand, the Department pointed out that it had not previously had
an opportunity to reply to Quebec’s figures, which it would have done had the
argument been made during the investigation. Commerce asserted that if certain
products that Quebec had excluded from its calculation were counted among
“all agricultural products,” the beneficiaries of FISI would be far less than 75
percent of all of Quebec’s agricultural products. Thus Commerce stuck by its
specificity determination. The Panel, looking at essentially the same discussion
a second time, was still not satisfied that the specificity test had been met, and
remanded again on this issue.”
B. From Corn to Hogs
Another program that Commerce had held to be countervailable in its first
subsidy determination concerned payments made to feed grain users, on the
basis of an estimate of the percentage of the consumption of feed grain attrib-
utable to consumption by hogs. In the first round the Department had put the
figure at 15 percent, and the Panel had said this figure was not supported by evi-
dence in the record. On remand, the Department asserted that such evidence as
existed on the ratio of feed grain consumption by hogs to total consumption sug-
gested that the percentage was either 10 or 15 percent, and so it would settle on
a finding of 12.5 percent. The Panel was not satisfied, and remanded again, with
some suggestions on how Commerce “should be able to arrive at a reasonable
estimate.””
C. Further Remand
On 11 April 1991, the Commerce Department considered the case for the
third time, and essentially went along with the Panel. As to the FISI program
in Quebec, the Department eliminated FISI benefits from the calculation of the
amount of the subsidy. As to the allocation of grain benefits to hogs, the Depart-
ment settled on a figure of 11.4 percent, as a “potential (albeit imperfect) proxy
for the amount of feed consumed by hogs.”43 Altogether, the Countervailing
Duty Deposit Rate came to C$0.036 per kilogram, as compared with the initial
determination of C$0.08 kilogram. It was estimated that the final determination
came to a subsidy of 2.9 per cent ad valorem.
Taking the subsidy phase of the Pork case as a whole, it seems that the
process worked pretty much as intended. The subsidy phase of Pork was a big
and complicated case, and it received thorough consideration –
almost cer-
tainly more thorough than it could have received from a single overworked
judge in the U.S. Court of International Trade. In the course of the review proc-
ess the issues were narrowed and clarified substantially. Of the nine issues that
had been challenged in the first review, the Panel had affirmed on three and
remanded on six. On the first remand, one of Commerce’s major revisions (the
conversion factor from hogs to pork) was not further challenged, and two pro-
41Supra, note 36 at 7034.
42Ibid.
43F.T.A.P.D. LEXIS 5, at * 2 (1991).
REVUE DE DROIT DE McGILL
(Vol. 37
grams previously deemed countervailable were found not to have benefited the
industry during the period of investigation. On the second round of panel
review, one determination of the Department was upheld,and two were further
remanded. While one would not say that the Department and the Panel were
thinking alike on all issues, the review process was at all times professional,
seeking to draw inferences from incomplete data and occasionally shifting the
burden of coming forward with “best information available,” or “substantial
evidence on the record.” Certainly nothing in the subsidy phase of the Pork case
came close to the disrespect and defiance evident in the injury phase of the case,
to which we now turn.
I1. Injury Phase: Round I”
Under the United States trade law since 1979 (as under Canadian law and
under the GATT), a finding that a product has been imported with the benefit
of a subsidy is not sufficient to support imposition of a countervailing duty. A
countervailing duty, in other words, is not seen as a penalty for improper con-
duct, but as a way to restore a balance that has been distorted. Before a coun-
tervailing duty may be imposed, there must be a finding of material injury to an
industry, or “threat of material injury.”45 The determination of injury is made in
the United States by the U.S. International Trade Commission (ITC), which, as
noted in the introduction, is an independent regulatory agency not subject to
direction by the executive branch, but subject, since the 1970s, to judicial
review.
The injury phase of the Pork case came before the ITC in May 1989 (i.e.,
after Commerce had made its preliminary determination of subsidy). The Com-
mission’s practice is to order a study by its own staff, and at the same time to
invite interested persons to make written submissions, both of economic data
and of legal arguments. In some instances, including the Pork case, the Com-
mission holds an oral hearing.
When the Commission came out with its decision on injury in September
1989, it was unanimous that imports of pork from Canada were not presently
causing material injury to the pork industry of the United States.46 Domestic
pork production had risen 11.6 percent during the period of investigation
(1986-88), and was continuing to rise in the first quarter of 1989, though at a
slower rate. Pork prices were falling, however, and three members of the Com-
mission concluded that the U.S. pork industry was “entering a period of vulner-
ability to the effects of subsidized imports from Canada,” i.e., the industry was
“threatened with material injury,” within the meaning of 705(b)(1)(A)(ii) of
the TariffAct of 1930, as amended.47 Both the Chairman and the Vice-Chairman
of the Commission dissented, and one member did not participate.
“Memorandum Opinion and Remand Order (24 August 1990), USA-89-1904-11,
reported as
Fresh, Chilled or Frozen Pork from Canada (1990), 3 T.C.T. 8276 (Ch. 19 Panel), 3 T.T.R. 281
[hereinafter Pork: Injury Phase cited to T.C.T.].
4519 U.S.C. 1671(a).
46Fresh, Chilled, or Frozen Pork from Canada, USITC Pub. No. 2218 (September 1989).
4719 U.S.C. 1671d(b)(1)(A)0ii).
19921
DISPUTE SETTLEMENT & THE PORK CASE
The majority quoted a decision of the Court of International Trade, to the
effect that since determination of a threat of injury involves projection of future
events, it is “inherently less amenable to quantification.”4 But the Commission-
ers reasoned that the subsidies found by the Commerce Department would
increase production in Canada faster than consumption, that as a result exports
to the United States would increase, and that as a consequence prices in the
United States would be suppressed. The dissenters pointed out that in the last
full year under investigation, imports of pork from Canada had constituted less
than 3 percent of U.S. consumption by volume, that the effect of imports on
prices of domestically produced pork was very small, and that the data on
employment and capacity utilization did not “remotely suggest” the likelihood
that the U.S. industry was in imminent danger of material injury.
Again, the Canadian parties applied for review by a panel under the FTA,
and a different panel from the one that was hearing the subsidy phase was con-
vened. Thus for most of the fall and winter 1989-90, two panels were sitting on
the two aspects of the Pork case.
As previous panels had done, the Panel in Pork: Injury Phase went over
the requirements of the FTA, that U.S. law be followed in its review, that it was
bound by the substantial evidence standard, but that it was not bound to agree
with the agency’s decision.49 The Panel in Pork emphasized, however, that a
finding of threat of injury must be based on a record before the Commission
showing more than a possibility of injury.” The Panel was not satisfied that the
record supported the finding of threat of injury, and remanded that case with
instruction to the Commission to reconsider the evidence. The Panel stated its
view that the burden on the ITC is higher than in other cases when a finding of
“threat of material injury” is coupled with a finding of “no present material
injury.” [emphasis added]”‘
A. A Question of Numbers
The issue before the Panel was unusual in that it turned largely on what the
Panel called questionable interpretation of unreliable statistics.5″ The argument
of the Commission majority had been that pork production (found by Com-
merce to have been subsidized) had grown from 2 billion pounds in 1986 to 2.6
billion pounds in 1988, while Canadian consumption had increased by only 110
million pounds. Ergo, the Commission had found (i) the excess was likely to be
exported to the United States; and (ii) the rate of increase was likely to continue.
But much of the reported increase in production turned out to reflect a change
of method of counting and reporting pork production in Canada. The actual
increase in production had been only 170 million pounds, not 600 million
48Hannibal Industries Inc. v. United States, 710 F. Supp. 332 at 338 (Ct. Int’l Trade 1989) [here-
inafter Hannibal Industries].
49Supra, note 44 at 8278-79.
510 bid. at 8279.
5tlbid. at 8281.
521bid. at 8283.
McGILL LAW JOURNAL
[Vol. 37
pounds, or 8.4 percent, not 31 percent –
tion in the United States. 3
less than the increase in pork produc-
This discrepancy, acknowledged by the ITC itself, made a remand inevita-
ble, and indeed the ITC had itself moved for a voluntary remand, which the
Panel had denied because the motion had come too late. The error affected not
only the estimate of increased exports to the United States that underlay the
finding of threat of injury, but drowned out other facts to which the ITC might
have attached greater importance, such as figures on hog breeding in Canada,
which declined after 1988 in accordance with the so-called “hog cycle.”‘ Once
these production figures were reconsidered, the ITC might well have reached a
different conclusion also on the causal link between the Canadian subsidy pro-
gram for pork and imports to the United States; in addition, the ITC might have
paid greater attention to the increased exports of hogs to the United States,
which the Commission had found irrelevant but the Panel thought “quite appar-
ently”5 would reduce Canada’s exports of pork.
B. A Question of Inferences
The Panel found other errors in the Commission’s reasoning. For instance,
Canadian exports of pork to Japan had increased because of quality problems
with pork supplies to Japan from Taiwan. Once Taiwan corrected its problem,
the Commission had reasoned, Canada’s exports to Japan would be likely to
drop, and the goods not sent to Japan would find their way to the United States.
“This strikes the Panel as inappropriate, selective fact finding,”56 the Panel
wrote, pointing out that Canadian exports to Japan had continued to increase,
quarter by quarter and month by month, with one exception seized upon by the
ITC.
Again, to show that an increase in exports from Canada could not be easily
absorbed, the ITC had argued that U.S. consumption of pork had declined over
the past decade. But the Panel pointed out that only per capita consumption of
pork in the U.S. had declined; given a growing population, overall consumption
had in fact increased. Here was one more example of a finding by the Commis-
sion not supported by the recordY
531bi
54The “hog cycle,” sometimes known as an illustration of the “cobweb theorem,” assumes that
as hog prices rise in Period I producers will increase the supply, so that in Period II supply exceeds
long-run equilibrium in relation to demand. Consequently, the price in Period II will fall to less
than an equilibrium position. In Period I supply will be reduced in response to the prior low price,
and thereafter the price will again rise and over-shoot equilibrium, so that there will be an excess
of supply, and so on. The basic operative decision is taken by the grower when he decides, on the
basis of economic signals in the preceding 2-6 months, whether to fatten the 5-month old female
for slaughter, or to retain her for breeding purposes, which occurs at about 8 to 10 months of age.
Hogs gestate for about 4 months, and are ready for slaughter about 6 months after birth. In the
United States, a hog cycle is typically 2 years in duration from peak to trough and 4 years from
peak to peak.
55Supra, note 44 at 8285.
561bid at 8286.
571bid.
1992]
DISPUTE SETILEMENT & THE PORK CASE
To take just one more point, the ITC had found the U.S. pork industry “par-
ticularly vulnerable.””8 The Panel did not find this conclusion, which differed
from an earlier investigation of the ITC, 9 to be consistent with the evidence.’
C. A Question of Logic
One of the Canadian panelists, a professor of economics, submitted addi-
tional views, to emphasize what he regarded as not only inaccuracies in the
interpretation of basic data but “incompleteness in the analytical logic”‘” linking
cause and effect. His point was that continuation of subsidies at existing levels
that do not cause material injury cannot, without some additional change, pro-
vide the basis for a finding of threat of injury in the future. Putting aside the sub-
stance of the argument, Professor Whalley’s opinion is striking in that it calls
for analysis in accordance with the “mainstream economic profession,”’62 a quite
different standard from the test of substantial evidence on the record applied by
the Court of International Trade and the Panel in Pork and all the other cases
under Chapter 19 of the FTA. On the merits, Professor Whalley’s critique of the
Commission’s reasoning is quite persuasive in challenging the effect in the large
economy (USA) of behaviour in the smaller economy (Canada) linked by sub-
stantially open borders. In terms of the mandate of the panels, the professor’s
approach, in effect placing the burden on the Commission to defend its eco-
nomic analysis and (because the issue is threat of material injury) its projec-
tions, is quite at variance with the deference that a reviewing court (and by def-
inition the binational panel) is expected to accord to the expertise of an
administrative agency. The other members of the Panel did not follow Professor
Whalley’s lead.
D. Summary of Round I
Each of the other panel decisions under Chapter 19 of the FTA – whether
they sustained or reversed the agency – had considered close questions of law
on which one could fairly come out either way. Also, each of the other panel
decisions that had remanded a decision to the rendering agency had done so
only in part. The injury phase of the Pork case was a total rejection of the deci-
sion of the ITC, for using erroneous data and then misinterpreting the data.
IV. Injury Phase: Round H
A. Back before the Commission 63
On 23 October 1990, within the 60 days provided in the remand order, the
ITC issued its Decision on Remand, and again found threat of-injury, though on
581bid. at 8288.
59Live Swine, supra, note 8.
60Supra, note 44 at 8288-89.
611bid. at 8289.
621bid. at 8290.
6 3Fresh, Chilled, or Frozen Pork from Canada, Views on Remand, Inv. No. 701-TA-298, USITC
Pub. No. 2230 (October 1990).
REVUE DE DROIT DE McGILL
[Vol. 37
somewhat different grounds. This time the vote in the Commission was 2-1, as
two of the members of the Commission who had heard the original petition had
in the meantime resigned.
The two members of the Commission who again found threat of injury,
Commissioners Rohr and Newquist, writing separately, made elaborate analyses
of the hog cycle, from which (with minor variations) they concluded that the
U.S. pork industry would be most vulnerable in the coming downward portion
of the cycle, which would coincide with higher Canadian subsidies on hogs
(because there was really a single North American market affected by a single
hog cycle). Higher Canadian subsidies on hogs, in turn, would lead to higher
U.S. countervailing duties on live swine, which in turn would lead to “product
shifting,” i.e., reduction of imports of live swine and correspondingly higher
imports of pork, thus creating a “threat of injury” to the U.S. pork industry.
“Certainly,” Commissioner Rohr conceded,
the impact of the subsidized Canadian imports is much smaller than many other
factors affecting the industry. But the standard I am legally required to apply is
whether the imports will be contributing even minimally … to material injury. I am
satisfied that this standard has been met.64
Commissioner Newquist wrote:
Although it is possible that pork imports from Canada will decrease, it is also
likely that production levels in the United States will decrease. Thus, Canadian
imports entering at a higher level than would be the case absent the subsidies (even
if they are not increasing absolutely) may well – given declines in U.S. produc-
tion –
Excess pork production in Canada, the likelihood of product shifting, and the
impending decline in domestic production, all lead me to reaffirm my earlier find-
ing that an increase in import penetration is likely. 65
take an increasing share of the market.
The argument about product shifting was new, at least in the form pre-
sented by the two commissioners. Acting Chairman Brunsdale, who had been
in the minority in the original decision, was not persuaded. She pointed out in
a brief dissent that Canadian pork had never achieved an import penetration
level into the United States higher than 3.4 percent –
and
that even if the total of Canadian increase in pork production in the two-year
period under review had been sold to the United States, Canadian import pen-
etration would not have reached 5 percent of the U.S. market.
not an injury level –
Altogether, one could not say that the performance of the Commission was
intellectually satisfying, as discussions of “threat of injury” rarely are. But the
debate remained within the bounds of professional discourse, members of an
independent commission doing their job as they saw it. As could be expected,
the Canadian parties promptly moved for review of the Commission’s Decision
on Remand by the same panel that had remanded the earlier decision, and that
motion was granted by the Panel.
64Ibid. at 21.
65Ibid. at 33.
19921
DISPUTE SETTLEMENT & THE PORK CASE
B. Back before the Panel
On 22 January 1991 (90 days after the Commission’s second decision), the
Panel rendered its second decision, in a Memorandum Opinion and Order of 38
pages.’ The Panel concluded –
that it was required by
article 1904.8 of the FTA to issue a “final decision,”’67 i.e., that the Agreement
did not contemplate or permit successive remands.68
it seems correctly –
1.
Reopening the Record
One issue not previously discussed affected the remand proceedings in a
major way. The ITC, having been criticized for using unreliable statistics, had
announced that it would reopen the record on which it had based its original
decision “on three narrow aspects.”6 9 Parties were invited to make submissions
on these three aspects only, within 10 days. Thereafter, however, the Commis-
sion, on its own, had taken into account documentary information –
particu-
larly Canadian government statistics – going beyond the points specified in its
notice and identified in the Panel’s Remand Order. In part in reliance on this
new information, the Commission’s second decision had relied on a new ground
to uphold its finding of threat of injury. The Canadian parties, complainants
before the Panel, questioned whether the Commission had the authority to open
the record, both on the “three narrow aspects” on which the parties could make
further submissions and on other aspects.
The question of the authority of an administrative agency to go beyond the
initial record in a remand from a court has been a controversial one in U.S.
administrative law, and (perhaps to its own surprise) the Panel selected for its
expertise in international trade law7″ now set out some of that controversy, quot-
ing at length from a decision of the U.S. Supreme Court in a broadcast appli-
cation case 50 years ago, Federal Communications Commission v. Pottsville
Broadcasting Co. 7 In Pottsville, the Supreme Court, overruling the Court of
Appeals, had sustained the FCC’s decision to admit a new applicant into a pro-
ceeding after remand, on the ground that administrative agencies should not be
66Memorandum Opinion and Order Regarding ITC’s Determination on Remand (22 January
1991), USA-89-1904-11, reported as Fresh, Chilled or Frozen Pork from Canada (1991), 4 T.C.T.
7014 (Ch. 19 Panel).
6 7Ibid. at 7016.
6 SFTA, art. 1904.8 reads, in pertinent part, as follows:
Where the panel remands a final determination, the panel shall establish as brief a time
as is reasonable for compliance with the remand, taking into account the complexity
of the facts and legal issues involved and the nature of the panel’s decision. In no event
shall the time permitted for compliance with a remand exceed an amount of time equal
to the maximum amount of time (counted from the date of the filing of a petition, com-
plaint or application) permitted by statute for the competent investigating authority in
question to make a final determination in an investigation. If review of the action taken
by the competent investigating authority on remand is needed, such review shall be
before the same panel, which shall issue a final decision within 90 days of the date on
which such remand action is submitted to it.
69Remand Notice, Inv. No. 701-TA-298 (19 September 1990).
70See FTA annex 1901.2, paragraph 1.
7’309 U.S. 134 (1940).
McGILL LAW JOURNAL
[Vol. 37
subjected to the tight rules applicable to lower courts, but “should be free to
fashion their own rules of procedure and to pursue methods of inquiry capable
of permitting them to discharge their multitudinous duties.”72 The Panel con-
cluded, however, that the powers of the ITC in a remand determination ordered
by a binational panel under the FTA might well be different. Had the Commis-
sion stuck to its Notice of Remand and enlarged the record only in respect to
the three aspects mentioned, the Panel would not, it seems, have reversed the
Commission on this ground.73 However, the fact that the Commission had con-
sidered new material on other issues as well, without notice to the parties or
opportunity for comment, struck the Panel as unacceptable. The Panel noted that
in review of determinations on remand it was limited to 90 days,74 and that it
could not comply with such a limit unless there were an end to new issues and
new evidence. “A line,” it wrote, “must be drawn somewhere.”75 Thus the Com-
mission had committed a legal error, and its second determination, like the first,
must be set aside.76 In view of the Commission’s response to the Remand Order,
as described below, it must be said that the Panel’s tone was polite and
restrained. Moreover, the Panel did not rest its decision on the procedural point,
even if it characterized the point as one of “fair play.”77 The Panel went on to
address on the merits the new issue raised by the Commission on remand, the
issue of product shifting.
2.
Product Shifting
It was understood by all that the Canadian and provincial governments had
been giving subsidies to producers of hogs (live swine),7” and that at least some
of these subsidies had led to countervailing duties imposed by the United States.
The ITC majority, as discussed above, had predicted that Canadian subsidies on
hogs and subsequently U.S. countervailing duties on live swine would be
increased in periods following the period under review, and had inferred that
this would lead to reduced exports of live swine to the United States and
increased exports of pork. This prediction of product shifting, the Commission-
ers had reasoned, supported the finding of “threat of injury” to the American
72Ibid. at 143.
73Supra, note 66 at 7020-21.
74See FTA, art. 1904.8, quoted at note 68 above.
75Supra, note 66 at 7020.
76A subsidiary argument concerned the contention of the Commission before the Panel that the
complainants (Canadian parties) did not have the right to due process under the Fifth and Four-
teenth Amendments to the U.S. Constitution. The Panel side-stepped this contention by pointing out
that art. 1904.3 of the PTA required it to apply general legal principles and that art. 1911 defined
general legal principles to include due process.
This writer’s view, expressed in numerous writings, is that the Bill of Rights, and particularly
the due process clause, constrains the conduct of officers of the United States acting in their official
capacity wherever they act; here, of course, they acted in Washington D.C. in respect of proposed
duties to be imposed on imports into the United States. Thus, whatever the merits of the contro-
versy about the fairness of enlarging the record upon remand, the argument that the Commission
is not bound by the requirements of due process is unfortunate, to put it as politely as possible,
and one may hope that it will not be repeated.
77Supra, note 66 at 7020.
78See Subsidy Phase Round I, supra, note 3 and accompanying text.
1992]
DISPUTE SETTLEMENT & THE PORK CASE
pork industry. The Panel was not impressed. Neither of the Commissioners’
findings, somewhat different from one another, rested, in the Panel’s view, on
substantial evidence. 79
In support of this observation, the Panel perhaps made an error in that it,
too, looked beyond the record by citing later findings of the Department of
Commerce to show that the predictions of the two Commissioners were not
sound.s In retrospect, the error was a tactical one, in that it opened up the Panel
to criticism of doing what it had criticized the Commission for doing. It seems,
however, that the Panel was correct in faulting the Commissioners for lack of
support for the product shifting theory, and as Commissioner Brunsdale later
pointed out,” the point was not necessary to the Panel’s conclusion.
3.
More on the Hog Cycle
The two Commissioners had made other arguments as well, based on their
interpretation of the so-called hog cycle, and the likelihood, as the Commission-
ers viewed it, of oversupply of pork in a declining market, and of price cutting
by sellers of Canadian pork. Again, the argument failed to persuade the Panel.
“[The Panel is forced to the conclusion,” it wrote, “that the theory is needed
because of an absence of evidence of causation. ‘ 2
At this point, the impatience of the Panel with the Commission majority
was beginning to show through:
[T]he Panel again remands the ITC’s Remand Determination for action … not
inconsistent with the Panel’s decision of August 24, 1990, and not inconsistent
with the Panel’s Memorandum Opinion [herein] … and instructs the ITC to con-
duct this second remand without any further reopening of its Record … The results
of this further remand shall be provided by the ITC to the Panel within 21 days
of the date of this decision.8 3
An outside observer might suggest that the traditional courtesies of international
dispute settlement, which had on the whole been observed in the earlier phases
of the Pork case and in all of the other cases under Chapter 19, were beginning
to wear thin. The Panel’s Memorandum was nothing, however, compared to
what followed on the second remand.
V.
Injury Phase: Round I
A. The Second Remand at the ITC
On its second remand, the ITC did as it was told. It unanimously deter-
mined that an industry in the United States was not materially injured or threat-
79Supra, note 66 at 7026.
80Ibid. at 7024-25. The Panel referred to preliminary results of an administrative review by the
Department of Commerce of results of a countervailing duty order on live swine originally issued
in 1985, covering the period 1 April 1987-31 March 1988,55 Fed. Reg. 20812 (1990), which indi-
cated that the deposit rate on sows and boars would be reduced to 0.71 Canadian cents per lb. In
its final determination, issued after all of the steps here discussed had been completed, the Depart-
ment reduced the deposit further, to 0.30 Canadian cents per lb. See 56 Fed. Reg. 10410 (1991).
81See infra, note 89 and accompanying text.
82Supra, note 66 at 7025.
831bid. at 7026.
REVUE DE DROIT DE McGILL
[Vol. 37
ened with material injury by reason of imports of pork from Canada found to
have been subsidized.’ But the Commissioners who had made up the majority
on the first remand left no doubt about how they felt:
The Panel now precludes the Commission, in this remand, from considering rel-
evant evidence as to both U.S. and Canadian pork production and from consider-
ing product shifting as a basis for a threat determination. Furthermore, the Panel
held that, in the absence of underselling the Commission is precluded from mak-
ing a finding of price suppression. The Panel further circumscribed the Commis-
sion’s discretion on remand by obliquely holding, without explanation, that Cana-
dian exports to the United States will not gain a higher relative share of the U.S.
market when U.S. production declines. We believe that these restrictions are con-
trary to the facts and the law, but because they are imposed by the Panel, they are
legally binding on us. Thus, we have no choice but to determine on remand that
the domestic industry is not threatened with material injury by reason of imports
of fresh, chilled or frozen pork from Canada which the Department of Commerce
has determined are being subsidized.
Notwithstanding this determination, this Second Panel Decision violates fun-
damental principles of the United States-Canada Free-Trade Agreement (FTA)
and contains egregious errors under U.S. law. Had this decision come from the
Court of International Trade, unlikely in light of the numerous CIT authorities
contrary to the Panel’s holding, we would have directed counsel to appeal it to the
Court of Appeals for the Federal Circuit. That avenue, however, is not available
to us in light of the provisions of the FTA. At the minimum, however, we find
many aspects of the Panel decision to lack “intrinsic persuasiveness” and, thus, we
will not change our practice or 8rocedure to conform with those aspects of the
Panel opinion discussed below.8
…And so on for more than thirty pages, full of statements referring to the
Panel’s “preordained outcome,” “counterintuitive, counterfactual, and illogical,
but legally binding conclusion,” “deliberate misunderstanding of the Commis-
sion’s views,” “woeful lack of knowledge,” “egregious intrusion into the factual
decision-making authority of the Commission,” “impermissible reweighing of
the evidence,” etc.
The two Commissioners who had twice been overruled by the Panel wrote,
as quoted above, that if the decision of the reviewing authority had come from
the Court of International Trade they would have instructed their counsel to
appeal, a remedy not open to them under the FTA. One may also safely surmise
that if the decision on remand had been issued by the Court of International
Trade, the Commissioners would not have used the tone they used to attack the
Panel. The tension was compounded by the separate views of the Acting Chair-
man of the Commission, Anne F. Brunsdale, who concurred in the result
because the finding of threat of injury had been rejected, but dissented, as
before, from the views of Commissioners Newquist and Rohr, and complained
that because her views on the merits were known, she had not even been per-
mitted to review the draft opinion prepared by the General Counsel pursuant to
the majority’s instructions. Though she understood that the majority “takes
umbrage with certain portions of the Panel’s opinion,” Brunsdale agreed with
the Panel that as a matter of due process, the Commission should not base its
84Fresh, Chilled, or Frozen Pork from Canada, Second Remand Determination, 12 February
1991 [on file with author].
851bid. at 4-5. Footnotes to the quoted text have been omitted.
19921
DISPUTE SETTLEMENT & THE PORK CASE
decisions on remand on grounds different from those supporting its original
decision.86 “While I differ on some of the particulars of the Panel’s decision,”
she wrote, “it does make good general points regarding the Commission’s obli-
gations when the case is remanded from an appellate authority. In the future, I
will be sure to keep these principles in mind and act accordingly.”‘
VI. Some Reflections on the Clash between the Panel and the Commission
A. Predictions, Predictions
Beyond the quarrel between the Commission majority and the Panel, and
among the members of the Commission, the Pork case raises an issue important
both to administrative law and to international trade law. In its Determination
on Remand on “threat of injury,” the Commission had developed a theory of
“product shifting,” which as discussed above predicted a rise in imports of pork
on the basis of an expectation of a rise in countervailing duties on live swine.88
By the time the Panel heard the case for the second time, it had turned out that
countervailing duties did not rise as much as the Commission had expected, as
shown by later findings by the Commerce Department published in the Federal
Register. In its opinion on the second remand, the majority of the Commission,
stung by the Panel’s criticism of its use of new information in the first remand,
in turn criticized the Panel for citing the later figures from the Federal Register
to support its rejection of the Commission’s prediction. But as Chairman Bruns-
dale pointed out, it is hard to review a finding of “threat of injury” – by def-
inition based on a prediction – while closing one’s eyes to information show-
ing that the prediction had been erroneous. Chairman Brunsdale chose to regard
this part of the Panel’s decision as obiter dictum. 9 To the present writer, it con-
firms a long-time scepticism about findings of threat of injury in the absence of
a finding of present injury. Of course the concept of “threat of injury” is con-
tained in article VI of the GATT, in the Subsidies and Anti-Dumping Codes
concluded during the Tokyo Round of trade negotiations, and in both United
States and Canadian law.’ A good case can be made that in certain situations,
typically when there have been rapid surges of imports and a clear trend has
developed, the law need not require proof of devastating injury before relief can
be granted. But given the uncertainty of predictions, it is sometimes said that
burden of proof of threat of injury should be heavier than for proof of actual
861bid. at 36-38 (concurring views of Acting Chairman Brunsdale).
87lbid. at 37. Compare the last sentence in the Acting Chairman’s opinion with the last sentence
of the long quotation from the majority, supra, note 85 and accompanying text.
88Supra, note 65.
89Supra, note 84 at 40.
90For the United States law, 19 U.S.C. 1671(a)(2)(A)(ii) (subsidies); 1673(2)(A)(ii)
(dumping). For the Canadian law, see Special Import Measures Act, R.S.C. 1985, c. S-15, s. 3,
which refers to “dumped and subsidized goods imported into Canada in respect of which the
[Canadian International Trade] Tribunal has made an order or finding … that the dumping or sub-
sidizing of goods of the same description has caused, is causing, or is likely to cause material injury
…. [emphasis added] See also the Canadian International Trade Tribunal Act, R.S.C. 1985, c.
C-18.3, s. 26(4).
McGILL LAW JOURNAL
[Vol. 37
injury.9 Could one go further and posit a general assumption that –
as seems
to have happened in the Pork case –
scrutiny of the finding of threat of injury
by the reviewing authority should be stricter, and deference to the administering
authority should be less than for other issues in anti-dumping and anti-subsidy
proceedings?
B. The Pork Case and the FTA
As to the binational review process itself, clearly it cannot continue if a
albeit an independent agency –
major agency of the U.S. government –
regards the process as fundamentally unfair. However only two Commissioners
have expressed this view, with the acting Chairman disagreeing and three
vacancies; the Panel, for its part, has been discharged, and it is unlikely that its
members will soon serve again as panelists –
certainly not together. Thus the
clouds of war may not be as threatening as one would have thought in the spring
of 1991 from first reading the decisions of the Panel and the Commission major-
ity in the Pork case. In fact the clouds became even darker in the ensuing
months, before a silver lining appeared, as described below.
VII. Extraordinary Challenge: Round IV
A. Raising the Challenge
The two disgruntled Commissioners of the ITC had contented themselves,
as described above, by venting their frustrations in their opinions on remand.
The private litigants, however, i.e., the National Pork Producers Council and its
members and associates, first changed counsel and then sought to invoke –
for
the first (and at this writing, the only) time since the FTA entered into effect –
the Extraordinary Challenge Procedure provided in article 1904.13 of the FTA.
Article 1904.13 and the Annex to that article provide for establishment of a
three-member committee selected from a roster of 10 retired judges (half Cana-
dian, half American) which is supposed to decide within 30 days from its estab-
lishment whether
a)
i) a member of the panel was guilty of gross misconduct, bias, or a serious
conflict of interest, or otherwise materially violated the rules of conduct,
ii) the panel seriously departed from a fundamental rule of procedure, or
iii) the panel manifestly exceeded its powers, authority or jurisdiction set forth
in this Article, and
b) any of the actions set out in subparagraph (a) has materially affected the panel’s
decision and threatens the integrity of the binational panel review process.92
A private party cannot itself invoke the Extraordinary Challenge Proce-
dure. That decision is up to the “Parties,” i.e., the two governments. For the
91See, for example, Hannibal Industries, supra, note 48 at 338 quoted by the Panel in its second
opinion in Pork, supra, note 66 at 7021. In that case the Court upheld the Commission’s determi-
nation (3-2) that threat of injury had not been proved. Commissioner Rohr, one of the members
of the majority in the Pork case, dissented in Hannibal Industries, i.e., he would have found threat
of injury, though not actual injury.
92FTA, art. 1904.13.
1992]
DISPUTE SETTLEMENT & THE PORK CASE
United States, the decision on whether to initiate an Extraordinary Challenge is
committed to the U.S. Trade Representative, who is to consider recommenda-
tions from an interagency group under her chairmanship.93
The request for establishment of the Extraordinary Challenge Committee
did not allege personal misconduct of any panelist; it did, however, invoke both
paragraphs (a)(ii) and (a)(iii). In brief, the petition criticized the Panel in five
respects. It asserted:
1. That the Panel had departed from a fundamental rule of procedure in holding
that the Commission had been wrong in its determination on the first remand
reopening the record beyond the limited notice it had given. According to the
petition, the Panel was wrong to apply principles of “fair play and due proc-
ess,” rather than determining what process was due under U.S. law. “If pan-
els are permitted to ignore U.S. law and make new FTA law,” the petition
asserted, “…the integrity of the binational panel process will be undermined.”
2. That the Panel itself erroneously considered evidence outside the administra-
tive record, i.e., the information on actual sales of live swine as disclosed by
the Department of Commerce after the Commission’s original determination.
3. That the Panel had “invented” a rule of finality in ordering that its second
review would be its last. The Panel, as noted above, had construed article
1904.8 of the FTA as requiring a “final decision” if review of a decision on
(first) remand was needed. The petition, however, asserted error in that panel
review is supposed to be like review by the Court of International Trade,
which has from time to time remanded a case to the Commission more than
twice. ‘”This perceived need for finality,” the petition asserted, “led to a rush
to judgment at odds with the governing law.”
4. That the Panel had disregarded the substantial evidence standard in rejecting
the finding of Commissioners Newquist and Rohr that there was likelihood
of product shifting from exports of live swine to pork. The Panel, it will be
recalled, had thought the whole theory of product shifting was needed
because the evidence was lacking.
5. That the Panel had erred in requiring evidence of price underselling, though
the Commission had not made a specific finding of price underselling, but
had simply inferred the likelihood of price suppression from the likelihood
of increased supplies.94
Finally, the petition sought to put pressure on the Trade Representative. “[T]he
Second Remand Decision,” the petition said, “not only undermines the integrity
of the binational review process with Canada but throws into doubt the wisdom
of using such a mechanism in any free trade agreement negotiated with Mex-
ico.”95
93See U.S.-Canada Free Trade Agreement Implementation Act of 1988, 405(a)(I)(A) and
(B)(iv), Pub. L. 100-449.
94petition to Invoke Extraordinary Challenge Procedure, at 18-20 [on file with author].
951bid.
REVUE DE DROIT DE McGILL
[Vol. 37
The decision for the U.S. Trade Representative on whether to initiate the
Extraordinary Challenge Procedure must not have been an easy one. On the one
hand the Trade Representative surely desired to keep the FTA on track and to
maintain the essentially non-political way that Chapter 19 disputes have been
addressed; on the other hand, she had to take into account the discontent on the
part of the ITC, as well as on the part of the National Pork Producers Council
and its friends in Congress, whose votes would be needed in the context of the
President’s request for extension of fast-track authority for the Uruguay Round
and for negotiation of a free trade agreement with Mexico.96 In the event, the
U.S. Trade Representative decided to request formation of an Extraordinary
Challenge Committee, and to include in the request, dated 29 March 1991,
every one of the points raised in the petition of the complainants.97
To the present writer, outside the line of fire, it seemed that the Extraordi-
nary Challenge Procedure was not suitable for the issues raised in the Pork case,
and I so stated in the Report to the Administrative Conference from which this
paper is adapted. The potential members of the Extraordinary Challenge Com-
mittee, by definition, are not experts in international trade, but as retired judges,
are expected to be able to recognize gross misconduct, conflict of interest,
denial of the right to be heard, and the like. They are not supposed to be a court
of appeal for the binational review procedure. If, as remarked earlier, patience
on all sides had worn thin in this case by the time of the second determination
and second remand, it did not seem to me that these manifestations “threat-
en[ed] the integrity of the binational panel review process.” I was worried that
Chapter 19 of the FTA might go the way of the World Bank Convention on the
Settlement of Investment Disputes between States and Nationals of Other
States” – which has been seriously undermined by repeated resort to a proce-
dure for annulment of arbitral awards” that was intended as a safety valve for
gross violations of due process but has come to be used by dissatisfied litigants
as a device for delay and repeated appeals.
B. The Decision of the Extraordinary Challenge Committee
It turns out that I need not have worried. An Extraordinary Challenge Com-
mittee was quickly chosen, consisting of two retired Canadian judges and one
96In fact letters to Ambassador Hills urging resort to the Extraordinary Challenge Procedure
were sent from some 26 members of the Senate and about 50 members of the House of
Representatives.
97I think it is not unfair to point out that Senator Baucus of Montana, the spokesman for the U.S.
pork producers (and author of the amendment, supra, notes 9, 16 and accompanying text, with
which the other Pork Panel had wrestled) was also the Chairman of the Subcommittee on Inter-
national Trade of the Senate Finance Committee. Senator Baucus had already promised his support
for fast-track authority, and with his help, the resolutions to reject fast-track authority were
defeated in both Houses of Congress. See 137 Cong. Rec. H3517, 3588 (23 May 1991) (House of
Representatives); 137 Cong. Rec. S6765, 6777, 6829 (24 May 1991) (Senate).
9818 March 1965, 17 U.S.T. 1270, T.I.A.S. No. 6090, 575 U.N.T.S. 159 (in force 14 October
1966).
“Ibid. art. 52.
10See, for example, W.M. Reisman, “The Breakdown of the Control Mechanism in ICSID Arbi-
tration” [1989] Duke L.J. 739.
19921
DISPUTE SETTLEMENT & THE PORK CASE
retired American federal judge, who served as chairman. The Committee
received extensive briefs and heard oral argument on 15 May 1991.01 By 14
June 1991, the Committee was ready with its opinion:
As its name suggests, the “extraordinary” challenge procedure is not intended to
function as a routine appeal. Rather the decision of a binational panel may be chal-
lenged and reviewed only in “extraordinary circumstances.” While the legislative
history of the extraordinary challenge committee mechanism is lacking in specif-
ics, it is clear that the extraordinary challenge procedure is intended solely as “a
safeguard against an impropriety or gross panel error that could threaten the integ-
rity of the [binational panel review] process …… …The challenge committee’s
function is to determine whether a panel or panel member violated the three-prong
standard of the extraordinary challenge procedure. 102 In contrast, a binational
panel is composed of five individuals with expertise in international trade law. The
panel members’ function is to review the record evidence and the trade law issues
that have been raised before the competent investigation authority. The committee
and the panel have separate roles and different expertise; it is not the function of
a committee to conduct a traditional appellate review regarding the merits of a
panel decision.’
0 3
The Committee pointed out further that under the FTA it has only 30 days
to complete its task, in contrast to the 315 days allotted to the ordinary panels
for review of agency decisions, and that only governments, not private parties,
could initiate challenge procedures.
The Committee reviewed all five allegations of error by the Panel in its
second review:’0 4
1. The claim that the Panel created a due process principle independent of U.S.
law. This had to do with the Panel’s criticism of the ITC for opening the rec-
ord on remand beyond what it had announced. The assertion that applying
a due process standard was contrary to U.S. law might have sounded strange
to the Committee. The Commissioners had argued that as foreign nationals
the Canadian parties were not entitled to the protection of the Fifth Amend-
mnent to the U.S. Constitution (a highly dubious argument), l0” and the Panel
had side-stepped this issue by pointing out that in any event article 1911 of
the FTA expressly incorporates due process as one of the “general legal prin-
ciples” to be applied by the panels. The Committee rejected the argument
raised in the challenge that this was using the FTA as an independent source
of law, pointing out that the Panel had carefully reviewed U.S. cases on
agency discretion following remand.
2. The claim that the Panel improperly considered non-record evidence. This
had to do with the Panel’s citation of later figures issued by the Commerce
‘0’Supporting the challenge were the U.S. Trade Representative, the ITC, and the National Pork
Producers Council; opposing the challenge were the governments of Canada and Alberta, the
Canadian Meat Council, Canadian Packers, and Moose Jaw Packers (1974) Ltd.
’02See FTA, art. 1904.13, excerpted supra, note 92.
10 3h the Matter of Fresh, Chilled or Frozen Pork from Canada (14 June 1991), 4 T.C.T. 7037
(Ex. Chall. Ctee).
ing text.
“Compare
the somewhat fuller statement of the allegations at supra, note 94 and accompany-
105See, for example, Russian Volunteer Fleet v. United States, 282 U.S. 481 (1931).
McGILL LAW JOURNAL
[Vol. 37
Department that showed that the Commissioner’s predictions about higher
countervailing duties on live swine had been erroneous. As suggested earlier,
this was probably a mistake on the part of the Panel, and the Extraordinary
Challenge Committee said so. But it probably did not affect the Panel’s basic
conclusion that Commissioner Newquist’s finding of product shifting was
unsupported by substantial evidence. In any event, even if the Panel’s look
at the Federal Register had constituted a serious departure from a fundamen-
tal rule of procedure or manifest excess of authority (see article
1904.13(a)(ii) and (iii)), which the Committee did not say, it had not “mate-
rially affected the panel’s decision” and had not threatened the integrity of
the binational review process (article 1904.13(b)), and so could not be the
basis for setting aside the Panel’s decision.
3. The claim that the Panel improperly applied a procedural rule of finality.
This had to do with the Panel’s instructions to the Commission to issue its
decision without further reopening of the record and within 21 days. The
Committee said that the Panel was within its authority to demand a final
decision after the record before the ITC had been combed “not once but
twice in search for evidence of threat of material injury.”‘ 6
4. The claim that the Panel effectively applied a “de novo standard of eviden-
tiary review” instead of the correct standard of “substantial evidence on the
record.” This again, had to do with the issue of product shifting for which
the Panel could find no evidence. The Committee pointed out that the Panel
had correctly stated the standard of review in its first decision, and had made
considerable effort in both of its decisions to determine presence or absence
of substantial evidence supporting the Commission’s decisions. The Com-
mittee was not willing to substitute its judgment for that of the expert pan-
elists on whether the Commission’s findings were supported by substantial
evidence, when it was clear that the correct standard had been conscien-
tiously applied.
5. The claim that the Panel reweighed the evidence in a manner contrary to
United States law by requiring that the ITC find “price underselling” in
order to find a likelihood of negative impact on the United States pork prices.
This had to do with the argument by the Commissioners about the hog cycle.
Petitioners claimed that the Panel had unfairly accused the Commission of
predicting underselling, for which it was conceded that there was no evi-
dence, when all the Commissioners had meant was that increased supplies
would lead to lower prices. The Committee concluded that the allegation
overstated the Panel’s finding that it had been troubled by Commissioner
Rohr’s argument; the Panel had merely mentioned the lack of evidence of
underselling as an absence of evidence of causation.
The Committee concludes that none of the allegations provide a basis for jurisdic-
tion for an extraordinary challenge under FTA Article 1904.13(a), and that none
of the alleged errors materially affected the panel decision or threaten the integrity
of the panel review process under FTA Article 1904.13(b). 0 7
106Supra, note 66 at 7017.
l0Supra, note 103 at 7042.
19921
DISPUTE SETTLEMENT & THE PORK CASE
Thus the Panel’s decision stood, the great case of Fresh, Chilled, and Fro-
zen Pork was finished (at least for the time being), and the crisis in administer-
ing the dispute settlement system under Chapter 19 of the FTA was over.
Conclusion
It is, of course, still early in the history of the FTA, though if one takes seri-
ously the provision in article 1906 that Chapter 19 is only an interim arrange-
ment for five (or at most seven) years, we are more than half way through the
try-out stage.” s I believe that everyone concerned with the FTA has learned
from the experience thus far, and I think the crisis in the Pork case may, in ret-
rospect, have proven to be a kind of cement, binding the two countries together
on their common adventure.
The Extraordinary Challenge Committee saw the danger that it would be
seen as an appellate forum, and rejected that approach firmly, and I venture to
predict decisively.
Both in the initial review and in the review of the decision on remand, all
five members of the binational panel in Pork: Injury Phase were agreed on the
principal findings. There is thus no basis for suggesting that the nationality of
the Panel members led to what turned out to be an ugly scene; on the contrary,
in both Panels in the Pork case, as well as in the Extraordinary Challenge Com-
mittee, there was no room for, and no evidence of, home town justice, party-
appointed arbitrators, or other aspects of “diplomacy by other means.”
On the merits, those persons in Canada who had never looked at their
country’s agricultural policy must have been startled that for a relatively minor
product, 36 different programs of government aid were identified, and first 18,
then eventually 15 or 16 (it is not easy to count) were held to be countervailable
subsidies. The romantic figure of Farmer Brown in a Grant Wood painting
seems ever more removed from reality –
in Canada, as in the United States,
in the European Community, and everywhere else the soil is cultivated or ani-
mals are raised for gain. On the other hand, the eventual conclusion that the
United States and Canada are indeed one market must be reassuring to those on
both sides of the border who really believe in the concept of a Free Trade
Agreement.
Whether we have learned something about the law of subsidies –
about
“specificity,” “upstream” and “downstream,” “pass-through,” and “product-
is hard to tell. Certainly the simple system of export subsidies
shifting” –
described in the textbooks, where a bonus is paid on each item (or each addi-
tional item) exported, counteracted by a duty in equal amount, is very far from
10FTA, art. 1906 provides:
The provisions of this Chapter shall be in effect for five years pending the development
of a substitute system of rules in both countries for antidumping and countervailing
duties as applied to their bilateral trade. If no such system of rules is agreed and imple-
mented at the end of five years, the provisions of this Chapter shall be extended for
a further two years. Failure to agree to implement a new regime at the end of the two-
year extension shall allow either Party to terminate the Agreement on six-month notice.
624
REVUE DE DROIT DE McGILL
[Vol. 37
the issues that had to be decided in Pork as well as in the other cases that have
come before panels under the FTA. I doubt that the drafters of the agreement on
dumping and subsidies that is supposed to replace Chapter 19 have been helped
by the experience of the panels, except to the extent that they could see the com-
plexity of the issues and the depth of the feelings involved.
Finally, though the government of Canada at one point viewed resort by the
United States government to the Extraordinary Challenge Procedure as an
unfriendly act, eight or nine months later no scars remain, as far as I can see.
If the dispute over pork –
first between the American and Canadian producers
and later between the two governments – was not exactly a lovers’ quarrel, it
does now seem more like a test of the durability of the relationship between the
two countries, and a test of the Free Trade Agreement. That test, I believe, has
been passed.
