(Cite as: 57 FR 22570, *22581)
(Ct. Int'l Trade 1985), dismissed as unappealable, 788 F.2d 1539 (Fed. Cir. 1986), vacated as moot Order dated Nov. 20, 1986
(Cabot). As we indicated in the Preliminary Determination, Cabot spawned multiple interpretations as to exactly what the court
meant in that decision. As in the Preliminary Determination, we will not go through the saga of Cabot, except to say that the
"identifiable recipients" test seemed to represent an initial attempt by the Court of International Trade (CIT) at distinguishing
government actions that benefit society generally, such as roads, bridges, schools, etc., from government actions that benefit
particular enterprises or industries. The former would be noncountervailable; the latter would be countervailable. Under this
scenario, in other words, if a firm received a check from the government with its name on it, the program pursuant to which the
firm received the check would be countervailable, notwithstanding the fact that every other firm in the country also received a
check under the program.
If this is in fact what the CIT had in mind in Cabot, it abandoned this test quickly, and in PPG the Federal Circuit ruled that such a
test was not required by the statute. The Department agrees with what the Federal Circuit said about the "identifiable recipients"
test, but we fail to see how we have applied that test here. We do not find stumpage programs to be specific merely because we can
identify the users of stumpage. Rather, we find stumpage programs to be specific because the industries using them are too few to
be
(Cite as: 57 FR 22570, *22581)
nonspecific.
Finally, Respondents refer to Georgetown Steel Corp. v. United States, 801 F.2d 1308 (Fed. Cir. 1986) (Georgetown), and its
discussion of "market distortion" and the discussion in the underlying determinations of the Department. Specificity Brief, pp.
II-18-21. Although we have read this portion of Respondents' argument several times, the relevance of this discussion to the
specificity of stumpage has not been demonstrated to us. We agree with Respondents that it is not enough that a program is
specifically provided in order for there to be a countervailable subsidy. There must also be a countervailable benefit. It appears to
us that Respondents' market distortion argument is more related to this second element, and we have addressed the argument
fully in the "Preferentiality" section below. We note that Respondents themselves point out that "(m)arket distortion is not strictly
part of the specificity test * * *" Respondents' Memorandum on Specificity, p. 18 (Feb. 19, 1992).
In conclusion, while a consideration of "inherent characteristics" or "purposeful government action" may not be precluded by the
statute, such consideration is not required by the statute. Certainly, in 1983, the Department believed that an inherent
characteristics test and, conversely, the lack of purposeful government action, were dispositive considerations in making a
finding of nonspecificity. With the amendments in the 1988 Act, however,
(Cite as: 57 FR 22570, *22581)
these considerations, at a minimum, became nondispositive.
Respondents have argued that in the absence of an inherent characteristics test, any program involving the provision of a good or
service that is not universally used is, per se, countervailable. See Hearing Transcript, p. 117. This assertion is inaccurate on at
least two counts. First, specificity is not the only requirement for a countervailable subsidy. In Department parlance, there also
must be a "countervailable benefit." See Proposed Regulations, § 355.42. Second, there are numerous instances where the
Department has not found specificity in so-called "natural resource cases." See, e.g., Portland Hydraulic Cement and Cement
Clinker from Mexico, 51 FR 44500 (1986).
However, more important than the inaccuracy of Respondents' assertion is the fact that Respondents, themselves, propose a test
which would result in a *22582
(Cite as: 57 FR 22570, *22582)
per se finding of noncountervailability. As discussed in the Preliminary Determination, Congress endorsed the Department's
finding in Carbon Black from Mexico, 51 FR 30385 (1986) (Carbon Black), that the provision of carbon black feedstock (CBFS) to
two firms was specific. Thus, Respondents have not, and cannot, dispute that if Carbon Black were decided today on the same
facts, the Department, in order to carry out Congress' intent, would have to find the provision of CBFS to be specific. Yet, if
"purposeful government action" were a prerequisite for a finding of specificity, one would have to find the provision of CBFS to be
nonspecific, and, therefore,
(Cite as: 57 FR 22570, *22582)
noncountervailable. Because both natural gas and CBFS were provided by the Government of Mexico, through PEMEX, without
limitation, there would be no basis for finding the provision of natural gas nonspecific and the provision of CBFS specific. In
neither instance was there "purposeful government action." In effect, if the Department applied the test urged by Respondents,
natural resource input subsidies would be beyond the reach of the countervailing duty law, because, except in rare cases,
they always would be found to be nonspecific. This would be true notwithstanding the fact, that in the 1988 Act Congress made the
negotiation of improved rules on these types of subsidies a principal negotiating objective of the United States. Omnibus Trade
and Competitiveness Act, Public Law 100-418, Section 1101(b)(8)(A), 102 Stat. 1122 (1988).
Thus, we conclude that the Department's use of the "inherent characteristics" test in Lumber I was ill-advised, and we decline to
follow it. This leaves the question, again assuming arguendo that Respondents' interpretation of the 1988 Act is correct, of
whether the Department may apply a revised interpretation of the statutory provisions on specificity, i.e., one that does not
include the inherent characteristics test, to the current investigation. [FN2]
FN2 If our conclusion that Congress overturned the inherent characteristics
(Cite as: 57 FR 22570, *22582)
test in the 1988 Act is correct, the following discussion is moot, because that Act made the revised provision on specificity
effective with respect to investigations initiated after August 23, 1988. Omnibus Trade and Competitiveness Act, Public Law
100-418, section 1337(b)(1), 102 Stat. 1211 (1988).
Respondents have not seriously questioned the Department's authority to change its interpretation of the statute and its
administrative practice. Respondents merely have argued that the Department must explain its reasons for changing and provide
an adequate rationale for its new position. As we have explained, regardless of the interpretation of the 1988 statutory changes,
our rationale is that the inherent characteristics test, as proposed by Respondents, leads to an absurd result: An automatic finding
of nonspecificity for all natural resource subsidies.
The final question is whether the Department may apply a revised interpretation of the statute to the instant investigation. The
federal courts have established a multifactor test to resolve questions concerning the retroactive application of changes in
administrative practice. The factors the courts consider are: (1) Whether the particular case is one of first impression; (2) whether
the new rule represents an abrupt departure from well- established practice or merely an attempt to fill a void in an unsettled area
of law; (3) the extent to which the party against whom the new rule is applied
(Cite as: 57 FR 22570, *22582)
relied on the former rule; (4) the degree of burden which a retroactive order imposes on a party; and (5) the statutory interest in
applying a new rule despite the reliance of a party on the old standard. See, e.g., District Lodge 64, Int'l Ass'n of Machinists and
Aerospace Workers, AFL-CIO v. NLRB, 949 F.2d 441 (9th Cir. 1991).
Applying these factors to this investigation, factor (1) might call for prospective application, because this is not a case of first
impression. With respect to factor (2), however, this determination does not constitute an abrupt departure from well-established
practice, because the application of the specificity test has been unsettled in general, and in the case of natural resource input
subsidies, has been particularly controversial. In particular, the review of Carbon Black clearly indicated that the 1983 finding on
specificity might not be followed. With respect to factor (3), we are aware of no evidence in the record of this case which would
indicate that the Canadian federal or provincial governments, Canadian lumber producers and exporters, or U.S. importers relied
on our 1983 finding concerning specificity as a basis for making decisions on whether to sell or buy stumpage rights, or to
produce, export, or import lumber. With respect to factor (4), we likewise are unaware of any undue burden imposed on any of
the parties, other than the normal consequences brought about by an affirmative determination of subsidization. Finally, with
respect to factor (5), we do not find it particularly relevant
(Cite as: 57 FR 22570, *22582)
given the fact that there is no evidence of reliance on the inherent characteristics test since 1986. On balance, we believe that the
Department is justified in abandoning the inherent characteristics test in this case.
Application of Specificity Factors
As stated in the Preliminary Determination, neither in the Trade Agreements Act of 1979 nor in the 1988 Act did Congress
attempt to define precisely the key phrase "specific enterprise or industry, or group of enterprises or industries." Instead,
Congress has delegated to the administering authority, currently the Department, the authority to establish the parameters of the
phrase. In this regard, the Department, in 1989, promulgated the Proposed Regulations. Section 355.43(b)(2) of the Proposed
Regulations summarized Department practice by stating that:
In determining whether benefits are specific (to an enterprise or industry, or group of enterprises or industries), the Secretary will
consider, among other things, the following factors:
(i) The extent to which a government acts to limit the availability of a program;
(ii) The number of enterprises, industries, or groups thereof that actually use a program;
(Cite as: 57 FR 22570, *22582)
(iii) Whether there are dominant users of a program, or whether certain enterprises, industries, or groups thereof receive
disproportionately large benefits under a program; and
(iv) The extent to which a government exercises discretion in conferring benefits under a program.
Respondents claim that the Department committed legal error and violated its Proposed Regulations by failing to consider fully
each of the four factors and by dismissing three of the four factors as irrelevant to its determination of specificity. [FN3]
FN3 We note that Respondents also argue, with respect to the preferentiality benchmarks outlined in the Proposed Regulations,
that such hierarchy was not intended to "prescribe an immutable formula by which all future cases should be decided." Further,
respondents argue that because the Proposed Regulations have never been promulgated in final form, the Department must give
careful consideration to other potential benchmarks that may be appropriate in particular cases (see Volume III-A of
Respondents' Case Brief). On the one hand, Respondents argue that the Proposed Regulations should be followed for purposes of
the specificity test, but that for purposes of preferentiality, the Proposed Regulations have no legal force and, therefore, the
Department should not follow them in every case without
(Cite as: 57 FR 22570, *22582)
an appropriate explanation.
The four specificity criteria are guidelines only. The Proposed Regulations state that we will consider these factors, among other
things. We agree with Respondents that no one factor is necessarily dispositive. However, we note that it is also not *22583
(Cite as: 57 FR 22570, *22583)
necessary to show all four. (See, e.g., Carbon Black at 13269.)
Respondents then point to the preamble of the Proposed Regulations, which states that the specificity test cannot be reduced to a
precise mathematical formula. Respondents misconstrue the thrust of that statement. All the Department meant was that it
cannot provide an exact numerical dividing line between specificity and nonspecificity. For example, the Department cannot
state as an absolute proposition that specificity will not exist when, say, 300 industries use a subsidy, because within this universe
of users there may be dominant users or enterprises or industries that receive disproportionate benefits. On the other hand, the
Department did not mean by its reference to the lack of a mathematical formula that in the absence of dominant users,
disproportionate use, or the exercise of discretion, a finding of specificity would be precluded if, in the Department's analysis, the
number of users (either enterprises or industries) were too few. Also, as discussed fully in the Preliminary Determination, other
factors, such as discretion, dominance, and disproportionality, were not considered in Carbon Black (which was later
(Cite as: 57 FR 22570, *22583)
endorsed by Congress), where the Department found the provision of CBFS specific to the only two enterprises that could use
CBFS.
This discussion does not mean that we have abandoned the specificity criteria in the Proposed Regulations, or that we have
ignored them in this case. On the contrary, we have considered all of them, and determine that one of them--the limited number of
users--requires a finding of specificity.
Group of Industries
As previously stated, we have determined that stumpage programs are in fact limited to a group of industries, the primary timber
processing industries. In our Preliminary Determination, we defined the primary timber processing group as comprised of two
basic manufacturing industries: Solid wood products (which includes logs) and pulp and paper products. See the Preliminary
Determination for the definitions cited in support of this conclusion. At verification, we received a British Columbia Ministry of
Forests policy paper regarding section 16.1 of the Small Business Forest Enterprise Program (see Exhibit S-11 of the BC
Verification Report). This report defines primary manufacturing as manufacturing which produces: (1) Logs; (2) timbers, defined
as softwood lumber that measures at least five inches in its least dimension (also called cants, beams, stringers, and girders); (3)
dimension lumber between two and five
(Cite as: 57 FR 22570, *22583)
inches thick (also called framing, joists, planks, rafters, etc.); (4) boards, which are less than two inches in thickness; (5) shakes
and shingles; and (6) pulp and paper, which includes kraft pulp, newsprint, linerboard, kraft paper, CTM pulp, refiner mechanical
pulp, sulfite pulp, and uncoated groundwood specialty paper. This definition further supports our conclusion that the primary
timber products group of industries is essentially two industries: The solid wood products industry and the pulp and paper
products industry.
Respondents argue that for the Department to establish that benefits have been targeted to a specific group, it is not enough that
participants of a program constitute an "identifiable" group merely by virtue of their being participants. Rather, some independent
characteristics must define the "specific" group.
In the Preliminary Determination, we placed excessive emphasis on what the two industries had in common, i.e., the milling
operation. In fact, we have consistently held that for a small number of industries (in this case two) to be considered a group of
industries for purposes of the countervailing duty law, there is no requirement for commonality between the industries.
Therefore, even though the two industries which make up the group of industries in this case may have many common features,
such as the same input, timber; and many of the same manufacturing processes, these common features in no way are
determinative of whether the two industries may be considered a group of
(Cite as: 57 FR 22570, *22583)
industries within the meaning of the Act. (See, e.g., Structural Shapes and Cold-Rolled Carbon Steel Flat-Rolled Products from
Korea, 49 FR 47284 (December 3, 1984).)
In the Preliminary Determination, although we combined two industries into one group, we could just as well have called the
beneficiaries of stumpage two industries, the solid wood products industry and the pulp and paper products industry, and we
would still have found stumpage to be specific. However, given the wording of section 771(5)(A)(ii) of the Act "* * * if provided to
a specific enterprise or industry, or group of enterprises or industries * * *," anything more than a single industry constitutes a
"group." Therefore, if, hypothetically, a program were available to only the chemical industry, the steel industry, and the dairy
industry, these three industries would constitute a "group," within the meaning of the Act, even though the three clearly have
nothing in common.
However, in this case, we have identified a group of industries that have many elements in common. The solid wood products
industry (including logs) is an industry because it uses the same input, timber, and produces it into a solid wood product such as
lumber, plywood, veneer, poles and posts, and shakes and shingles. The pulp and paper products industry also uses the same
input, timber, and uses one of two methods to produce pulp: (1) The mechanical process; and (2) the chemical process, (see, the
BC Verification Report, p. 5).
(Cite as: 57 FR 22570, *22583)
Respondents also argue that the Department has not justified its definition of a "group of industries" and must apply the accepted
definition of the word "industry" which, according to Respondents, is generally described under U.S. trade law in terms of
products. They state that a product-based analysis is consistent with Article 6 of the GATT Subsidies Code, which provides that
the relevant inquiry in an injury investigation is into the volume of subsidized imports and their effect on prices in the domestic
market of the like product.
We agree that an important aspect of our industry groupings for purposes of defining standing is the identification of the industry
that produces the "like product" in the United States. That is, we are concerned with the competing U.S. industry when defining the
foreign industry. In this case, the Coalition is predominantly made up of lumber mills. These lumber mills in the United States may
also produce the number and variety of products produced by stumpage holders in Canada, such as tissue paper and
cardboard boxes, but the like product, and the product that the ITC is investigating as a potential cause of material injury, is
lumber, not tissue paper or cardboard boxes. As explained below, however, this aspect of industry grouping is not determinative
of the way "industries" are identified for purposes of specificity.
Respondents argue that, in defining the industry, the Department inappropriately focused on the raw material rather than the
actual number and variety of products made by the industries that hold stumpage rights. They
(Cite as: 57 FR 22570, *22583)
contend that companies and divisions of companies holding stumpage rights produce a variety of downstream products, other
than what they consider to be "primary timber" products, and that these products fall into several "groups of industries" according
to the Canadian and United *22584
(Cite as: 57 FR 22570, *22584)
States Standard Industrial Classification (SIC) codes and, therefore, are beyond the Department's definition of the primary timber
products group of industries.
In support of this contention, Respondents conducted a survey for purposes of this investigation in which companies were asked
to indicate which products they produce, excluding products produced by subsidiaries, affiliates, and joint-ventures, (see the BC,
Alberta, and Quebec Verification Reports). In Ontario, Respondents conducted a survey of tenure holders to determine, what
products, other than lumber, tenure holders produce (see Ontario Verification Report, p. 14).
Our emphasis on the raw material used is relevant insofar as the program under investigation is the provision of a raw material to
users of that raw material. Our focus remains the two industries which receive the good provided by the government (stumpage).
Regarding the use of an "accepted" definition of the term "industry," we believe that there is a realm of acceptable definitions of the
term "industry." We agree that the definition cited by Respondents ("The term 'industry' means the domestic producers as a whole
of a like product * * *.") is relevant to,
(Cite as: 57 FR 22570, *22584)
for example, an analysis of whether the U.S. industry is injured by subsidized imports or whether the U.S. industry has standing to
bring a petition. It is not relevant to our analysis of what constitutes an industry or a group of industries. For example, in Certain
Fresh Cut Flowers from Canada, Chile, Colombia, Costa Rica, Ecuador, Israel, and the Netherlands, USITC, Pub. 1956, Inv.
Nos. 701-TA-275 to -278 (March 1987) (final), the ITC found that each of the seven different types of flowers (carnations,
chrysanthemums, etc.) constituted a separate like product and thus a separate industry. Using Respondents' logic, the
Department should find that, because there were seven domestic industries identified by the ITC in its like product analysis,
flowers constitute more than a group of industries. Such an analysis would lead to absurd results, at odds with years of
Department precedents and several court rulings (see below). See also, Anti-Friction Bearings (Other Than Tapered Roller
Bearings) and Parts Thereof from the Federal Republic of Germany, France, Italy, Japan, Romania, Singapore, Sweden, Thailand,
and the United Kingdom, USITC, Pub. 2185, Inv. Nos. 303-TA-19 and 20, 731-TA-391 to -399 (May 1989) (final).
The Department, in its application of the specificity test, has viewed the term "industry" in much broader terms. In the Final
Affirmative Countervailing Duty Determination: Certain Fresh Cut Flowers from the Netherlands 52 FR 3301, 3312
(February 3, 1987), we explicitly rejected the argument that benefits
(Cite as: 57 FR 22570, *22584)
provided to the horticulture and greenhouse industries were too broad to be considered specific because they were provided to
over 60 different categories of products. In that case we stated:
Although the horticulture and greenhouse industries contain many separable categories of products, the Department considers
these industries to be a specific subset of all agriculture and not so broad as to consider them more than a specific enterprise or
industry.
Moreover, in many countervailing duty investigations and administrative reviews, where we have found programs
to be nonspecific because they were available to a wide variety of industries, we have used the term "industry" as broader
industrial categories such as "food, steel, non-ferrous metals, machinery, wood products, textiles, rubber, chemical and paper
industries," taken together, to describe the nonspecific universe of industries receiving benefits (Final Affirmative
Countervailing Duty Determination; Carbon Steel Wire Rod from Malaysia 53 FR 13303, 13305 (April 22, 1988)). The
nonspecific universe is not defined by the number of products produced by each of these industrial categories.
With respect to Respondents' product-based argument, the alleged end products produced by the Canadian companies and
divisions of these companies are merely the downstream products that can be made from the base products, i.e., solid wood and
pulp (we note that paper cannot be produced until after
(Cite as: 57 FR 22570, *22584)
pulp is produced). The same situation exists for other groups of basic industries as well. For example, if a steel company produces
flat-rolled sheet as well as other downstream products made from the sheet, such as machine parts, and this company receives a
benefit that is also provided to all steel producers, the fact that this and other steel producers produce a number of downstream
steel products would not negate the fact that the benefit was provided to encourage the production of steel. Therefore, the
program would be specific to the steel industry.
Furthermore, with respect to the end product surveys used as support for this argument, we found at verification that many of the
"end products" listed by the companies were not products actually made by that company or divisions of that company. For
example, as explained in the Verification Report of Canadian Forest Products, Ltd. (CANFOR) in BC, the invoices showed that the
company sold logs and lumber to unrelated companies for the intended manufacture of a certain end product. The company
indicated on its survey response that it made products that it actually did not make, e.g., the company indicated that it produced
flooring (hence Respondents' contention that the "flooring industry" should be considered a separate industry benefitting from
stumpage). However, a review of the sales invoice showed that the company produced dimension lumber, not flooring. Likewise,
West Fraser, another BC company, indicated that it produced certain end products such as "appearance framing." Appearance
(Cite as: 57 FR 22570, *22584)
framing is lumber of a particular size and quality, and is included in the scope of this investigation. Furthermore, officials from
both companies stated that all of the end products "produced" by their companies are made from either a solid wood product such
as logs, lumber, or plywood, or from pulp and paper products. (See, e.g., pp. 14 and 15 of the BC Public Verification Report.)
Our review of the end product survey in Alberta produced similar results. One survey response indicated that the stumpage
holder produced chicken coops. However, at verification, we found that the producer did not sell chicken coops. (See p. 33 of the
Alberta Verification Report.) In Quebec, the end product survey showed that the overwhelming majority of stumpage holders
produced either traditional paper products or traditional sawmill products. We found only two companies out of the 124 that
produced other types of products. Further, we were not able to trace some of the items Respondents claimed were produced by
stumpage holders to any of the individual end product responses.
Therefore, even if we agreed (which, for the reasons stated above, we do not), with Respondents' argument that downstream
products produced by companies holding stumpage leads to the inevitable conclusion that there are groups of industries that are
not included in the Department's definition of a group of industries, based on the facts on the record, Respondents have not
demonstrated that companies which hold stumpage rights actually produce the wide variety of products indicated in the survey
responses.
(Cite as: 57 FR 22570, *22584)
Respondents also contend that because some stumpage holders only produce logs, which according to Respondents are not
processed at all, these producers do not fit into the Department's definition of a group of *22585
(Cite as: 57 FR 22570, *22585)
industries. First, we do not agree that logs are not processed. The only timber that is not processed at all is standing timber. Once
the standing timber is harvested, the processing of that standing timber has begun. Furthermore, as previously stated, we placed
excessive emphasis in the Preliminary Determination on the common features necessary to find the solid wood products industry
and pulp and paper products industry to be a single group of industries. One of those common features we referred to is the
milling process. It is absurd to suggest that a log producer does not benefit from stumpage.
Respondents also argue that the Department should use the definition of groups of industries as set forth in the SIC, a
product-based classification system, as the Department did in Lumber I. Further, they contend that all of the end products
"produced" by stumpage holders fall into, at a minimum, 27 groups of industries at the four-digit level, according to both the
Canadian and U.S. SIC. Respondents cite the Final Negative Countervailing Duty Determination: Certain Granite Products
from Italy 53 FR 27197 (July 19, 1988) (Granite), in support of their assertion that the Department "generally respects and
applies" a country's own classification system in classifying and counting industries.
(Cite as: 57 FR 22570, *22585)
In Granite, we found a loan program not countervailable because it was used by virtually every productive sector in the country,
and we listed these productive sectors according to Italy's own statistical categories. Nowhere does this case imply that the
Department generally uses a country's own classification system for purposes of applying the specificity test. In fact, the
Department has frequently used other means besides SIC classifications to define a group of industries (or an industry). For
example, we stated "SIC classification was not dispositive of the question of whether software design was in the service industry or
the manufacturing industry" in Certain Computer Aided Software Engineering Products from Singapore, 55 FR 12248 (April 2,
1990).
As stated in our Preliminary Determination, the use of the word "group" in the United States and Canadian SIC does not interfere
with the meaning of "group of industries" as used in the Act. Moreover, there is no evidence that Congress intended the term
"group," as used in section 771(5), to equate to SIC terminology.
We note, however, that the International SIC (ISIC) codes classifies the wood products industry as one industry at the four-digit
level, which includes wooden railway sleepers, coniferous sawn wood, veneer sheets, plywood, and particle board. The ISIC also
classifies the pulp and paper products industry as one industry at the four-digit level. This classification includes products
(Cite as: 57 FR 22570, *22585)
such as mechanical wood pulp, pulp of fibers other than wood, newsprint, other printing and writing paper, kraft paper and kraft
paperboard, cigarette papers, and fiberboard. The ISIC's use of the term industry underscores the fact that the term "industry" can
be used in a variety of ways even within the different SIC regimes.
We also note that the U.S. SIC codes classify the steel industry into separate major groups that are in no way related to the manner
in which we have defined the steel industry in previous cases. Further, the ISIC classifies the steel industry as one industry, which
includes products such as pig iron, wire rod, plates, and sheets.
We maintain that SIC codes were not intended by Congress to be the dispositive definition of industry for purposes of
administering the countervailing duty law any more than our tariff classifications are dispositive in defining the scope of a
countervailing duty proceeding. As pointed out, even within the different SIC regimes, the classifications of products differ.
Respondents also contend that economic integration or interdependence should not affect the Department's analysis of industries,
and that a company producing both a solid wood product and a pulp and paper product does not preclude that company's
participation in more than one industry. We agree with Respondents that the level of integration is not dispositive with regard to
the number of industries represented by the integrated firm. Just as we do not
(Cite as: 57 FR 22570, *22585)
hold that two industries found to be a group of industries must have common features, they also need not be integrated.
Respondents argue that in nine years the Department has come up with three different definitions describing the industries which
use stumpage. They allege that the industry has not changed in that time, only the Department's definition has changed in order to
reach a finding a specificity.
In Lumber I, we found that stumpage programs were not specific because:(1) Any limitation on use was not a result of government
action, but rather was due to the inherent nature of the products under investigation, and (2) stumpage was used by several
specifically-named groups of industries (the lumber and wood products industries, the pulp and paper industries, and the
furniture industries). Although we found that nonstumpage benefits provided to the forests products industries were specific, in
the case of stumpage, we reasoned that a finding of no specificity was warranted because the universe of users of stumpage was
limited by the inherent characteristics and uses of raw timber.
We agree with Respondents that the users of stumpage have not changed dramatically in the past nine years. However, we note
that our definition of the group of industries which use stumpage has not changed as radically as Respondents suggest in each of
the three Canadian lumber investigations. For example, the industries in Lumber I were broken down into lumber and wood
products industries (solid wood products), pulp and paper industries, and
(Cite as: 57 FR 22570, *22585)
furniture manufacturing industries. In Lumber II, we stated that we no longer believed that furniture manufacturers should be
included in our analysis of the group of industries because such manufacturers held negligible stumpage rights. We also called into
question the earlier conclusion that stumpage rights were not in fact limited to one group of industries. Furthermore, we pointed
out in Lumber II that our reference in Lumber I to SIC codes was misplaced. We still consider there to be two industries that use
stumpage: the solid wood products and the pulp and paper products industries. The only thing that has changed during the past
nine years is our consideration of inherent characteristics and how that consideration affects our specificity determination.
With respect to Lumber I, the users of stumpage did not undergo some transformation that reduced them from three groups of
industries when cutting stumpage to one group when borrowing money. The use of two different descriptions of exactly the same
users of stumpage in that notice, leading to a finding of specificity for nonstumpage programs and a finding of no specificity for
stumpage programs, was inconsistent on its face and underscores the fact that, in 1983, the Department was entering into
uncharted legal waters.
The Courts have recognized that the "application of the de facto aspect of the specificity test requires a 'case-by-case' analysis to
determine whether 'there has been a bestowal upon a specific class."' Cabot I, as quoted in PPG
(Cite as: 57 FR 22570, *22585)
(emphasis added). The critical focus of a determination of specificity must be an *22586
(Cite as: 57 FR 22570, *22586)
analysis of whether a benefit "has been bestowed on a discrete class of grantees despite nominal availability, program grouping, or
the absolute number of grantee companies or industries." Roses, Inc., California Floral Trade Council and Floral Trade Council v.
United States, 743 F. Supp. 870 (1990) (emphasis added).
In this investigation, stumpage is clearly provided to a specific class of beneficiaries: the pulp and paper products and solid wood
products industries. While the statute uses the terms "enterprises or industries or groups of enterprises or industries," the courts
have made clear that these terms are not to be applied in a narrow manner. The concept of a "discrete class of beneficiaries," as
used by the court in Cabot I and Roses, Inc., is broader than any of the narrow, forced definitions that the Respondents would
have us use, and the concept is consistent with the Department's long-standing practice in implementing the specificity test. (See
also, Final Affirmative Countervailing Duty Determination Certain Fresh Atlantic Groundfish from Canada 51 FR
10041 (March 24, 1986) (Groundfish) and Netherlands Flowers.)
In addition, Respondents argue that because the forestry sector in Ontario and BC is larger than the agriculture sector in each of
those provinces, and because the Department has determined that it will not regard a program as being specific solely because the
program is limited to agriculture, it is
(Cite as: 57 FR 22570, *22586)
absurd to find that a program available to all of a sector that is larger than agriculture is specifically provided.
We disagree. A program is not necessarily specific if it is limited to agriculture because all of agriculture cannot be considered a
discrete class and is certainly more than several groups (see, e.g., Final Negative Countervailing Duty Determination; Fresh
Asparagus from Mexico 48 FR 21618 (May 13, 1983)). In agriculture, there are a large number of distinct industries producing a
vast number of products from many raw materials, growing methods, production and distribution techniques. In response to a
comment in Groundfish, the Department stated that fisheries programs were bestowed to "two specific industries, the salt water
fishing industry and the seafood products industry, as such, these programs were available to no more than a group of industries."
Although the saltwater fishing industry and the seafood products industry produce a large number of products (arguably at least
as large as the pulp and paper and solid wood products industries), the Department rejected the argument that the fisheries sector
is like agriculture, so broad as to not constitute "a specific group of industries."
Similarly, regardless of the size of its economic contribution in a particular political jurisdiction, forestry is not like agriculture as
a sector, because all forestry products come from one basic raw material, timber. The number and diversity of products produced
in the forest industries
(Cite as: 57 FR 22570, *22586)
are simply not comparable to those produced in agriculture. Therefore, the Department does not regard a program as nonspecific
if it is available throughout the forestry sector.
In fact, agriculture is defined as "The production of plants and animals useful to human beings, including the cultivation of soil,
management of crops, and the feeding, breeding, and managing of livestock." [FN4] Arguably, therefore, forestry could be
considered part of the agricultural sector. If stumpage programs were analyzed in terms of whether a specific group of industries
within agriculture had received benefits, the Department would most certainly consider the program specific. Thus, regardless of
whether forestry is considered as part of agriculture, even if a program is available throughout the forestry sector it is still
specific. Because we have found stumpage programs limited to a group of industries, we find them specifically provided within the
meaning of the Act.
FN4 McGraw-Hill Encyclopedia of Science and Technology, McGraw-Hill Book Company, p. 143.
Having determined that the stumpage programs administered by Alberta, BC, Manitoba, Ontario, Quebec, Saskatchewan, and the
Territories are provided to the primary timber processing industries group, the next section addresses whether stumpage is
provided at preferential rates.
(Cite as: 57 FR 22570, *22586)
Preferentiality
Having determined that the stumpage programs administered by Alberta, BC, Manitoba, Ontario, Quebec, Saskatchewan, and the
Territories are specific within the meaning of the statute, the Department must next determine whether stumpage is provided
under these programs at preferential rates pursuant to section 771(A) (ii) (II) of the Act. The standard employed by the
Department for making this determination is set forth in § 355.44(f) of the Proposed Regulations, which itself reflects an approach
first articulated in detail by the Department in its Preferentiality Appendix (Preferentiality Appendix to the Department's
Preliminary Determination in the Administrative Review of Carbon Black from Mexico, 51 FR 13271 (1986)). Section
355.44(f) reads as follows:
(1) Provision of goods or services at preferential rates. The provision by a government of a good or service pursuant to a domestic
program confers a countervailable benefit to the extent the Secretary determines that the price charged by the government for
the good or service is less than the benchmark price, which normally will be the nonselective prices the government charges to
the same or other users of the good or service within the same political jurisdiction.
(2) Where the Secretary determines that there is no benchmark price under
(Cite as: 57 FR 22570, *22586)
paragraph (f) (1) which is not selective within the meaning of § 355.43, the Secretary will determine the existence of a
countervailable benefit based upon, in order of preference, the following alternative benchmarks:
(i) The price, adjusted for any cost differences, the government charges for a good or service which is similar or related to the
good or service in question, provided that the similar or related good or service and its price is (sic) not selective within the
meaning of section 355.43;
(ii) The price charged by other sellers to buyers within the same political jurisdiction for an identical good or service;
(iii) The government's cost of providing the good or service; or
(iv) The price paid for the identical good or service outside of the political jurisdiction in question.
In the Preliminary Determination, the Department relied on the traditional measure of preference--price discrimination-- to
determine preliminarily that stumpage was being provided at preferential rates to softwood lumber producers in the provinces of
BC, Ontario and Alberta. In Quebec, the Department found preliminarily that stumpage was being preferentially provided to
softwood lumber producers on the basis of a comparison with prices charged by private sellers of stumpage within Quebec, as
there was not an adequate basis to make a determination based on the provincial government's sales of an identical or
(Cite as: 57 FR 22570, *22586)
similar good. Finally, with respect to the stumpage programs of Manitoba, Saskatchewan and the Territories, the Department did
not reach the question of preferentiality because the small amount of exports from these provinces and territories meant that use
of even the highest estimates of subsidization on the record would have led to no more than a de minimis effect on the country-
wide rate.
In this final determination, the Department confirms its preliminary finding that stumpage is being provided to softwood lumber
producers in the provinces of Alberta, BC, Ontario and *22587
(Cite as: 57 FR 22570, *22587)
Quebec at preferential rates. Furthermore, the Department has determined that the benchmarks it applied in the Preliminary
Determination for purposes of identifying and measuring preferentiality remain the appropriate determinants of whether
preference and, hence, a countervailable benefit exists in these circumstances (although, as explained below, refinements have
been made in our comparisons on the basis of the results of verification and our analysis of parties' comments).
However, before turning to our explanation of the specific bases for finding preferentiality in the context of each provincial
experience, we note that the parties to this proceeding have made several arguments of a more fundamental and cross-cutting
nature than whether a particular benchmark of the proposed § 355.44(f) (1) has been correctly selected or applied. These
arguments call into question the very standard followed by the Department in determining
(Cite as: 57 FR 22570, *22587)
preference and (even more fundamentally) whether stumpage programs administered in the manner which Canadian provincial
governments currently administer them can ever be found to confer countervailable subsidies, irrespective of whether
preference has been exercised. Because these arguments go to the very heart of whether and/or how the provision of a natural
resource good can constitute a subsidy, and because our acceptance of any of these arguments would have radically altered the
manner in which we addressed the preferentiality issue, we address these arguments first.
Legal Analysis
Respondents argue that the provision of stumpage by the Canadian provinces does not constitute a countervailable subsidy
because there is no market distortion in this case. Although Respondents' economic argument is described in more detail below,
Respondents essentially rely on an analysis submitted by Prof. William D. Nordhaus to support the proposition that stumpage
charges are in the nature of economic rents. Because of this fact, they argue, stumpage charges cannot result in increased
production (i.e.,a market distortion) unless the provinces confer net harvest-related benefits on tenureholders. They then claim
that there has been no showing in this case that stumpage charges confer such net benefits. Respondents then cite the
Department's determination in