Countervailing Duties

SUMMARY: The Department of Commerce (``the Department'') hereby issues

final countervailing duty regulations to conform to the Uruguay Round

Agreements Act, which implemented the results of the Uruguay Round

multilateral trade negotiations. The Department has sought to issue

regulations that: Where appropriate and feasible, translate the

principles of the implementing legislation into specific and

predictable rules, thereby facilitating the administration of these

laws and providing greater predictability for private parties affected

by these laws; simplify and streamline the Department's administration

of countervailing duty proceedings in a manner consistent with the

purpose of the statute and the President's regulatory principles; and

codify certain administrative practices determined to be appropriate

under the new statute and under the President's Regulatory Reform

Initiative.

DATES: The effective date of this final rule is December 28, 1998,

except that Sec. 351.301(d) is effective on November 25, 1998. See

Sec. 351.702 for applicability dates.

FOR FURTHER INFORMATION CONTACT: Jennifer A. Yeske at (202) 482-1032 or

Jeffrey May at (202) 482-4412.

SUPPLEMENTARY INFORMATION:

Background

The publication of this notice of final rules, which deals with

countervailing duty (``CVD'') methodology, completes a significant

portion of the process of developing regulations under the Uruguay

Round Agreements Act (``URAA''). The process began when the Department

took the unusual step of requesting advance public comments in order to

ensure that, at the earliest possible stage, we could consider and take

into account the views of the private sector entities that are affected

by the antidumping (``AD'') and CVD laws. On February 26, 1997, the

Department published proposed rules dealing with CVD methodology

(``1997 Proposed Regulations''). The Department received over 200

written public comments regarding the 1997 Proposed Regulations. On

October 17, 1997, the Department held a public hearing, and thereafter,

received over 50 additional post-hearing written public comments on the

1997 Proposed Regulations.1

---------------------------------------------------------------------------

\1\ The prior notices published by the Department as part of its

URAA rulemaking activity are: (1) Advance Notice of Proposed

Rulemaking and Request for Public Comments (Antidumping Duties;

Countervailing Duties; Article 1904 of the North American Free Trade

Agreement), 60 FR 80 (January 3, 1995); (2) Advance Notice of

Proposed Rulemaking; Extension of Comment Period (Antidumping

Duties; Countervailing Duties; Article 1904 of the North American

Free Trade Agreement), 60 FR 9802 (February 22, 1995); (3) Interim

Regulations; Request for Comments (Antidumping and Countervailing

Duties), 60 FR 25130 (May 11, 1995); (4) Proposed Rule; Request for

Comments (Antidumping and Countervailing Duty Proceedings;

Administrative Protective Order Procedures; Procedures for Imposing

Sanctions for Violation of a Protective Order), 61 FR 4826 (February

8, 1996); (5) Notice of Proposed Rulemaking and Request for Public

Comments (Antidumping Duties; Countervailing Duties), 61 FR 7308

(February 27, 1996); (6) Extension of Deadline to File Public

Comments on Proposed Antidumping and Countervailing Duty Regulations

and Announcement of Public Hearing (Antidumping Duties;

Countervailing Duties), 61 FR 18122 (April 24, 1996); (7)

Announcement of Opportunity to File Public Comments on the Public

Hearing of Proposed Antidumping and Countervailing Duty Regulations

(Antidumping Duties; Countervailing Duties), 61 FR 28821 (June 6,

1996); (8) Notice of Proposed Rulemaking and Request for Public

Comment (Countervailing Duties), 62 FR 8818 (February 26, 1997); (9)

Final Rules (Antidumping Duties; Countervailing Duties), 62 FR 27295

(May 19, 1997); (10) Extension of Deadline to File Public Comments

on Proposed Countervailing Duty Regulations, (Countervailing

Duties), 62 FR 19719 (April 23, 1997); (11) Extension of Deadline to

File Public Comments on Proposed Countervailing Duty Regulations,

(Countervailing Duties), 62 FR 25874 (May 12, 1997); (12) Notice of

Public Hearing on Proposed Countervailing Duty Regulations and

Announcement of Opportunity to File Post-Hearing Comments,

(Countervailing Duties), 62 FR 38948 (July 21, 1997); (13) Notice of

Public Hearing on Proposed Countervailing Duty Regulations and

Announcement of Opportunity to File Post-Hearing Comments;

Correction, (Countervailing Duties), 62 FR 41322 (August 1, 1997);

(14) Notice of Postponement of Public Hearing on Proposed

Countervailing Duty Regulations and of Opportunity to File Post-

Hearing Comments, (Countervailing Duties), 62 FR 46451 (September 3,

1997); (15) Interim Final Rules; Request for Comments (Procedures

for Conducting Five-Year (``Sunset'') Reviews of Antidumping and

Countervailing Duty Orders), 63 FR 13516 (March 20, 1998); and (16)

Final Rule; Administrative Protective Order Procedures; Procedures

for Imposing Sanctions for Violation of a Protective Order,

(Antidumping and Countervailing Duty Proceedings), 63 FR 24391 (May

4, 1998).

---------------------------------------------------------------------------

In drafting these final rules, the Department has carefully

reviewed and considered each of the comments it received. While we have

not always adopted suggestions made by commenters, we found the

comments to be very useful in helping us to work our way through the

many legal and policy issues addressed in the regulation. Therefore, we

are extremely grateful to those who took the time and trouble to

express their views regarding how the Department should administer the

CVD laws in the future.

In addition, in these final rules, the Department has continued to

be guided by the objectives described in the 1997 Proposed Regulations.

Specifically, these objectives are: (1) Conformity with the statutory

amendments made by the URAA; (2) the elaboration through regulation of

certain statements contained in the Statement of Administrative Action

(``SAA''); 2 and (3) consistency with President Clinton's

Regulatory Reform Initiative and his directive to identify and

eliminate obsolete and burdensome regulations.

---------------------------------------------------------------------------

\2\ See Statement of Administrative Action accompanying H.R.

5110, H.R. Doc. No. 316, Vol. 1, 103d Cong., 2d Sess. 911-955

(1994).

---------------------------------------------------------------------------

In the case of CVD methodology, the Department previously issued

proposed regulations in 1989 (``1989 Proposed

Regulations'').3 Because the Department never issued final

rules, the 1989 Proposed Regulations were not binding on the Department

or private parties. Nevertheless, to some extent both the Department

and private parties relied on the 1989 Proposed Regulations as a

restatement of the Department's CVD methodology as it existed at the

time. Thus, notwithstanding statutory amendments made by the URAA and

subsequent developments in the Department's administrative practice,

the 1989 Proposed Regulations still serve as a point of departure for

any new regulations dealing with CVD methodology.

---------------------------------------------------------------------------

\3\ See Notice of Proposed Rulemaking and Request for Public

Comments (Countervailing Duties), 54 FR 23366 (May 31, 1989).

---------------------------------------------------------------------------

In an earlier rulemaking (see item 9 in note 1), we consolidated

the AD and CVD regulations into a single part 351. For the most part,

the regulations contained in this notice constitute subpart E of part

351.

Explanation of the Final Rules

In drafting these Final Regulations, the Department carefully

considered each of the comments received. In addition, we conducted our

own independent review of those provisions of the 1997 Proposed

Regulations that were not the subject of public comments. The following

sections contain a summary of the comments we received and the

Department's responses to those comments. In addition, these sections

contain an explanation of changes the Department has made to the 1997

Proposed Regulations either in response to

[[Page 65349]]

comments or on its own initiative. Finally, these sections contain a

restatement of principles that remain unchanged from the 1997 Proposed

Regulations and that were not the subject of any public comments.

The Department is also hereby issuing interim final rules to set

forth certain procedures for establishing the non-countervailable

status of alleged subsidies or subsidy programs pursuant to section

771(5B) of the Tariff Act of 1930, as amended (``the Act''). Pursuant

to authority at 5 U.S.C. 553(b)(A), the Assistant Secretary for Import

Administration waives the requirement to provide prior notice and an

opportunity for public comment because this action is a rule of agency

procedure. This interim final rule is not subject to the 30-day delay

in its effective date under 5 U.S.C. 553(d) because it is not a

substantive rule. The analytical requirements of the Regulatory

Flexibility Act (5 U.S.C. 601 note) are inapplicable to this rulemaking

because it is not one for which a Notice of Proposed Rulemaking is

required under 5 U.S.C. 553 or any other statute.

Section 351.102

These regulations add several definitions to Sec. 351.102. Many of

these definitions are identical (or virtually identical) to definitions

contained in Sec. 355.41 of the 1989 Proposed Regulations, and some are

based on definitions contained in the Illustrative List of Export

Subsidies (``Illustrative List'') annexed to the Agreement on Subsidies

and Countervailing Measures (``SCM Agreement''). We have made some

changes to the definitions contained in the 1997 Proposed Regulations.

While we have not changed the definition of consumed in the

production process, we are clarifying that the definition is not to be

used as a way to expand significantly the rights of countries to apply

border adjustments for a broad range of taxes on energy, particularly

in the developed world. See SAA at 915.

The definition of firm is based on Sec. 355.41(a) of the 1989

Proposed Regulations, but an additional clause has been added to

clarify that the purpose of this term is to serve as a shorthand

expression for the recipient of an alleged subsidy. While other terms

could be used, the use of the term ``firm'' in this manner has become

an accepted part of CVD nomenclature. For clarification, we have added

``company'' and ``joint venture'' to the entities listed in the

definition in the 1997 Proposed Regulations.

Similarly, the term government-provided is used as a shorthand

adjective to distinguish the act or practice being analyzed as a

possible countervailable subsidy from the act or practice being used as

a benchmark. As made clear in the regulation, the use of ``government-

provided'' does not mean that a subsidy must be directly provided by a

government. This definition is unchanged from our 1997 Proposed

Regulations.

As in our 1997 Proposed Regulations, loan is defined to include

forms of debt financing other than what one normally considers to be a

``loan,'' such as bonds or overdrafts. Again, this definition is

intended as a shorthand expression in order to avoid repetitive use of

more cumbersome phrases, such as ``loans or other debt instruments.''

In this regard, the Department considered codifying its approach

with respect to so-called ``hybrid instruments,'' financial instruments

that do not readily fall into the basic categories of grant, loan, or

equity. In the 1993 steel determinations (see Certain Steel Products

from Austria (General Issues Appendix), 58 FR 37062, 37254 (July 9,

1993) (``GIA'')), the Department developed a hierarchical approach for

categorizing hybrid instruments, an approach that was sustained in

Geneva Steel v. United States, 914 F. Supp. 563 (CIT 1996). However,

notwithstanding this judicial imprimatur, the Department has relatively

little experience with hybrid instruments. Therefore, although the

Department has no present intention of deviating from the approach set

forth in the GIA, the codification of this approach in the form of a

regulation would be premature at this time.

Many commenters proposed definitions of the phrase ``entrusts or

directs'' as it is used in section 771(5)(B)(iii) of the Act, which

deals with ``indirect subsidies.'' Indirect subsidies generally involve

situations where a government provides a financial contribution through

a private body. Under section 771(5)(B)(iii) of the Act, a subsidy

exists when, inter alia, a government ``makes a payment to a funding

mechanism to provide a financial contribution, or entrusts or directs a

private entity to make a financial contribution * * *'' (emphasis

added). In our 1997 Proposed Regulations, we did not address indirect

subsidies in detail. Instead, we noted that the SAA directs the

Department to proceed on a case-by-case basis (see SAA at 925-26), and

we requested comments on the factors we should consider in making our

case-by-case determinations.

One commenter suggested that an indirect subsidy need only be

linked to a government action or program to satisfy the ``entrusts or

directs'' standard. This same commenter asked the Department to include

an illustrative list of situations that would meet the ``entrusts or

directs'' standard. A second commenter believed that the standard is

met when a government takes an action that causes a private party to

confer a benefit. This same commenter asked the Department to clarify

that the term ``private body'' is not limited to a single entity, but

also includes a group of entities or persons. A third commenter

proposed that the ``entrusts or directs'' standard be considered

satisfied whenever a government takes an action that proximately

results in a private entity providing a financial contribution. Certain

commenters also asked the Department to confirm that the standard is no

narrower than the prior U.S. standard for finding an indirect subsidy.

The issue of what ``entrusts or directs'' means was debated

extensively at the Department's hearing on its 1997 Proposed

Regulations. This debate prompted the submission of additional proposed

definitions. Two commenters argued that an indirect subsidy occurs

whenever a government action has the inevitable result of compelling a

private party to provide a benefit. A second commenter proposed a ``but

for'' test, i.e., if the government did not act, the subsidy would not

exist.

As the extensive comments on this issue indicate, the phrase

``entrusts or directs'' could encompass a broad range of meanings. As

such, we do not believe it is appropriate to develop a precise

definition of the phrase for purposes of these regulations. Rather, we

believe that we should follow the guidance provided in the SAA to

examine indirect subsidies on a case-by-case basis. We will, however,

enforce this provision vigorously.

We agree with those commenters who urged the Department to confirm

that the current standard is no narrower than the prior U.S. standard

for finding an indirect subsidy as described in Certain Steel Products

from Korea, 58 FR 37338 (July 9, 1993) and Certain Softwood Lumber

Products from Canada, 57 FR 22570 (May 28, 1992). Also, we believe that

the phrase ``entrusts or directs'' subsumes many elements of the

definitions proposed by commenters. With respect to the suggestion that

we include an illustrative list of situations that would fall under the

``entrusts or directs'' standard, we do not believe this is necessary.

The SAA at 926 lists a number of cases where the Department

[[Page 65350]]

has found indirect subsidies in the past, and these cases serve to

provide examples of situations where we believe the statute would

permit the Department to reach the same result. Similarly, regarding

the request that we define the phrase ``private entity'' to include

groups of entities or persons, the SAA is clear that groups are

included (see SAA at 926). Therefore, we have not promulgated a

regulation with this definition.

Although the indirect subsidies that we have countervailed in the

past have normally taken the form of a foreign government requiring an

intermediate party to provide a benefit to the industry producing the

subject merchandise, often to the detriment of the intermediate party,

indirect subsidies could also take the form of a foreign government

causing an intermediate party to provide a benefit to the industry

producing the subject merchandise in a way that is also in the interest

of the intermediate party. We believe the phrase ``entrusts or

directs'' could encompass government actions that provide inducements,

other than upstream subsidies, to a private party to provide a benefit

to another party.

One commenter argued that the Final Regulations should include a

definition of consultations. Consistent with Article 13 of the SCM

Agreement, section 702(b)(4)(A)(ii) of the Act requires the Department

to provide the government of the exporting country named in a petition

an opportunity for consultations with respect to the petition. This

commenter suggested that the definition of consultations should include

a statement of purpose as articulated in the SCM Agreement (i.e.,

clarifying the allegations in the petition and arriving at a mutually

agreed solution). Furthermore, the commenter argued, in the Final

Regulations the Department should commit to consult with the foreign

government both prior to initiating and during the course of the

investigation. Finally, the commenter proposed that the definition

contain a requirement that all government-to-government exchanges (oral

and written) be placed on the record of the proceeding.

We do not believe that a regulation is required to define

``consultations.'' We agree that, in accordance with Article 13 of the

SCM Agreement, the purpose of consultations is to clarify the

allegations presented in a petition and arrive at a mutually agreed

solution. Section 351.202(h)(2)(i)(2) of Antidumping Duties;

Countervailing Duties; Final rule, 62 FR 27295, 27384 (May 19, 1997)

clearly states that the Department will invite the government of any

exporting country named in a CVD petition to hold consultations with

respect to the petition. Further, consistent with Article 13.2 of the

SCM Agreement, the Department affords foreign governments reasonable

opportunities to consult throughout the period of investigation. In

regard to communications, it is the Department's longstanding practice

that all ex parte communications with Department decisionmakers be

placed on the record of a proceeding through memoranda to the file.

Section 351.501

Section 351.501 restates very generally the subject matter of

subpart E. To be more specific, the arrangement of subpart E is as

follows. After dealing with the specificity of domestic subsidies in

Sec. 351.502 and the concept of ``benefit'' in Sec. 351.503,

Secs. 351.504 through 351.513 deal with the identification and

measurement of various general types of subsidy practices. Sections

351.514 through 351.520 focus on export subsidies, incorporating the

appropriate standards from the Illustrative List of Export Subsidies

contained in Annex I of the SCM Agreement. Sections 351.521 through

351.523 deal with import substitution subsidies (currently designated

as ``Reserved''), green light and green box subsidies, and upstream

subsidies, respectively. Section 351.524 addresses the allocation of

benefits to a particular time period. Section 351.525 sets forth rules

regarding the calculation of an ad valorem subsidy rate and the

attribution of a subsidy to the appropriate sales value of a product.

Finally, Secs. 351.526 and 351.527 contain rules regarding program-wide

changes and transnational subsidies, respectively. The section

numbering in these Final Regulations reflects minor changes from the

1997 Proposed Regulations. As discussed below, we have decided to

codify a final rule on the concept of ``benefit.'' This rule is now

Sec. 351.503. We have also moved the rules regarding the allocation of

benefits, which were included in the section on grants in the 1997

Proposed Regulations to a separate section, Sec. 351.524. Finally, we

have moved Sec. 351.520 of the 1997 Proposed Regulations to

Sec. 351.514(b) because general export promotion activities are more

appropriately addressed as an exception to export subsidies.

The last sentence of Sec. 351.501 acknowledges that subpart E does

not address every possible type of subsidy practice. However, the same

sentence provides that in dealing with alleged subsidies that are not

expressly covered by these regulations, the Secretary will be guided by

the underlying principles of the Act and subpart E.

In this regard, the Act and the SCM Agreement serve to eliminate

much of the confusion and controversy surrounding the necessary

elements of a countervailable subsidy. First, under section 771(5)(B)

of the Act and Article 1.1(a)(1) and (2) of the SCM Agreement, there

must be a financial contribution that a government provides either

directly or indirectly, or an income or price support in the sense of

Article XVI of the General Agreement on Tariffs and Trade 1994 (``GATT

1994''). Although the precise parameters will have to be determined on

a case-by-case basis, this element provides a framework for analysis

that previously was not directly addressed.

Second, under section 771(5)(B) of the Act and Article 1.1(b) of

the SCM Agreement, the financial contribution (or income or price

support) must confer a benefit. Section 351.503 sets out the principles

we will generally follow in determining whether a benefit has been

conferred.

Finally, under section 771(5)(A) of the Act and Article 1.2 of the

SCM Agreement, a subsidy must be specific in order to be

countervailable. The ``specificity test'' is addressed in Sec. 351.502,

but we note here that by clarifying the purpose of the specificity test

and the manner in which it is to be applied, the URAA, the SAA and the

SCM Agreement should serve to reduce the controversies and volume of

litigation concerning this issue.

In the preamble to our 1997 Proposed Regulations we discussed our

decision not to include two topics in our proposed changes to subpart

E: Indirect subsidies (with the exception of upstream subsidies) and

privatization. The numerous comments regarding our decision not to

promulgate regulations on these two topics are addressed below.

Indirect Subsidies

In our 1997 Proposed Regulations, we discussed only briefly the

topic of indirect subsidies. We received several comments on this

issue. Comments concerning the adoption of a definition of the phrase

``entrusts or directs'' have been addressed previously (see

Sec. 351.102). The remaining comments relating to indirect subsidies

are addressed here.

One commenter asked the Department to codify a rule stating that

indirect subsidies are countervailable. In this commenter's view, this

would eliminate any uncertainty that could become the cause of

litigation. Another commenter requested that the Department include a

[[Page 65351]]

broad definition of indirect subsidies in our regulations.

We have not adopted either suggestion. We believe that section

771(5)(B)(iii) of the Act clearly states that subsidies provided by

governments through private parties are covered by the CVD law.

Additionally, section 771(5)(C) of the Act states that the

determination of whether a subsidy exists shall be made ``without

regard to whether the subsidy is provided directly or indirectly * *

*'' (emphasis added). Therefore, no regulation is needed on this point.

Regarding the second comment, as discussed previously, the phrase

``entrusts or directs'' as used in section 771(5)(B)(iii) of the Act

could encompass a broad range of meanings. As such, we do not believe

it is appropriate to develop a precise definition of the phrase for

purposes of these regulations.

One commenter singled out subsidies involving the provision of

goods and services for less than adequate remuneration and asked the

Department to confirm that indirect subsidies can be conferred through

the provision of goods or services by private parties. This same

commenter also asked the Department to state in the preamble to the

Final Regulations that the new statute will not alter the Department's

practice of finding export restraints to be countervailable. Other

commenters objected to this position. They argued that: (1) The

practices constituting financial contributions under the Act are

payments of cash or cash equivalents, while government regulatory

measures do not entail any financial contribution; (2) export

restraints do not direct private parties to make any type of payment;

they simply limit the parties' ability to export; (3) regulatory

measures that distort trade are separately covered by other World Trade

Organization (``WTO'') Agreements (e.g., GATT 1994 Articles I-V, VII-

IX, Agreement on Sanitary and Phytosanitary Measures, Agreement on

Technical Barriers to Trade, and Agreement on Trade-Related Investment

Measures); and (4) expanding the definition of subsidy to include

regulatory measures would extend that term to absurd dimensions far

beyond the limited scope intended by the SCM Agreement and the Act.

These same commenters urged the Department to issue a regulation which

clarifies what they see as a conflict between the clear language in the

statute (regulatory measures are not financial contributions within the

meaning of the Act and, hence, cannot confer subsidies) and the

language in the SAA at 926 (suggesting that regulatory measures can be

countervailed as indirect subsidies).

Regarding the issue of whether indirect subsidies can arise through

the provision of goods and services, we believe this is clearly

answered by the Act. Section 771(5)(D)(iii) states that financial

contributions include the provision of goods or services. Hence, if a

private entity is entrusted or directed to provide a good or service to

producers of the merchandise under investigation, a financial

contribution exists. With regard to export restraints, while they may

be imposed to limit parties' ability to export, they can also, in

certain circumstances, lead those parties to provide the restrained

good to domestic purchasers for less than adequate remuneration. This

was recognized by the Department in Certain Softwood Lumber Products

from Canada, 57 FR 22570 (May 28, 1992) (``Lumber'') and Leather from

Argentina, 55 FR 40212 (October 2, 1990) (``Leather''). Further, as

indicated by the SAA (at 926), and as we confirm in these Final

Regulations, if the Department were to investigate situations and facts

similar to those examined in Lumber and Leather in the future, the new

statute would permit the Department to reach the same result.

We agree that regulatory measures that distort trade normally may

be subject to the provisions of other WTO Agreements. We do not

believe, however, that this negates our ability to address them through

the application of our CVD law when such measures meet the definition

of a countervailable subsidy. We disagree that countervailing such

measures goes beyond the ambit of the SCM Agreement and the Act. As

discussed above in response to an earlier comment, the SCM Agreement

clearly permits, and the Act clearly requires, that we countervail

subsidies provided through private parties. Also, Article VI of GATT

1994 continues to refer to subsidies provided ``directly or

indirectly'' by a government.

Change in Ownership

The SAA and the House and Senate Reports emphasize the importance

of considering the facts of individual cases to determine whether, and

to what extent, change-in-ownership transactions eliminate previously

conferred countervailable subsidies. In the 1997 Proposed Regulations,

we did not include a provision dealing with change in ownership.

Rather, we invited comment on a broad array of factors concerning this

topic and whether we should promulgate a final rule that integrates

some or all of the factors identified in the preamble.

The comments we received on this issue largely fell along two

lines. On the one hand, several commenters argued that the Department

should promulgate a regulation stating that change-in-ownership

transactions, even if conducted at arm's-length and at fair market

value, have no effect on non-recurring subsidies bestowed prior to the

sale of a firm, and that non-recurring subsidies, in most instances,

pass through in their entirety to the sold or privatized entity.

Conversely, other commenters contended that a change-in-ownership

regulation should establish a rebuttable presumption that, in general,

the sale or change in ownership of a firm at fair market value

eliminates the benefit conferred by prior non-recurring subsidies.

According to the first group of commenters, under section 771(5)(F)

of the Act, the change in ownership of a firm has no effect on the

Department's ability to countervail fully subsidies bestowed prior to

the change in ownership. In fact, in these commenters' view, Congress

expected the Department to continue countervailing prior subsidies,

unless something serves to eliminate those subsidies. The sale of a

firm at fair market value does not serve to eliminate prior subsidies;

thus, after such a sale, prior subsidies would continue to be

countervailed until fully amortized. The only instance where partial

repayment of prior subsidies can exist is where economic resources have

been returned to the government, i.e., where the investor has paid more

than fair market value for a productive unit. The Department should

specify this in its regulations.

These same commenters argued that recent court decisions support

the conclusion that subsidies continue to be countervailable after the

privatization of a firm at fair market value. See, e.g., Saarstahl AG

v. United States, 78 F.3d 1539 (Fed. Cir. 1996); British Steel plc v.

United States, 127 F.3d 1471 (Fed. Cir. 1997). In light of these

decisions, one commenter stated that it would be ironic for the

Department now to conclude under the URAA that subsidies are no longer

countervailable after the sale of a firm at fair market value. This

commenter also claimed that such a conclusion would result in anti-

subsidy practices weaker than those of the European Union (``EU''),

because EU Guidelines on State Aid recognize that the sale of a company

does not extinguish previously bestowed subsidies. Rather, according to

this commenter, the EU requires subsidy recipients to repay illegal

subsidies, including principal and interest, from the time the aid was

disbursed, without

[[Page 65352]]

regard to whether the recipient is later sold or

privatized.4

---------------------------------------------------------------------------

\ 4\ In support of this proposition, the commenter cites

Community Guidelines on State Aid for Rescuing and Restructuring

Firms in Difficulty, O.J. Eur. Comm. No. C283/2 at 283/4 (September

19, 1997) (``The assessment of rescue or restructuring aid is not

affected by changes in the ownership of the business aided. Thus, it

will not be possible to evade control by transferring the business

to another legal entity or owner.'')

---------------------------------------------------------------------------

These commenters opposed the Department's attempt to develop a

``flexible'' approach toward privatization. They expressed concern that

ascribing any significance to the broad array of factors listed in the

1997 Proposed Regulations may lead to all or some pre-privatization

subsidies being extinguished in a fair market privatization, which

would involve reevaluating the amount, and possibly the existence, of

prior subsidies based on post-bestowal events and conditions. This

would violate the statute's prohibition against considering the effects

of subsidies and the Department's practice of not examining subsequent

events to determine whether the subject merchandise continues to

benefit from subsidies. See section 771(5)(C) of the Act and GIA at

37261. For example, one commenter stated that taking account of current

market conditions, such as global overcapacity, in determining the

extent to which pre-privatization subsidies pass through, is tantamount

to considering effects. Similarly, another commenter rejected the

suggestion that subsidies that reduce excess capacity are not

countervailable because this too depends on an impermissible ``use''

analysis. Whatever the use of the subsidy, these commenters argued, the

benefit from the subsidy continues unabated after privatization.

Finally, this first group of commenters asserted that the

privatization or sale of a productive unit, even at fair market value,

does not result in any partial or full repayment of prior subsidies. To

conclude otherwise would conflict with Congress' mandate that the

Department's privatization methodology be ``consistent with the

principles of the countervailing duty statute.'' S. Rep. No. 103-412,

at 92 (1994). Those principles include prohibitions against (1)

focusing on subsequent events, (2) analyzing alleged effects of

subsidies, (3) granting offsets not included in the exclusive statutory

list, and (4) valuing subsidies based on the cost-to-government

standard. Some in this first group of commenters asserted that the

logical reading of Congress' instruction to evaluate change-in-

ownership transactions on a case-by-case basis is to determine whether

a privatization or sale involving a productive unit elicits some non-

commercial activity, i.e., whether under- or overpayment for the

productive unit has occurred. In the case of underpayment, the

Department should find that additional subsidies have been bestowed; in

the case of overpayment, the Department should find that certain prior

subsidies have been repaid.

In contrast to these arguments, the second group of commenters

asserted that the Department should issue regulations establishing a

rebuttable presumption that the arm's-length sale of a firm, including

a government-owned enterprise, at a price that reflects the current

market value of its assets, in most cases extinguishes any previously

received subsidies. This group argued that Congress' instruction to

examine change-in-ownership transactions on a case-by-case basis

indicates that the URAA contemplates extinguishment of prior subsidies,

at least in certain circumstances. In these commenters' view, the

arm's-length sale of a company at full market value is such a

circumstance, because the market price takes into account prior

subsidies, and the benefit is, therefore, eliminated. However, if the

price paid for the firm does not reflect full market value, the

question of a continuing benefit can reasonably be raised. According to

several of these commenters, any other approach would be

counterproductive, because it would discourage potential buyers from

bidding on subsidized government-owned enterprises about to be

privatized. One commenter further stressed that restructuring of, and

foreign investment in, countries such as those in Eastern Europe, may

be inhibited, which is a concern for U.S. investors and the United

States' wider economic and political interests.

One member of this group of commenters found support for the

proposition that an arm's-length sale at fair market value must

extinguish prior subsidies with the following statutory analysis. The

commenter claimed that the URAA requires the Department to determine

whether and to what extent government financial contributions confer a

benefit on the production or sale of the investigated merchandise in

each CVD proceeding. Such a determination is based on the nature of the

subsidy benefit, which is the artificially reduced cost of an input

used in the production of the merchandise. Thus, where the subsidy is

provided for a specific use, e.g., the acquisition of capital assets,

the continuing subsidy benefit is the reduced cost of that asset

allocated over the useful life of the asset. Where government financial

contributions are not tied to specific applications, as in the case of

an equity infusion, the Department should normally view the money

itself as the continuing subsidy benefit.

In light of this, the commenter contended that the Department's

privatization analysis must first examine what inputs were acquired by

the subsidy recipient at an artificially reduced cost. Then, the

Department must determine whether the cost for those inputs was

artificially reduced for the privatized company as well. According to

this commenter, where the privatization transaction occurs at arm's-

length and at fair market value, the privatized company would not

continue to benefit from the past subsidies. Similarly, where

government financial contributions are not tied to specific

applications, meaning that the money itself is the continuing subsidy

benefit, the Department's focus should be on the price and terms of the

privatization transaction. If the privatization of the company,

including all its physical and financial assets, was at fair market

value, the Department would not find any benefit to have passed

through, because the privatized company would not be operating with any

capital for which it paid less than market value. According to this

commenter, if the privatization of a firm were at full market value,

the new owners of the company have paid for all of the inputs at market

value. Therefore, the privatized firm no longer operates with inputs

acquired at a cost that is less than what would have been paid without

a government financial contribution.

This commenter stressed that there are several possible exceptions

to this rule. For example, where an asset would not have been created

or acquired absent the government financial contribution, and where the

creation or acquisition of the asset was not economically viable, the

Department may conclude that the very existence of the asset is the

continuing benefit and not the reduced costs of the asset. In such an

instance, the benefit could be deemed to continue, even after a full

market privatization. However, this commenter asserted that this would

represent an exception to the general rule.

This commenter rejected the argument that this analysis is

tantamount to an ``effects'' test. If a subsequent event does in fact

eliminate subsidization, limited Departmental resources should not

prevent examination of that event. The commenter stated that, in the

case of

[[Page 65353]]

subsidies not tied to any particular use, the only event that the

Department would need to consider is one which would eliminate the

artificially reduced cost of the company's inputs as a whole. The sale

of an entire company for market value is such an event, in the

commenter's view. Where a subsidy is tied to a particular use, the only

event that the Department would need to consider is one that would

affect or eliminate the benefit arising from that specific use.

Moreover, according to the commenter, in numerous contexts the

Department traces the use of a subsidy. These include instances where

subsidies are provided for certain uses that may be greenlighted or

that may benefit a company over time, i.e., non-recurring subsidies.

Most commenters also found fault with the Department's existing

repayment or reallocation methodology, under which pre-sale subsidies

are partially repaid to the seller as part of the purchase price.

Several commenters argued that the repayment/reallocation methodology

should be abandoned, because it is not defensible, economically or

legally. According to these commenters, the repayment/reallocation

methodology violates the offset provision of the statute (section

771(6) of the Act), because this provision does not include repayment

or reallocation of subsidies in the context of a privatization at fair

market value. Moreover, a fair-market-value privatization does not

offset the distortion caused by government subsidies, a fact recognized

by EU law, according to which subsidy repayment can occur only if the

illegal aid is returned.5 According to these commenters, the

repayment/reallocation methodology is also inconsistent with the

Department's and the Court's ``conceptual model of subsidies,'' which

presumes that subsidies distort market processes and result in a

misallocation of resources (citing Carbon Steel Wire Rod from Poland,

49 FR 19374, 19375 (May 7, 1984), and Georgetown Steel Corp. v. United

States, 801 F.2d 1308, 1315-16 (Fed. Cir. 1986) (``Georgetown Steel'').

Under this model, repayment or reallocation can only occur if an

equivalent ``distortion'' takes place, that is, a return of the

illegally provided resources from the subsidized entity. This does not

occur, the commenters emphasized, in a fair-market privatization.

Further, the repayment/reallocation methodology is inconsistent with

the benefit-to-recipient standard because it is based on the assumption

that the government was paid more money upon privatization than it

would have received absent the subsidy, a fact that is only relevant

under a cost-to-government standard. These commenters stated that while

the cost of the subsidy to the government may be diminished in a fair-

market privatization, the value of the subsidy to the recipient is

unchanged. According to these commenters, by finding that repayment/

reallocation occurs in a fair-market-value transaction, the Department

is encouraging subsidization. This violates the basic purpose of the

CVD law, which is intended to deter subsidization. These commenters

also argued that the Court of International Trade's (``CIT'') decision

in British Steel plc vs. United States, 879 F. Supp. 1254, 1277 (CIT

1995), aff'd in part and rev'd in part, 127 F.3d 1471 (Fed. Cir. 1997),

casts doubt on the permissibility of finding repayment in the context

of a privatization at fair market value. One commenter also argued that

the repayment/reallocation methodology is inconsistent with the URAA

and the SAA's instruction to examine carefully the facts of each case

in determining the effects of privatization on prior subsidies, because

it is an automatic rule that always assumes a portion of the purchase

price represents repayment or reallocation of prior subsidies.

---------------------------------------------------------------------------

\5\ Citing Commission notice pursuant to Article 93(2) of the EC

Treaty to other Member States and interested parties concerning aid

which Germany has granted to Fritz Egger Spanplattenindustrie GmbH &

Co. KG at Brilon, O.J. Eur. Comm. No. C369/6, 369/8-369/9 (1994),

and Agreement Respecting Normal Competitive Conditions in the

Commercial Shipbuilding and Repair Industry, opened for signature

December 21, 1994, art. 8, para. 5.

---------------------------------------------------------------------------

Another commenter asserted that the repayment/reallocation

methodology does not capture the full extent of the benefit bestowed

upon a company because it does not capture the benefit from the

government's assumption of risk. According to this commenter, to

encourage investment in risky industry sectors, governments can assume

some of the risk, for example by providing start-up capital. If the

government privatizes the company, the trade-distorting effect of the

government action continues, and the production of the company

continues to enjoy the benefit of the government subsidy. This

commenter argued that if the Department maintains the repayment/

reallocation methodology, it should also consider whether the industry

could attract private capital at the time the subsidies were provided.

Where an industry could not attract private capital, the Department

should find that all subsidies passed through after privatization.

Alternatively, if the Department finds that privatization can

extinguish or repay a subsidy, this should only be permitted when the

price paid for the privatized company is equal to the net worth of the

firm without the subsidy, plus the residual value of the subsidy. For

example, a firm receives a $1 million countervailable subsidy, which

the Department allocates over 10 years. In year two, the residual value

of the subsidy (for countervailing duty purposes) is $900,000. In that

year, the firm is privatized and its pre-subsidy assets are valued at

$18 million. If the firm is sold for $18.9 million, the subsidy would

be repaid. If it is sold for $18 million, the subsidy would pass

through in its entirety. According to this commenter, this approach

recognizes that the buyer of a firm is paying for the assets as well as

the residual value of the subsidy, while the current repayment/

reallocation approach fails to do this.

Another modification suggested by some commenters to the repayment/

reallocation methodology is to alter the calculation of ``gamma,''

which measures the proportion of the purchase price that the Department

considers to be repaid to the government in a privatization

transaction, or reallocated to the previous owner in a private-to-

private sale. This commenter stated that the gamma ratio should be

calculated using the total remaining value of the subsidies at the time

of the privatization to the company's total net worth in the same year,

rather than using the average of the historical values of the subsidies

to the firm's net worth starting in the years the subsidies were

received. This approach would give more weight to subsidies received

immediately preceding privatization.

Finally, several commenters addressed the issue of whether

subsidies provided in anticipation, or in the process, of privatization

should be given special consideration. On the one hand, one commenter

argued that subsidies provided shortly before, and in preparation for,

the sale, such as debt forgiveness, asset revaluations, tax breaks, and

other measures to ``clean up'' balance sheets, should be considered new

subsidies and not ``pre-privatization'' subsidies. According to this

commenter, under no circumstance should these subsidies be eliminated

as part of the privatization transaction. On the other hand, another

commenter suggested that steps taken by a government just prior to

privatization to make a company more ``saleable,'' such as closing

inefficient operations, should not by themselves be considered

[[Page 65354]]

subsidies that pass through to the privatized company.

Except for the comments on our current repayment/reallocation

methodology and the comments on subsidies given in the process of

privatization, which we address below, the commenters have presented

two general positions with respect to the impact of changes in

ownership on subsidies bestowed prior to the sale: (1) That the arm's-

length sale of a company at fair market value has no effect on the

countervailability of prior subsidies; and (2) that the fair-market

sale of a firm, in general, excuses the purchaser from any CVD

liability for prior subsidies. While the commenters suggest possible

exceptions to these general positions that theoretically would give

effect to the statutory direction to consider the facts of each case,

the exceptions are narrowly defined to fit improbable circumstances. In

most cases, the proposals, with their narrowly defined exceptions,

would lead to either total pass-through or total extinguishment of pre-

sale subsidies.

Although we see merit in some of the arguments presented, we

believe that adopting either of these extreme positions would require a

strained interpretation of the statute. The statute, SAA, and

legislative history plainly state that the arm's-length sale of a firm

does not by itself require a determination that prior subsidies have

been extinguished. See section 771(5)(F), SAA at 928, and S. Rep. No.

103-412, at 92 (1994); see also the discussion in the 1997 Proposed

Regulations at 8821. Moreover, we continue to disagree with the claim

that in order to impose countervailing duties on a privatized or post-

sale firm, the Department must affirmatively demonstrate how subsidies

continue to benefit the subject merchandise after the fair-market sale

of a company. See GIA at 37263. Our refusal to read a continuing

competitive benefit test (sometimes called an ``effects test'') into

the CVD law was upheld by the Federal Circuit in Saarstahl v. United

States, 78 F.3d 1539 (Fed. Cir. 1996) (``Saarstahl'') and British Steel

plc v. United States, 879 F. Supp. 1254 (CIT 1995), aff'd in part and

rev'd in part 127 F.3d 1471 (Fed. Cir. 1997) (``British Steel''). As

the CIT explained in British Steel plc v. United States, ``Commerce has

consistently maintained that it does not measure the effects of

subsidies once they have been determined by Commerce. In other words,

whether subsequent events mitigate these effects is irrelevant. This

Court, for the purposes of this proceeding, has no quarrel with that

practice.'' 879 F. Supp. at 1273. Further, section 771(5)(C) of the Act

specifically states that the Department ``* * * is not required to

consider the effect of the subsidy in determining whether a subsidy

exists * * *'' See also Certain Hot-Rolled Lead and Bismuth Carbon

Steel Products from the United Kingdom, 61 FR 58377, 58379 (November

14, 1996) (1994 Administrative Review UK Lead Bar).

In this regard, it is useful to clarify what we mean in saying that

we would not attempt to determine whether a subsidy had any ``effect''

on the recipient, or whether ``subsequent events'' might have mitigated

or eliminated any potential effects from the subsidy. The term

``effect,'' as used in the statute and SAA, and the term ``subsequent

events,'' as used by the Courts, refer to the question of whether a

subsidy confers a competitive benefit upon the subsidy recipient or its

successor. There is no requirement that the Department determine

whether there is a competitive benefit, as is made clear in the SAA (at

926):

* * * the new definition of subsidy does not require that Commerce

consider or analyze the effect (including whether there is any

effect at all) of a government action on the price or output of the

class or kind of merchandise under investigation or review.

In the course of the 1993 steel investigations, certain respondents

argued that: (1) A subsidy cannot be countervailed unless it bestows a

``competitive benefit'' on merchandise exported to the United States;

(2) the arm's-length sale of a subsidized company eliminates any

competitive benefit from prior subsidies (because the price paid for

the company includes payment for any continuing value the subsidies

might have); and (3) therefore, the arm's-length sale of a subsidized

company frees the new owner from any countervailing duty liability for

prior subsidies to that company. We rejected this argument (see GIA at

37260-61), explaining that the statute did not require that a subsidy

bestow a competitive benefit on imports to the United States as a

condition of liability for countervailing duties. Just as we would not

attempt to determine whether a subsidy conferred a competitive benefit

on the original recipient in the first place (that is, whether the

subsidy had any effect on the original recipient's subsequent

performance (usually an effect upon its output or prices)), we would

not attempt to determine whether any potential competitive benefit

continued with respect to the new owner in light of a subsequent event

such as a change in ownership. The Federal Circuit upheld this position

in Saarstahl and British Steel. As one commenter noted, the law is

concerned with the benefit originally received, not with what the

recipient does with it.

When we say we do not consider ``subsequent events'' in the

calculation of a subsidy, we generally are referring to events that

arguably affect the subsequent performance (normally in terms of output

or prices) of the subsidy recipient or its successor. We have never

implied, however, that no subsequent event could ever affect the

allocation of a subsidy. The Department may consider whether government

or private actions occurring after the receipt of a subsidy should

result in the reallocation of a subsidy as long as there is no tracing

of the uses of the subsidy or the effect of the subsidy on the output

or price of subject merchandise. Clearly, a post-subsidy change in

ownership is an event that occurs subsequent to the receipt of the

subsidy, and we have reallocated subsidies based on changes in

ownership. It is entirely appropriate and consistent with the statute

to consider whether a change in ownership is an appropriate occasion to

reallocate countervailing duty liability for prior subsidies to the

company that is sold. Section 771(5)(F) of the Act implies that such an

exercise is warranted and, as explained above, a post-subsidy change in

ownership is not the type of subsequent event or effect that is

envisioned in section 771(5)(C).

The language of section 771(5)(F) of the Act purposely leaves much

discretion to the Department with regard to the impact of a change in

ownership on the countervailability of past subsidies. Specifically, a

change in ownership neither requires nor prohibits a determination that

prior subsidies are no longer countervailable. Rather, the Department

is left with the discretion to determine, on a case-by-case basis, the

impact of a change in ownership on the countervailability of past

subsidies. The SAA at 928 specifically states that ``Commerce retain[s]

the discretion to determine whether, and to what extent, the

privatization of a government-owned firm eliminates any previously

conferred countervailable

subsidies. . . .''

The repayment/reallocation methodology that we currently use

achieves this objective. See 1994 Administrative Review UK Lead Bar at

58379-80. Depending on the amount of prior subsidies in relation to the

company's net worth and the amount paid for the company, we might find

that a considerable amount of prior subsidies passes through or that a

[[Page 65355]]

significant amount of subsidies has been repaid to the government or

reallocated to the previous owner. Nonetheless, we are not codifying

the current repayment/reallocation methodology. This methodology has

been heavily criticized by various parties, and we recognize that it

may not provide sufficient flexibility to deal with the ``extremely

complex and multifaceted'' nature of changes in ownership. See SAA at

928. We will address comments related to the calculation of gamma in

the context of specific cases.

While we have developed some expertise on the issue of changes in

ownership over the past five years, and the comments submitted in

response to the 1997 Proposed Regulations have provided us with

additional ideas to consider, we do not think it is appropriate to

promulgate a regulation on this issue at this time. As noted above,

many of the ideas presented by the commenters would move us in the

direction of adopting extreme positions. Another factor weighing

against codification of any privatization methodology at this time is

that the Courts may, in the course of their review of the current

methodology, adopt an interpretation of the law that would either

validate or overturn some of the options that we have considered,

including those proposed by the commenters. Finally, given the rapidly

changing economic conditions around the world, particularly with

respect to the issue of state ownership, we believe we should continue

to develop our policy in this area through the resolution of individual

cases. These changing economic conditions pose additional challenges in

developing a unified framework in which to analyze change-in-ownership

transactions. In the 1997 Proposed Regulations, we identified many of

these additional issues and new challenges that may warrant

consideration in this context and raised questions about them. However,

it is our view that the comments we received did not sufficiently

address many of these concerns.

An additional issue that merits further discussion concerns

subsidies received just prior to, or in conjunction with, the

privatization of a firm. While we have not developed guidelines on how

to treat this category of subsidies, we note a special concern because

this class of subsidies can, in our experience, be considerable and can

have a significant influence on the transaction value, particularly

when a significant amount of debt is forgiven in order to make the

company attractive to prospective buyers. As our thinking on changes in

ownership continues to evolve, we will give careful consideration to

the issue of whether subsidies granted in conjunction with planned

changes in ownership should be given special treatment.

Our decision not to include a provision on changes in ownership in

these Final Regulations does not preclude us from issuing such a

regulation at a later date. We will continue to examine this issue and

consider whether an alternative analytical framework can be developed

that addresses the variety of change-in-ownership scenarios we have

encountered and that, like the present methodology, satisfies

Congressional intent that we examine changes in ownership on a case-by-

case basis. In the interim, we will continue to apply our current

methodology for ongoing CVD cases and carefully examine the facts of

each case. However, we will consider whether modifications to the

methodology may be appropriate.

Section 351.502

Section 351.502 deals with the ``specificity'' of domestic

subsidies. Unlike its predecessor, Sec. 355.43 of the 1989 Proposed

Regulations, Sec. 351.502 does not contain a ``general'' specificity

test. As we noted in the preamble to the 1997 Proposed Regulations,

section 771(5A) of the Act and the SAA provide much more detail and

clarity regarding the application of the ``specificity test'' than did

the prior statute and its legislative history. Thus, on the subject of

specificity, there are far fewer interpretative gaps for the Department

to fill than there were in 1989 and, thus, less need for regulations.

We received numerous comments arguing that we should codify the

policies articulated in the preamble to the 1997 Proposed Regulations,

especially those dealing with sequential analysis, purposeful

government action, characteristics of a ``group,'' and integral

linkage. These commenters claimed that even where the SAA is clear on a

particular point, it is unclear how the Courts will view the SAA. In

their opinion, detailed specificity regulations would prevent costly

litigation of these issues.

We have continued to limit Sec. 351.502 to those aspects of the

specificity test that are not addressed explicitly in the statute or

the SAA. Section 102(d) of the URAA provides that the SAA ``shall be

regarded as an authoritative expression by the United States concerning

the interpretation and application of (the Agreements and the URAA) in

any judicial proceeding in which a question arises concerning such

interpretation or application.'' 19 U.S.C. Sec. 3512(d). Therefore, we

see no need to repeat this principle. However, in reviewing the

comments and the relevant provisions of the statute and the SAA, we

have identified particular issues on which the SAA may usefully be

clarified. In particular, we found that the statute and the SAA do not

fully address sequential analysis and the characteristics of a group.

Accordingly, we have included final regulations on these topics.

Sequential analysis: Paragraph (a) is a new paragraph which

addresses the ``sequential approach'' to specificity. We received

several requests that we codify the sequential approach. Under this

approach, if a subsidy is de jure specific or meets any one of the

enumerated de facto specificity factors, in order of their appearance

in section 771(5A)(D)(iii) of the Act, further analysis is unnecessary

and is not undertaken. In support of their position, these commenters

emphasized the language contained both in section 771(5A)(D)(iii) of

the Act and the SAA that a subsidy will be considered specific ``if one

or more'' of the factors exists. See SAA at 931. Furthermore, these

commenters contended, the SAA and the legislative history of the URAA

make clear that the specificity test was intended to be generally

consistent with the Department's previous practice, a practice that

included this sequential approach. SAA at 929-31; S. Rep. No. 103-412,

at 93-94 (1994).

In opposition to this view, other commenters maintained that the

sequential approach contradicts the SAA, because the SAA states that

the Department will ``seek and consider information relevant'' to all

four of the de facto specificity factors. SAA at 931. Moreover, these

commenters maintained, the language in the SCM Agreement requires that

all of the de facto specificity factors be considered and that any

specificity determination ``shall be clearly substantiated on the basis

of positive evidence.'' Articles 2.1(c) and 2.4 of the SCM Agreement.

The apparent disagreement over the interpretation of the SAA

regarding the use of a sequential approach indicates that it is

necessary to clarify our position in a regulation. Therefore,

Sec. 351.502(a) provides that the de facto specificity factors will be

examined in sequence, in order of their appearance in section

771(5A)(D)(iii) of the Act, and that the Department may find a domestic

subsidy to be specific based on the presence of a single de facto

specificity factor. For example, the Department will first look to see

if there is a limited number of users. If the number of users is

limited, we will look

[[Page 65356]]

no further. In accordance with the SAA, the Department will continue

its practice of collecting information regarding each of the four de

facto specificity factors; however, our analysis of the issue will stop

if we determine that a single factor justifies a finding of

specificity. As for the SCM Agreement, none of the provisions cited

precludes a finding of specificity based on the presence of a single

factor. Moreover, a finding that a certain industry receives

disproportionate amounts under a particular government program, for

example, constitutes positive evidence of specificity even if there are

numerous users of the program and there is little discretion in

awarding benefits.

Discretion: In endorsing the use of a sequential approach in the

preamble to the 1997 Proposed Regulations, we stated, ``with the

exception of the government discretion factor, the Department may find

a domestic subsidy to be specific based on the presence of a single de

facto specificity factor.'' (1997 Proposed Regulations at 8824.)

Certain commenters objected to the exception of the discretion factor,

arguing that the statute accords the exercise of government discretion

equal status with the other de facto specificity factors. They asked

the Department to clarify that the Department may find a subsidy to be

specific solely based on the degree of discretion exercised in the

administration of a subsidy program.

There appears to be a great deal of confusion and controversy over

the role of the fourth factor, discretion, in the finding of de facto

specificity. Based on the comments received and a review of the statute

and SAA, we are elaborating on the statements we made in the preamble

to the 1997 Proposed Regulations. As stated in the 1997 Proposed

Regulations, we do not believe that a finding of specificity may be

based solely on the fact that some measure of discretion may have been

exercised in the administration of a subsidy program. This position is

consistent with the SAA, which states that if a subsidy program is

broadly available and widely used and there is no evidence of dominant

or disproportionate use, the mere fact that government officials may

have exercised discretion in administering the program is insufficient

to justify a finding of specificity. SAA at 931.

Based on our experience in administering the CVD law, some measure

of administrative discretion exists in the operation of almost every

alleged subsidy program. At the most basic level, an administrator of a

program typically must exercise judgment or discretion in evaluating

the facts and merits of an application for a subsidy to determine

whether the applicant qualifies for the subsidy. If we were to find

specificity based simply on the exercise of this type of discretion,

the other de facto factors would be rendered meaningless, because

virtually every subsidy program in the world could be declared specific

on the basis of the discretion factor alone. This is clearly an absurd

result and could not have been the intent of Congress.

Instead, section 771(5A)(D)(iii)(IV) of the Act provides that a

subsidy is specific if:

The manner in which the authority providing the subsidy has

exercised discretion in the decision to grant the subsidy indicates

that an enterprise or industry is favored over others. (Emphasis

added.)

This language does not focus on discretion alone. Rather, it states

that discretion is relevant only to the extent that it is exercised in

a manner that favors one enterprise or industry over others. This

distinction is important because it supports the statements made in the

SAA and the position we are taking in these regulations. Haphazard,

random, or purposeless discretion cannot by itself indicate

specificity. Only discretion that shows favoritism toward some

enterprises or industries over others can inform the question of

specificity. In the Department's experience, favoritism generally will

manifest itself as one of the first three de facto factors: A limited

number of users, dominant users, or one or a few users receiving a

disproportionate amount of the subsidy. For example, administrators of

a program could exercise discretion in selecting some industries

instead of others as beneficiaries. If the selected industries

constituted a limited number of industries, there would be specificity.

Similarly, if benefits were distributed such that there was a

predominant user or such that certain users received disproportionate

benefits, there would be specificity. However, if the selected

industries constituted more than a limited number of industries, if

there were no dominant users or disproportionate benefits to certain

users, or if there were no other indication that one or a group of

enterprises or industries was favored over others, the program would

not be specific.

As indicated in the SAA at 931, the discretion factor is generally

more valuable as an analytical tool that enhances the analysis of the

other de facto specificity factors and criteria. The example given in

the SAA is the case of a new subsidy program for which there have been

few applicants and few recipients. In accordance with section

771(5A)(D)(iii) of the Act, in evaluating the four de facto factors,

the Department must take into account ``* * * the length of time during

which the subsidy program has been in operation.'' In the case of a new

program, the first three factors--limited number of users, dominant

user, or disproportionately large user--may provide little or

misleading indication regarding whether the program is de facto

specific. Therefore, the manner in which authorities have exercised

their discretion in the early days of a new program (e.g., by excluding

certain applicants and limiting the benefit to a particular industry)

might be more useful for the Department in making a specificity

determination. See SAA at 931.

Discretion can also come into play where evidence relating to the

first three factors is inconclusive. As an example, where the number of

users is borderline, discretion may help to inform whether there is

specificity. In this situation, the factors we might consider in

analyzing the relevance of discretion include the number of applicants

that are turned down, the reasons they are turned down, and the reasons

successful applicants are chosen.

Characteristics of a ``group'': New paragraph (b) clarifies the

Department's position regarding whether the Department must examine the

``actual make-up'' of a group of beneficiaries when performing a

specificity analysis. Citing PPG Industries, Inc. v. United States, 978

F.2d 1232, 1240-41 (Fed. Cir. 1992) (``PPG II''), one group of

commenters argued that, to be consistent with judicial precedent, the

Department must undertake such an analysis. According to these

commenters, if a group of recipients does not share similar

characteristics but, instead, consists of companies in a variety of

industries, the Department cannot conclude that the subsidy in question

is limited to a ``group of industries.'' Moreover, they argued, nothing

in the Act or the SAA requires the Department to ignore the

characteristics of the group receiving the benefits from an alleged

subsidy program.

Other commenters argued that the Department can identify a

``group'' of subsidy recipients without regard to any shared

characteristics of the individual group members. According to these

commenters, a proper understanding of what may constitute a specific

``group of industries'' flows directly from the

[[Page 65357]]

purpose of the specificity test as articulated in Carlisle Tire &

Rubber Co. v. United States, 564 F. Supp. 834 (CIT 1983)

(``Carlisle''); namely, that subsidy recipients should be considered a

specific group unless the recipient industries are numerous and

distributed very broadly throughout the economy. Moreover, these

commenters maintained that the Department has on several occasions

found subsidy programs specific even when the ``group'' of recipients

has not shared common characteristics. See, e.g., Steel Wheels from

Brazil, 54 FR 15523, 15526 (April 18, 1989) and Cold-Rolled Carbon

Steel Flat-Rolled Products from Korea, 49 FR 47284, 47287 (December 3,

1984).

As noted in the preamble to the 1997 Proposed Regulations, we

disagree with the first set of comments. Section 771(5A)(D) of the Act

provides that a subsidy may be found to be specific if it is limited to

a ``group'' of enterprises or industries. There is no requirement that

the members of a group share similar characteristics. The purpose of

the specificity test is simply to ensure that subsidies that are

distributed very widely throughout an economy are not countervailed.

There is no basis for adding the further requirement that subsidies

that are not widely distributed are also confined to a group of

enterprises or industries that share similar characteristics. See,

e.g., Certain Refrigeration Compressors from the Republic of Singapore,

61 FR 10315 (March 13, 1996).

Assuming, arguendo, that PPG II is relevant under the new law, this

decision upheld the Department's determination that the program in

question was not specific. To put PPG II in its proper context, it is

necessary to understand the facts presented in the underlying CVD case.

In that case, there were numerous enterprises that used the program

under investigation. Therefore, when looked at in terms of the number

of enterprises, the actual recipient enterprises did not appear to be

limited. However, this conclusion says nothing about whether the number

of industries that received benefits under the program was limited. To

answer this question, the Department (and the Court) correctly focused

on the makeup of the users. If the numerous enterprises that received

benefits had comprised a limited number of industries, then the program

would have been specific. However, because the users represented

numerous and diverse industries, the program was found not to be

specific. There is no basis in PPG II or in the language of section

771(5A)(D) of the Act for concluding that there is a requirement that

the limited users also share similar characteristics. Moreover, such a

requirement would undermine the purpose of the specificity test as

articulated in the SAA.

Several commenters have urged the Department to codify our position

with respect to this issue. Because this issue is not addressed in the

statute or the SAA, we have adopted this suggestion. Accordingly,

Sec. 351.502(b) provides that the Secretary is not required to

determine whether there are shared characteristics among enterprises or

industries that are eligible for, or actually receive, a subsidy in

determining whether that subsidy is specific.

Integral linkage: Paragraph (c) is a new paragraph which sets out

our revised test for considering two or more subsidy programs to be

``integrally linked.'' Section 355.43(b)(6) of the 1989 Proposed

Regulations provided that, for purposes of applying the specificity

test, the Department would consider two or more subsidy programs as a

single program if the Secretary determined that the programs were

``integrally linked.'' Section 355.43(b)(6) also set forth factors to

be considered in making this determination.

In the 1997 Proposed Regulations, we opted not to incorporate

Sec. 355.43(b)(6) into these regulations. We noted that claims of

integral linkage were relatively rare, and that when they did arise, we

did not find the factors set forth in Sec. 355.43(b)(6) particularly

helpful. We did not, however, rule out the possibility of considering

two or more ostensibly separate subsidy programs as constituting a

single program for specificity purposes, and we outlined circumstances

that might lead us to do so.

We received a number of comments requesting that we promulgate a

regulation which allows for integral linkage. Two commenters argued

that, in addition to the factors discussed in the preamble, the

regulation should re-codify certain of the factors found in the 1989

Proposed Regulations. These commenters also suggested that programs

should not be considered to be integrally linked unless they were

linked ``at their inception.'' These commenters asked the Department to

clarify that it will view claims of integral linkage narrowly and that

respondents will be required to establish that the programs are linked

by clear and convincing evidence. Other commenters argued that the

factors enumerated in both the 1989 Proposed Regulations and in the

preamble to the 1997 Proposed Regulations are too restrictive and that

any integral linkage test should not be applied narrowly.

We have given further consideration to our earlier decision not to

codify an integral linkage test. In light of the interest in this

issue, and the fact that we have had experience with a regulation on

this topic, we have concluded that it would be beneficial to parties to

promulgate a rule describing when two or more separate programs may be

integrally linked and treated as one program for specificity purposes.

We have not codified the 1989 rule because, as we stated in the

preamble to our 1997 Proposed Regulations, we did not find the factors

enumerated in that provision to be particularly useful. Instead,

Sec. 351.502(c) provides that integral linkage is possible in

situations where the subsidy programs have the same purpose (e.g., to

promote technological innovation), bestow the same type of benefit

(e.g., long-term loans or tax credits), confer similar levels of

benefits on similarly situated firms, and were linked at their

inception.

We believe these factors are more useful for finding integral

linkage than those contained in the 1989 Proposed Regulations because

they require evidence of similarities in the purposes and

administration of the programs which are more than coincidental. For

example, where a government claims that a program is integrally linked

with another program, Sec. 351.502(c)(4), which calls for the programs

to be linked at inception, requires evidence that, in establishing the

most recent program, the government's clear and express purpose was to

complement the other program.

As stated in the preamble to the 1997 Proposed Regulations, when an

interested party believes that two or more programs should be

considered in combination for purposes of the Department's specificity

analysis, that party will have the burden of identifying the relevant

programs and supporting its contention that the programs are integrally

linked by providing information and documentation regarding the

purpose, type and levels of benefit associated with the programs.

Agricultural subsidies: Paragraph (d) is based on Sec. 355.43(b)(8)

of the 1989 Proposed Regulations and is the same as Sec. 351.502(a) of

the 1997 Proposed Regulations. It provides that the Secretary will not

consider a domestic subsidy to be specific solely because it is limited

to the agricultural sector. Instead, as under prior practice, the

Secretary will find an agricultural subsidy to be countervailable only

if it is specific within the agricultural sector, e.g., a subsidy is

limited to livestock, or

[[Page 65358]]

livestock receive disproportionately large amounts of the subsidy. See,

e.g., Lamb Meat from New Zealand, 50 FR 37708, 37711 (September 17,

1985).

One commenter suggested that the Department should abandon the

special specificity rule for agricultural subsidies, citing the fact

that under section 771(5B)(F) of the Act and Article 13(a) of the WTO

Agreement on Agriculture, so-called ``green box'' agricultural

subsidies are non-countervailable. With respect to this comment, we

note that the Department's application of the specificity test to

agricultural subsidies was upheld in Roses, Inc. v. United States, 774

F. Supp. 1376 (CIT 1991) (``Roses''). Given the absence of any

indication that Congress intended the ``green box'' rules to change the

Department's practice or to overturn Roses, we are retaining the

special specificity rule for agricultural subsidies.

Subsidies to small- and medium-sized businesses: Paragraph (e) is

based on Sec. 355.43(b)(7) of the 1989 Proposed Regulations, and

continues to provide that the Secretary will not consider a subsidy to

be specific merely because it is limited to small or small- and medium-

sized firms. Instead, as under prior practice, the Secretary will find

such a subsidy to be countervailable if, either on a de jure or a de

facto basis, the subsidy is limited to certain small or small- and

medium-sized firms. As in the case of the special specificity rule for

agricultural subsidies, there is no indication that Congress intended

to alter this aspect of the Department's specificity practice. We

received no comments regarding this rule.

Disaster relief: Paragraph (f) provides that the Secretary will not

regard disaster relief as a specific subsidy if the relief constitutes

general assistance available to anyone in the affected area. Although

paragraph (f) has no counterpart in the 1989 Proposed Regulations, the

rule contained in paragraph (f) has been part of the Department's

specificity practice since Certain Steel Products from Italy, 47 FR

39356, 39360 (September 7, 1982), in which the Department stated that

``[d]isaster relief is not selective in the same manner as other

regional programs since there is no predetermination of eligible areas

and no part of the country, and no industry, is excluded from

eligibility in principle.'' However, before declaring a subsidy to be

non-specific under paragraph (f), the Department would have to be

satisfied that the subsidy in question was, in fact, bona fide disaster

relief. See Certain Steel Products from Italy, 58 FR 37327, 37332 (July

9, 1993). We received no comments regarding this rule.

Purpose of the specificity test: Some commenters requested that the

Department restate in the regulations the policy rationale behind the

specificity test. According to these commenters, the underlying purpose

of the specificity test is to identify those domestic subsidies that

confer a competitive advantage and thereby distort international trade.

Other commenters pointed out that the new statute expressly states that

the Department is not required to examine the effects of a subsidy or

establish that the subsidy has any effect at all. These commenters,

citing the reference to the Carlisle decision in the SAA, maintain that

the sole purpose of the specificity test is to ``winnow out only those

foreign subsidies which truly are broadly available and widely used

throughout an economy.'' SAA at 929-30.

In our view, the language from the SAA cited above makes the

purpose of the specificity test abundantly clear. Given the clarity of

the SAA on this point, the authoritative nature of the SAA (see 19

U.S.C. 3512(d)), and our general reluctance to issue regulations that

merely repeat the statute or the SAA, we do not consider it appropriate

to issue a regulation that restates the purpose of the specificity

test.

Use of presumptions: Some commenters suggested that in applying the

specificity test, the Department should employ certain presumptions.

These commenters maintained that, when investigating a domestic subsidy

program (and when considering whether to initiate an investigation of

such a program), the Department should presume that the foreign

government in question exercises discretion in the administration of

the program, and that the program is specific. These commenters

maintained that, because information regarding applications and

approvals generally is not available to petitioners prior to the filing

of a petition, the burden should be on respondent interested parties to

provide such information and to rebut the presumption of specificity.

One commenter also suggested that the Final Regulations should state

that a previous finding that a subsidy was de facto non-specific should

have no relevance when the same subsidy program is alleged in a new

investigation involving different merchandise and different facts.

Other commenters argued that there is no legal basis for making

presumptions regarding specificity. With respect to de facto

specificity, the SAA states that the Department is obligated to ``seek

and consider'' information relevant to each of the four factors listed

in section 771(5A)(D)(iii) of the Act. SAA at 931. One of these

commenters also asserted that a petitioner alleging that a subsidy is

specific should be required to provide a reasonable amount of

information supporting the allegation.

As was true under the law prior to the URAA, we note that a

petition to initiate an investigation of alleged domestic subsidies

must provide reasonably available information supporting the allegation

that the subsidy is specific. See section 702(b) of the Act. On the

other hand, we recognize that because detailed information regarding

the distribution of program benefits usually either is not published or

is not widely available, information supporting specificity often is

not reasonably available to a petitioner at the time a petition is

filed. Therefore, in deciding whether to include alleged domestic

subsidies in our investigation, we carefully consider the information

the petitioner has put forward, the reasons that more information may

not be available, and any arguments the petitioner makes regarding the

specificity of the program. Because the types of allegations and

information available will vary from case to case, it is not possible

to state a general rule for accepting or rejecting specificity

allegations. However, we believe that the threshold we have used in the

past for including alleged subsidies in CVD investigations has been

sufficient to ensure that all potentially countervailable subsidies are

investigated. We intend to continue employing this initiation

threshold.

In this regard, we note that when a subsidy program has been

previously investigated and found to be non-specific, it would be a

waste of administrative resources to re-investigate that program

without a reasonable basis to believe that the facts supporting the

previous finding have changed. In situations where a previous finding

may be pertinent to one industry, e.g., that the paper clip industry

did not receive dominant or disproportionate benefits under a

particular program, petitioners seeking investigation of benefits under

that program to the staple industry should allege that the program has

changed or that the situation of the staple industry differs, and they

should support their allegation with reasonably available information.

Where domestic subsidy programs are included in an investigation,

we will not presume such programs are specific. Instead, we will seek

in our questionnaire all of the information

[[Page 65359]]

necessary to apply the specificity test according to section 771(5A)(D)

of the Act. Based on our analysis of the information provided in the

questionnaire responses, verification, and other information that may

be collected, we will make the necessary specificity determination. If

a respondent refuses to provide the information requested by the

Department to conduct its specificity analysis, we may draw adverse

inferences in the application of ``facts available.'' See section

776(b) of the Act. However, the use of an adverse inference in these

situations is not the same thing as relying on a rebuttable presumption

of specificity.

Purposeful government action: In our 1997 Proposed Regulations, we

noted that certain commenters, citing such cases as Saudi Iron and

Steel Co. (Hadeed) v. United States, 675 F. Supp. 1362, 1367 (CIT

1987), maintained that a finding of specificity does not require a

finding of targeting or some other sort of purposeful government action

that limits the number of subsidy program beneficiaries. They cited the

statute and its legislative history for the proposition that the

Department should deem irrelevant the fact that program usage may be

limited by the ``inherent characteristics'' of the thing being provided

by the government. SAA at 932; S. Rep. No. 103-412 at 94 (1994).

In the preamble to the 1997 Proposed Regulations, we agreed with

these commenters, stating:

[e]xcept in the special circumstances described in section 771(5A),

i.e., where respondents request the Department to take into account

the extent of economic diversification in the jurisdiction of the

granting authority or the length of time during which the program

has been in operation, the Department is not required to explain why

the users of a subsidy may be limited in number.

Several of the same commenters objected to this statement, arguing

that it could be misinterpreted to mean that evidence of purposeful

action is required in some instances. These commenters requested that

the Department clarify, in a regulation, that purposeful government

action is never required.

As we stated in the 1997 Proposed Regulations, the SAA and other

legislative history are clear on this point. The SAA clearly indicates

that the Department does not need to find ``targeting'' or ``purposeful

government action'' to conclude that a domestic subsidy is specific.

See SAA at 932 (``(E)vidence of government intent to target or

otherwise limit benefits would be irrelevant in de facto specificity

analysis''). Thus, for example, the fact that users may be limited due

to the inherent characteristics of what is being offered would not be a

basis for finding the subsidy non-specific. SAA at 932; S. Rep. No.

103-412 at 94 (1994). Regarding situations where the Department is

asked to consider the economic diversification in the jurisdiction or

the length of time during which the program has been in operation,

neither purposeful government action nor targeting is required to find

specificity. However, evidence indicating that the government has taken

or will take actions to limit benefits to certain industries would be

sufficient to find specificity.

Universe: One commenter argued that, in determining whether

subsidies are specific, the Department generally should focus on the

level of benefits provided to recipients, rather than the number of

recipients to whom subsidies are provided. This commenter also argued

that, in analyzing the level of benefits provided, the Department's

point of reference should be the economy as a whole, as it was for the

preferential loan programs used by the Korean steel industry in Certain

Steel Products from Korea, 58 FR 37338 (July 9, 1993) (``Korean

Steel''), rather than those enterprises or industries that were

eligible to receive the subsidy.

For the most part, we disagree. The starting point of the

Department's analysis of specificity will always be the number of

users. We normally will not analyze the level of benefits provided

(that is, whether the recipients were dominant or disproportionate

users of the program) unless the subsidy in question was provided to

numerous and diverse industries. Even in that situation, it may be

impracticable or impossible to determine the relative level of

benefits.

Once we have decided to analyze the level of benefits provided, our

point of reference normally will be the enterprises or industries that

received benefits under the program. In other words, we will attempt to

determine whether one or a limited number of the recipient enterprises

or industries were, in fact, dominant or disproportionate users. In

certain limited circumstances, however, it may be appropriate to

determine whether the benefits received by a particular enterprise or

industry or group thereof were disproportionate in relation to the

economy as a whole. The Department employed this approach in Korean

Steel, because the type of subsidy under investigation--governmental

use of the economy-wide banking system to direct credit to steel

producers--required the broader analysis. We consider the Korean

situation to be unusual compared with the majority of cases in which we

have analyzed specificity. In addition, we agree that the analysis of

whether an enterprise or industry or group thereof is a dominant user

of, or has received disproportionate benefits under, a subsidy program

should normally focus on the level of benefits provided rather than on

the number of subsidies given to different industries.

Section 351.503

Section 351.503 deals with the concept of benefit. Under section

771(5)(B) of the Act and Article 1.1(b) of the SCM Agreement, a

government action must confer a benefit in order to be considered a

countervailable subsidy. Hence, the notion of benefit is central to the

administration of the CVD law. In the preamble to the 1997 Proposed

Regulations, we included a lengthy discussion of this topic. We

described a benefit as being conferred when a firm pays less for an

input than it otherwise would pay or receives more revenue than it

otherwise would earn. Given the crucial role that benefit plays in our

analysis of whether a government action confers a countervailable

subsidy, we have decided to codify a final rule regarding benefit that

reflects the principles outlined in the 1997 Proposed Regulations.

Paragraph (a) states that, where a specific rule for the

measurement of a benefit is contained in these regulations, we will

determine the benefit as provided in that rule. Where a government

program is covered by a specific rule contained in these regulations,

such as a program providing grants, loans, equity, direct tax

exemptions, or worker-related subsidies, we will not seek to establish,

nor entertain arguments related to, whether or how that program

comports with the definition of benefit contained in this section.

Paragraph (b) outlines the principles we will follow when dealing

with alleged subsidies for which these regulations do not establish a

specific rule. In such instances, we will normally consider a benefit

to be conferred where a firm pays less for its inputs (e.g., money, a

good, or a service) than it otherwise would pay in the absence of the

government program, or receives more revenues than it otherwise would

earn.

We have adopted this definition because it captures an underlying

theme behind the definition of benefit contained in section 771(5)(E)

of the Act and, in our estimation, reflects the fundamental principles

that we have

[[Page 65360]]

articulated over the years with respect to programs and practices that

we have determined confer either direct or indirect countervailable

subsidies. One common element the four illustrative examples set forth

in the statute share is that, in the overwhelming majority of cases,

the recipient of a government financial contribution, income or price

support, or indirect subsidy, enjoys a reduction in input costs or

revenue enhancement that it would not otherwise have enjoyed absent the

government action. As explained below, we are using the terms ``input''

and ``cost'' broadly.

While we believe that this definition will provide useful guidance,

we recognize that there may be programs or practices not fitting the

input cost reduction or revenue enhancement definition in some economic

or accounting senses that may still give rise to a benefit in the sense

that the program or practice is similar to the illustrative examples

listed in section 771(5)(E) of the Act. For example, without attempting

to create a hypothetical program or practice not yet encountered in our

experience, we would argue that a program that is similar to a

countervailable equity infusion constitutes a reduction in a firm's

cost of capital, or that a program that is similar to a countervailable

provision of a freight forwarding service constitutes a reduction in a

firm's input costs. Since both practices constitute a reduction in the

cost of an input, there would be a benefit. We recognize that some

might take issue with whether equity or a freight forwarding service is

in fact an input into subject merchandise, or whether equity or freight

forwarding constitutes a cost of producing subject merchandise.

Nonetheless, in these and other instances in which a program or

practice contains elements similar to those in the illustrative

examples in the statute, a benefit would still exist. As explained

further below, when we talk about input costs in the context of the

definition of benefit, we are not referring to cost of production in a

strict accounting sense. Nor are we referring exclusively to inputs

into subject merchandise. Instead, we intend the term ``input'' to

extend broadly to any input into a firm that produces subject

merchandise.

When we talk about a firm paying less for its inputs than it

otherwise would pay (or receiving more revenues than it otherwise would

earn), we are referring to the lower price it pays to acquire the thing

provided by the government (e.g., money, a good, or a service), or the

increased revenue it receives as a result of a government action. We

believe that the definition of benefit outlined here is consistent with

the various standards (or ``benchmarks'') used to identify and measure

the benefit from different subsidy programs that are contained in

section 771(5)(E) of the Act and Article 14 of the SCM Agreement. For

example, when the amount that a firm pays on a government-provided loan

is less than what the firm ``would pay on a comparable commercial loan

that the (firm) could actually obtain on the market,'' the firm's cost

of borrowing money is reduced. See section 771(5)(E)(ii) of the Act.

Similarly, when a firm sells its goods to the government and ``such

goods are purchased for more than adequate remuneration,'' the firm's

revenues are increased beyond what it would otherwise earn. See section

771(5)(E)(iv) of the Act. In neither instance need the Department do

more than apply the test enumerated by the statute in order to find

that a benefit has been conferred.

Paragraph (b)(2) cautions that the definition of benefit as an

input cost reduction or revenue enhancement does not limit our ability

to impose countervailing duties when the facts of a particular case

indicate that a financial contribution has conferred a benefit, even if

that benefit does not take the form of a reduction in input costs or an

enhancement of revenues. We will examine the concept of benefit in this

broader sense by looking to see whether the alleged program or practice

contains elements similar to the examples listed in sections

771(5)(E)(i) through (iv) of the Act. We cannot possibly foresee all

the types of government actions we will encounter in administering the

CVD law and, hence, cannot write a definition of benefit that would be

sufficiently broad to capture all possible countervailable subsidies.

In this regard, it is important to note here our practice of not

applying the CVD law to non-market economies. The CAFC upheld this

practice in Georgetown Steel Corp. v. United States, 801 F.2d 1308

(Fed. Cir. 1986). See also GIA at 37261. We intend to continue to

follow this practice. Where the Department determines that a change in

status from non-market to market is warranted, subsidies bestowed by

that country after the change in status would become subject to the CVD

law.

We received several comments regarding the proposed definition of

benefit. Two commenters expressed the opinion that the definition is

too restrictive. These parties identified examples of benefits which

they believed would not be captured under the proposed definition. The

first example is where a domestic purchaser is the only customer for an

input provided by a government entity or where non-domestic purchasers

are not allowed to purchase an input. In these situations, the

commenter maintains that there could be a benefit even though the price

paid is not less than any other domestic price. The second example is

where a transaction is structured so that the firm pays market value

for the input but receives other perquisites, such as a higher-quality

input or additional services or goods as part of a package.

We disagree that our definition of a benefit is not comprehensive

enough to include these types of scenarios. The definition of a benefit

(in the absence of a specific rule for the measurement of the benefit)

does not call for comparisons only to other domestic prices. Rather, it

calls for a determination of whether the input costs were reduced

relative to what they would be in the absence of the financial

contribution. In the first example, a benefit exists to the extent that

the domestic purchaser would have paid more for the input absent the

government provision or absent the restrictions placed on foreign

purchasers. Likewise, in the second example, if the firm would have had

to pay more in order to receive the additional perquisites without the

government assistance, a benefit exists. Section 351.511, governing the

provision of goods and services, contains more detailed guidance on how

such subsidies would be valued.

Another commenter supported the proposed definition, but urged the

Department to leave itself enough flexibility so that we could find a

benefit when government action enables a firm to sell a product that

would not have been created but for the government assistance. For

example, if the government assists in the development of a new product,

this commenter asserted that the benefit is not the reduced development

cost of the new product, but the continuing existence of the product.

We believe that in situations such as that described by the

commenter, the existence of a benefit is directly dependent upon the

nature of the financial contribution. If a financial contribution has

been provided, either directly or indirectly, in a form which is

specifically identified in the statute or regulations (e.g., a loan, a

grant, an equity infusion, etc.), we will identify and measure the

resulting benefit in accordance with the rules contained in the statute

and regulations. If the financial contribution takes a form

[[Page 65361]]

which has not been specifically dealt with in these regulations, we

will identify and measure the benefit in accordance with the definition

of benefit contained in paragraph (b). Moreover, as noted above,

paragraph (b) provides sufficient flexibility to accommodate

circumstances in which the facts of a particular case indicate that a

financial contribution has conferred a benefit, even if the benefit

does not take the form of a reduction in input costs or an enhancement

of revenues.

Finally, one commenter objected to the following statement which

was included in the preamble to the 1997 Proposed Regulations: ``By the

same token, where a firm does not pay less for an input than it

otherwise would pay (or its revenues are not increased) as a result of

a financial contribution, it would be very difficult to contend that a

benefit exists.'' This commenter argued that we should not define the

types of practices which do not confer benefits as this would invite

the creation and exploitation of loopholes.

We agree that we need only provide a definition of what constitutes

a benefit. We believe we have given ourselves the flexibility to apply

the concept of benefit in such a way that we will be able to find a

benefit in situations in which the regulations do not contain specific

rules for identifying and measuring the benefit from a particular

government program or practice.

We received several comments regarding the extent to which the

Department should consider the overall ``effect'' a government program

has on a firm's behavior in determining whether a benefit exists. One

group of commenters requested an affirmative statement preserving the

Department's discretion to consider ``effects'' in appropriate

circumstances. Another group of commenters urged us to renounce any use

of our discretion and to state that the effects of government actions

are irrelevant to the existence of a countervailable subsidy.

As we explained in the preamble to the 1997 Proposed Regulations,

the determination of whether a benefit is conferred is completely

separate and distinct from an examination of the ``effect'' of a

subsidy. In other words, a determination of whether a firm's costs have

been reduced or revenues have been enhanced bears no relation to the

effect of those cost reductions or revenue enhancements on the firm's

subsequent performance, such as its prices or output. In analyzing

whether a benefit exists, we are concerned with what goes into a

company, such as enhanced revenues and reduced-cost inputs in the broad

sense that we have used the term, not with what the company does with

the subsidy. Our emphasis on reduced-cost inputs and enhanced revenues

is derived from elements contained in the examples of benefits in

section 771(5)(E) of the Act and in Article 14 of the SCM Agreement. In

contrast, the effect of government actions on a firm's subsequent

performance, such as its prices or output, cannot be derived from any

elements common to the examples in section 771(5)(E) of the Act or

Article 14 of the SCM Agreement.

For example, assume that a government puts in place new

environmental restrictions that require a firm to purchase new

equipment to adapt its facilities. Assume also that the government

provides the firm with subsidies to purchase that new equipment, but

the subsidies do not fully offset the total increase in the firm's

costs--that is, the net effect of the new environmental requirements

and the subsidies leaves the firm with costs that are higher than they

previously were.

In this situation, section 771(5B)(D) of the Act, which deals with

one form of non-countervailable subsidy, makes clear that a subsidy

exists. Section 771(5B)(D) of the Act treats the imposition of new

environmental requirements and the subsidization of compliance with

those requirements as two separate actions. A subsidy that reduces a

firm's cost of compliance remains a subsidy (subject, of course, to the

statute's remaining tests for countervailability), even though the

overall effect of the two government actions, taken together, may leave

the firm with higher costs. As another example, if a government

promulgated safety regulations requiring auto makers to install seat

belts in back seats, and then gave the auto makers a subsidy to install

the seat belts, we would draw the same conclusion. In the two examples,

the government action that constitutes the benefit is the subsidy to

install the equipment, because this action represents an input cost

reduction. The government action represented by the requirement to

install the equipment cannot be construed as an offset to the subsidy

provided to reduce the costs of installing the equipment.

Thus, if there is a financial contribution and a firm pays less for

an input than it otherwise would pay in the absence of that financial

contribution (or receives revenues beyond the amount it otherwise would

earn), that is the end of the inquiry insofar as the benefit element is

concerned. The Department need not consider how a firm's behavior is

altered when it receives a financial contribution that lowers its input

costs or increases its revenues.

If there were any doubt on this score, section 771(5)(C) of the Act

eliminates it by clarifying that the ``benefit'' and the ``effect'' of

a subsidy are two different things. While, as stated above, there must

be a benefit in order for a subsidy to exist, section 771(5)(C) of the

Act expressly provides that the Department ``is not required to

consider the effect of the subsidy in determining whether a subsidy

exists.'' This message is reinforced by the SAA at 926, which states

that ``the new definition of subsidy does not require that Commerce

consider or analyze the effect (including whether there is any effect

at all) of a government action on the price or output of the class or

kind of merchandise under investigation or review.''

Paragraph (c) of the new regulation further reinforces this

principle by stating affirmatively that, in determining whether a

benefit is conferred, the Department is not required to consider the

effect of the government action on the firm's performance, including

its prices or output, or how the firm's behavior otherwise is altered.

When we examine indirect subsidies, we are inquiring into whether a

government is entrusting or directing a private entity to provide a

reduced-cost input or enhanced revenue to a firm that produces the

subject merchandise. For example, we have investigated whether below-

market loans or reduced-cost goods have been provided by means of

indirect subsidies. This analysis in no way implies that we are

examining whether the indirect subsidy has an effect on the price or

output of the subject merchandise. It merely means that we are

investigating, in fulfillment of other statutory requirements, whether

loans were provided on non-commercial terms or whether goods were

provided for less than adequate remuneration.

In addition to those comments relating specifically to our proposed

definition of a benefit, we received comments on other topics which we

believe are appropriately addressed in the context of a discussion on

benefits. First, one commenter objected to the absence of a regulation

regarding so-called ``tiered'' programs. Tiered programs are those

programs which provide varying levels of government assistance based

upon differing eligibility criteria. Our longstanding practice

regarding such programs has been to countervail only the difference

between the assistance provided at a

[[Page 65362]]

non-specific level (within the meaning of section 771(5A) of the Act)

and the assistance provided to a specific enterprise or industry (or

group thereof). This practice was reflected in Sec. 355.44(n) of the

1989 Proposed Regulations.

Our omission of a similar rule in this round of regulations was an

oversight. To correct for this, we have added paragraph (d), which

provides that where varying levels of financial contributions are

provided, a benefit will be conferred to the extent that a specific

enterprise or industry or group thereof receives a greater level of

financial contribution than that provided at the non-specific level.

The varying financial contribution levels must be set forth in a

statute, decree, regulation, or other official act, and they must be

clearly delineated and identifiable (e.g., the investment tax credit

program in Certain Fresh Atlantic Groundfish from Canada, 51 FR 10041

(March 24, 1986)). We note, however, that this exception cannot apply

where the statute specifies a commercial test for determining the

benefit, such as with respect to loans and loan guarantees.

Another related topic involves the treatment of taxes on subsidies.

Typically, we have referred to this issue as the ``secondary tax

consequences'' of subsidies. Section 351.527 of the 1997 Proposed

Regulations stated that we would not take account of secondary tax

consequences. For example, if receipt of a grant increases the amount

of income tax paid by a firm, we do not reduce the amount of the

benefit from the grant to reflect the higher taxes paid. In these Final

Regulations, we have retained this rule and have relocated it to

Sec. 351.503(e).

We received two comments expressing support for the 1997 Proposed

Regulations. One of these commenters requested that we include in the

regulation the following corollary, which flows from the same basic

principle: where a subsidy is exempt from income tax, we will treat the

tax exemption as a separate benefit in addition to the benefit from the

original subsidy. An additional commenter requested that the regulation

be expanded to clarify that we will not consider any secondary

consequences or effects of the granting of the subsidy outside the

exclusive list of subsidy offsets designated by the statute. To this

end, this commenter advocated including the list of allowable offsets

in the regulations and stating that we will not consider secondary

consequences of the benefit. We have not added the requested language

because the statute is clear regarding what is considered to be an

allowable offset. Nor have we broadened the regulation as requested by

either commenter. We believe that the impact of the benefit under one

subsidy program should not be considered in calculating the benefit

under a separate program. However, in our experience, this question has

only arisen with respect to the impact of tax programs on other

programs. Therefore, a broader regulation is not necessary.

Section 351.504

Section 351.504 deals with the benefit attributable to the most

basic type of subsidy, a grant. In the 1997 Proposed Regulations,

paragraph (c) of this section (which was then numbered Sec. 351.503)

included our methodology for allocating over time the benefit from a

grant, or the benefit from a subsidy that the Department treated as a

grant. In these Final Regulations, we have broken out the allocation

issues from the grant section and created a separate section

(Sec. 351.524) which deals with the allocation of benefits to a

particular time period. Therefore, Sec. 351.504 now pertains only to

grants.

As in our 1997 Proposed Regulations, paragraph (a) provides that in

the case of a grant, a benefit exists in the amount of the grant.

Paragraph (b) sets forth the rule for determining when a firm is

considered to have received a subsidy provided in the form of a grant.

This paragraph provides that the Secretary will normally consider the

benefit as having been received on the date on which the firm received

the grant. In these Final Regulations, we have added the word

``normally'' for reasons explained in the preamble discussion of

Sec. 351.524. Finally, paragraph (c) provides that the benefit from a

grant will be allocated to a particular time period pursuant to the

methodology set forth in Sec. 351.524.

All the comments that we received regarding grants dealt with the

allocation of benefits. These comments are, therefore, discussed in the

preamble to Sec. 351.524.

Section 351.505

Section 351.505 deals with loans and other forms of debt financing.

Paragraph (a) deals with the identification and measurement of the

benefit attributable to a loan. Paragraph (a)(1) tracks the general

standard set forth in section 771(5)(E)(ii) of the Act, which directs

the Department to use a ``comparable commercial loan that the recipient

could actually obtain on the market'' as the benchmark in determining

whether a government-provided loan confers a benefit.

Use of Effective Interest Rates: Paragraph (a)(1) restates the

Department's current practice of normally seeking to compare effective

interest rates rather than nominal rates in making this comparison.

``Effective interest rates'' are intended to take account of the actual

cost of the loan, including the amount of any fees, commissions,

compensating balances, government charges (such as stamp taxes) or

penalties paid in addition to the ``nominal'' interest. However, where

effective rates are not available, we will compare nominal rates or, as

a last resort, nominal to effective rates, as under current practice.

If the ``loan'' is a bond (see definition of ``loan'' in Sec. 351.102),

we normally will treat the yield on the bond as the effective interest

rate.

One commenter asked that the regulations clarify that only payments

legitimately made on a loan will be used when calculating the effective

interest rate. The commenter urged the Department to exclude other,

unrelated payments to the government which the borrower might make

along with the loan payments.

We agree with this commenter that payments unrelated to the loan

should not be included when we calculate the effective interest rate,

but we do not believe that the regulation needs to be modified to

address this concern. The preamble clearly describes the types of

payments that would be included in calculating an effective interest

rate. However, we will examine whether there are requirements placed on

either the government loan or the benchmark loan affecting the cost of

borrowing that should be factored into the calculation of the benefit

amount.

Selection of Benchmark Loans and Interest Rates

Paragraphs (a)(2) and (a)(3) elaborate on the criteria for

selecting the benchmark. The criteria contained in these two paragraphs

are much more general (and, thus, much more flexible) than the detailed

hierarchies contained in Sec. 355.44(b) of the 1989 Proposed

Regulations. The Department seldom used these hierarchies because, in

practice, the information required in the 1989 Proposed Regulations was

seldom available.

``Comparable commercial loan'' defined: Paragraph (a)(2) sets forth

the criteria the Department normally will consider in selecting a

comparable commercial loan. First, paragraph (a)(2)(i) defines the term

``comparable.'' In the preamble to the 1997 Proposed Regulations, we

stated that in order to be used as a benchmark, a comparable

[[Page 65363]]

commercial loan should represent a financial instrument that is similar

to the government-provided loan and that was taken out (or could have

been taken out) at the same time. To identify a loan that is comparable

to the government-provided loan, the 1997 Proposed Regulations called

for primary emphasis to be placed on the structure of the loans (e.g.,

fixed interest rate v. variable interest rate), the maturities of the

loans (e.g., short-term v. long-term), and the currencies in which the

loans are denominated.

Several commenters maintained that it is not enough to look at the

structure, maturity, and currency denomination to identify a benchmark

loan that is comparable to the government-provided loan. These

commenters argued that the Department should also consider the level of

risk associated with the loans by comparing the security or collateral

that the borrower is required to provide for each loan. One of the

commenters observed that this approach would be consistent with the

Department's practice in Laminated Hardwood Trailer Flooring from

Canada, 62 FR 5201 (February 4, 1997). This commenter also noted that,

while the risk element was discussed in the preamble of the 1997

Proposed Regulations, it did not appear in the regulation.

In opposition, another commenter argued that a commercial loan

should be considered sufficiently comparable to a government loan when

the structures and maturities of the two loans are identical or similar

and the loans are provided in the same currency. This commenter argued

that in the interest of predictability and uniformity, no further

analysis, particularly with regard to the level of security of a loan,

should be necessary. This commenter asserted that, where these three

criteria are met, the loans would generally require the same level of

security. Comparing the value of different assets securing different

loans would create an unworkable test, according to the commenter, who

suggested that the Department at least make it a rebuttable presumption

that a commercial and a government-provided loan are comparable if the

three criteria listed above match.

We have not adopted the proposals put forward by either set of

commenters. As in the 1997 Proposed Regulations, Sec. 351.505(a)(2)(i)

states that we intend to place primary emphasis on three basic

characteristics in determining whether particular loans are comparable

to a government-provided loan: The structure, maturity, and currency

denomination of the loans. This does not mean, however, that a loan in

the same currency with a similar structure and maturity will always be

found comparable to the government-provided loan. Nor should our

decision to place primary emphasis on these three characteristics be

seen as a rebuttable presumption.

Instead, we recognize that many characteristics could factor into a

decision of whether a loan should be considered comparable to the

government-provided loan. Certainly, as the first set of commenters has

pointed out, the levels of security or collateral on the two loans

could be relevant in determining comparability. Similarly, the amounts

of principal might differ so greatly that the two loans should not be

compared. However, rather than identifying numerous characteristics for

finding loans to be comparable, and thereby limiting our ability to

find benchmarks, we have continued to place primary emphasis on what we

believe to be the three most important characteristics. Regarding other

characteristics that might render particular loans not comparable to

the government-provided loan, such as collateral and size, we will

consider arguments made by the parties based on the facts presented in

their cases.

Paragraph (a)(2)(ii) provides a definition of the term

``commercial.'' The 1997 Proposed Regulations stated that we would

normally treat a loan as ``commercial'' if it were taken out from a

commercial lending institution or if it were a bond issued by the firm

in commercial markets. We also stated that a loan provided under a

government program, even if the program is not specific to an

enterprise or industry, would not be considered a ``commercial'' loan

for benchmark purposes. Finally, the 1997 Proposed Regulations stated

that the Department would treat a loan from a government-owned bank as

a commercial loan, unless there was evidence that the loan was provided

at the direction of the government or with government funds.

We received several comments on this issue, all of which urged us

not to use loans from government-owned banks for benchmark purposes.

One commenter asserted that a loan from a government-owned bank is the

same as a loan from the government, regardless of whether the loan is

provided under a government program, because the actions of a

government-owned bank are presumably consistent with the policies of

its owner, the government. A second commenter maintained that the

distinction between ``a government program'' and ``government control''

is blurred and pointed to the Department's determination in Certain

Steel Products from Korea, 58 FR 37338 (July 9, 1993), where the

Department found that a countervailable benefit was conferred by

government-directed, preferential access to specific sources of credit

offered at favorable terms. Because of the availability of ``directed

credit'' such as that found in the Korean case, this commenter argued

that the Department should not use rates from loans provided by

government-owned banks as benchmark rates. A third commenter argued

that the Department should not use loans from government-owned banks

for benchmark purposes unless the respondent can demonstrate the

commercial nature of such loans. This and other commenters objected to

the burden that the 1997 Proposed Regulations allegedly placed upon a

petitioner to show that a loan from a government-owned bank is provided

at the direction of the government or with government funds. Noting

that the 1989 Proposed Regulations directed the Department to use

financing provided or directed by the government as a benchmark only

under certain exceptional circumstances, several commenters urged the

Department to continue to apply this narrow standard.

We have traditionally recognized that government-owned banks may

operate as commercial banks in some countries. It is not appropriate to

maintain that loans from government-owned banks per se are not

commercial. Therefore, we continue to take the positions that: (1) We

will not consider loans provided under government programs to be

commercial loans, and (2) we will not automatically disqualify loans

from government-owned commercial banks as benchmarks. However, we will

not use loans from government-owned special purpose banks, such as

development banks, as benchmarks because such loans are similar to

loans provided under a government program or at the direction of the

government. Regarding loans from government-owned commercial banks, we

will treat such loans as being commercial and use them as benchmarks

unless they are made on non-commercial terms or are provided at the

direction of the government. We do not believe that this standard

imposes an unreasonable burden on petitioners because this is the type

of information they would routinely provide when alleging that

government-provided loans are countervailable.

Further, regarding the definition of ``commercial,'' where a firm

receives a financing package including loans from both commercial banks

and from the government, we intend to examine the package closely to

determine whether

[[Page 65364]]

the commercial bank loans should in fact be viewed as ``commercial''

for benchmark purposes. In particular, we will look to whether there

are any special features of the package that would lead the commercial

lender to offer lower, more favorable terms than would be offered

absent the government/commercial bank package.

Paragraphs (a)(2)(iii) and (iv) specify the time period from which

the Department will select comparable financing. Paragraph (a)(2)(iii)

addresses long-term loans and is unchanged from the 1997 Proposed

Regulations. This regulation directs us to use a loan whose terms were

established during or immediately before the year in which the terms of

the government-provided loan were established. Paragraph (a)(2)(iv)

addresses short-term loans. In the 1997 Proposed Regulations, we stated

that we would use as the benchmark rate an annual average of the

interest rates on comparable commercial loans taken out during the

period of investigation or review. However, in cases with significantly

fluctuating interest rates, the 1997 Proposed Regulations allowed us to

use ``the most appropriate'' interest rate as the benchmark rate.

We received two comments regarding the benchmark interest rate for

short-term loans. Both commenters argued against using a simple average

of the interest rates on comparable commercial short-term loans

obtained by the respondent. Instead, they asked the Department to

weight the rates by the associated principal amount of each loan in

order to prevent small, one-time loans from distorting the benchmark

calculation. According to the commenters, this change would also

address the Department's concern about significantly fluctuating

interest rates.

We have adopted the commenters' proposal in part and have amended

paragraph (a)(2)(iv) to provide that we will calculate a weighted

rather than a simple average benchmark interest rate for short-term

loans. However, we do not share the commenters' view that this change

addresses situations where the interest rate fluctuates significantly

over the year, e.g., in economies with a high inflation rate. We are,

therefore, retaining the provision that allows us to use benchmarks

other than annual weighted averages in these situations.

We also wish to clarify that we intend to follow our practice of

calculating short-term benchmarks on a calendar year basis. In most

instances, the period of investigation or review is a calendar year, so

the short-term benchmark will be calculated using commercial loans that

were obtained (or could have been obtained) during the period of

investigation or review. In situations where the loans under

investigation span two calendar years, we will calculate two annual

benchmarks corresponding to the two years.

Finally, we received one comment on the selection of benchmark

interest rates to be used in administrative reviews of suspension

agreements. In the preamble to the 1997 Proposed Regulations, we stated

that in administering a suspended investigation, we would monitor

developments in commercial benchmarks outside of the normal

administrative review process and that this monitoring activity should

serve to ensure that the commercial benchmarks used were timely. The

commenter, however, claimed that a special regulation requiring the

Department to monitor commercial benchmark rates is needed because

otherwise there is no guarantee that the Department will do so. In the

commenter's experience, the Department has not always undertaken this

type of monitoring activity. Specifically, pointing to Miniature

Carnations and Roses and Other Fresh Cut Flowers from Colombia, 59 FR

52514 (October 18, 1994), the commenter alleged that the Department set

new benchmarks at the conclusion of each administrative review, with

the result that the interest rates used for purposes of the suspension

agreement always lagged behind the contemporaneous commercial rates.

For short-term loans, the commenter argued, the Department should

monitor commercial interest rates on at least a quarterly basis in

order to keep the suspension agreement current.

We do not agree with the commenter's view that a regulation is

needed on this issue. In the case of suspension agreements, we will

revise the benchmarks for long- and short-term loans whenever

appropriate, regardless of whether we are conducting an administrative

review of the suspension agreement. To ensure that the benchmarks are

kept as current as possible, we intend to review them once a year or

more frequently, if information available to the Department indicates

that a change is necessary.

``Could actually obtain on the market'' defined: In accordance with

section 771(5)(E)(ii) of the Act, paragraph (a)(3) addresses the

requirement that the comparable loan be one that the firm ``could

actually obtain on the market,'' and reflects a change in our practice

with respect to short-term loans. In the past, we have used national

average interest rates to determine the benefit from government-

provided short-term loans. This practice was codified in

Sec. 355.44(b)(3) of the 1989 Proposed Regulations. However, as early

as 1989, we announced that we would consider using company-specific

benchmarks for short-term loans. Based upon our experience in the

interim, and especially because of the ability to computerize our loan

calculations, we have concluded that we have the capability to use

company-specific benchmarks. Moreover, we believe that company-specific

benchmarks provide a more accurate measure of the benefit, if any, to a

recipient of a government-provided short-term loan. Therefore,

paragraph (a)(3)(i) states a preference for using company-specific

benchmarks for both short- and long-term loans. Under paragraph

(a)(3)(ii), we normally would use national averages only in the event

that the firm did not take out any comparable commercial loans during

the relevant period. Except for a minor clarification (adding ``for

both short- and long-term loans'' to paragraph (a)(3)(i)), these

paragraphs are unchanged from the 1997 Proposed Regulations.

Two commenters warned against using the interest rates on

hypothetical loan offers as benchmark rates. One of the commenters

pointed to a perceived loophole in the preamble to the 1997 Proposed

Regulations, which stated that ``a comparable commercial loan used as a

benchmark should represent a financial instrument * * * that was taken

out (or could have been taken out) at the same point in time.'' Another

commenter suggested that the acceptance of hypothetical loan offers for

benchmark purposes might tempt respondents to manipulate the benchmark

rate by soliciting offers of loans that they do not intend to take.

Both commenters asserted that the interest rates on such hypothetical

loan offers would be very low and that they would, thus, distort the

benchmark rate.

We agree that respondents should not be permitted to submit

hypothetical loans for use as benchmarks. The language in the preamble

cited by the commenter was meant to address another situation: Where

the respondent did not actually take out any commercial loans during

the relevant period and where we, therefore, would use an appropriate

alternative benchmark interest rate * * * such as a national average

interest rate. The national average interest rate is representative of

a loan that ``could have been taken out.''

Benchmark for uncreditworthy companies: Paragraph (a)(3)(iii),

which deals with long-term loans provided to firms considered to be

uncreditworthy, describes our methodology for

[[Page 65365]]

calculating the benchmark that we will use in identifying and measuring

the benefit attributable to a government-provided, long-term loan

received by an uncreditworthy firm. One important aspect of this

methodology has changed from the 1997 Proposed Regulations.

Our methodology is based explicitly on the notion that, when a

lender makes a loan to a company that is considered to be

uncreditworthy (as opposed to a safer, creditworthy company), the

lender faces a higher probability that the borrower will default on

repayment of the loan. As a consequence of this higher probability of

default, the lender will charge a higher interest rate. The calculation

described in paragraph (a)(3)(iii) addresses the increased probability

of default for an uncreditworthy company by adjusting upward the

interest rate for a creditworthy company in the country in question.

As stated in the 1997 Proposed Regulations, in making this

adjustment, we are not proposing to calculate the probability that a

particular uncreditworthy firm will default on a particular loan. Such

a calculation would require extensive data and analysis, and any

conclusion would be highly speculative. Instead, similar to the method

we have used since 1984, we will rely on information regarding the U.S.

debt market. In the 1997 Proposed Regulations, we stated that we would

use the weighted average one-year default rate for speculative grade

bonds, as reported by Moody's Investor Service. This weighted average

default rate would be reflected indirectly in our formula for

calculating the benchmark interest rate for uncreditworthy companies,

which is based on the probability that these risky loans will be

repaid.

We received numerous comments on our new methodology. One commenter

expressed support for the methodology, stating that it seemed to

calculate accurately the full benefit of a loan subsidy. Certain other

commenters supported the new methodology as long as it resulted in a

``substantial spread'' between the observed commercial interest rates

in the country under investigation and the benchmark interest rate used

for uncreditworthy companies.

One commenter did not object to the new methodology but argued

that, in calculating the risk premium, the Department should use data

pertaining to the country under investigation, not U.S. data, which

should only be used as facts available.

Another commenter criticized the reliance upon default rates in the

U.S. ``junk'' bond market, arguing that U.S. data do not reflect the

risk of lending to uncreditworthy companies in foreign countries,

especially developing countries where the default rate is likely to be

much higher. This commenter also criticized the use of a one-year

default rate in the calculation of the risk premium, arguing that this

significantly understates the overall default rate because default is

more likely after the first year of the life of a loan. Should the

Department decide to rely on U.S. market data, the commenter asked that

the Department, at a minimum, examine the default rate over 10 years.

Another commenter stated that the Department's new methodology

implies a serious departure from the statutory mandate to determine an

interest rate that the borrower could actually obtain on the market.

First, the commenter argued, a default-based premium does not take into

account all the costs associated with lending to an uncreditworthy

company, e.g., collection costs and lost opportunity costs and, as a

result, the premium is understated. Second, the commenter asserted, the

new methodology treats all uncreditworthy borrowers as if they were

large corporate borrowers able to issue junk bonds of the kind reported

by Moody's. According to this commenter, many companies cannot obtain

long-term loans even at junk bond rates and are forced to rely on

borrowing from the venture capital market at substantially higher

interest rates. In reality, the commenter argued, a private lender

would assess a company's creditworthiness on a case-by-case basis using

the same financial indicators that the Department has relied upon in

the past (see Sec. 355.44(b)(6)(i) of the 1989 Proposed Regulations).

The regulations, therefore, should reflect such private lender behavior

by directing the Department to determine the risk premium on a case-by-

case basis.

Finally, two commenters noted that the European Union (``EU'')

takes a tougher stance on government loans to uncreditworthy borrowers

by treating the entire loan as a grant when the recipient company's

financial position is so weak that it could not have obtained a

commercial loan, and implied that the Department should follow the EU's

example.

As stated in the 1997 Proposed Regulations, we are changing our

methodology because we believe that the new methodology more

appropriately reflects the risk involved in lending to firms with

little or no access to commercial bank loans from conventional sources.

By adjusting upward the interest rate that an average, creditworthy

company would pay to account for the greater likelihood of default by

an uncreditworthy company, we recognize the speculative nature of loans

to uncreditworthy borrowers and the premium they would have to pay the

lender to assume that risk.

We have continued to rely on default information pertaining to the

United States in our formula because we believe it would be difficult

to locate detailed and comprehensive default information for many of

the countries that we investigate. However, if such data do exist and

are brought to our attention in the course of an investigation or

review, and the data indicate that the default experience in the

country in question differs significantly from that in the United

States, we would consider using the default rate from the country under

investigation. Therefore, we have amended the 1997 Proposed Regulation

to say that the Secretary ``normally'' will calculate the benchmark for

uncreditworthy companies using U.S. data.

We have not adopted the suggestion that we follow the EU's practice

of treating loans to uncreditworthy firms as grants. Under our

definition, uncreditworthy firms are those that cannot obtain long-term

loans from conventional commercial sources. This does not mean,

however, that they cannot borrow funds from other sources. Hence, we

would not equate loans to these companies with grants. Instead, the

purpose of our methodology is to capture the increased risk of lending

to these companies.

Regarding the new calculation methodology, we agree that using a

one-year default rate would not accurately reflect the risk that an

uncreditworthy borrower will default on a long-term loan. We have,

therefore, changed this aspect of our methodology and will use the

average cumulative default rate for the number of years corresponding

to the length of the loan, as reported in Moody's study of historical

corporate bond default rates. In other words, we would use a five-year

default rate for a five-year loan, a 15-year default rate for a 15-year

loan, and so forth. We believe that using a default rate that is

directly linked to the term of the loan is a better reflection of the

risk associated with long-term lending to uncreditworthy borrowers.

Our formula for calculating the benchmark interest rate for an

uncreditworthy company is based upon the assumption that a lender's

expected return on all loans should be equal. Under this assumption,

the interest rate differential on loans charged to

[[Page 65366]]

creditworthy and uncreditworthy companies is such that the lender's

expected (total) return on a loan to an uncreditworthy company equals

the expected (total) return on a loan to a creditworthy company, after

accounting for differences in the risk of default. A second assumption

is that, in the event of default, no portion of the principal or

interest is recovered by the lender. The following equation relates the

loan rate to a creditworthy company and the loan rate to an

uncreditworthy company:

(1-qn)(1+if)n = (1-pn)(1 +

ib)n,

Where:

n = the term of the loan;

ib = the benchmark interest rate for uncreditworthy

companies;

if = the long-term interest rate that would be paid by a

creditworthy company;

pn = the probability of default by an uncreditworthy company

within n years; and

qn = the probability of default by a creditworthy company

within n years.

Default means any missed or delayed payment of interest and/or

principal, bankruptcy, receivership, or distressed exchange. For values

of pn, we will normally rely on the average cumulative

default rates reported for the Caa to C-rated categories of companies

in Moody's study of historical default rates of corporate bond issuers.

For values of qn, we will normally rely on the average

cumulative default rates reported for the Aaa to Baa-rated categories

of companies in Moody's study of historical default rates of corporate

bond issuers.

Solving for ib in the above equation yields a formula

for the benchmark interest rate that should be paid by an

uncreditworthy borrower:

ib = [(1-qn)(1+if)n/

(1-pn)]1/n-1.

One commenter urged the Department to apply a risk premium also to

short-term loans taken out by uncreditworthy borrowers. Another

commenter supported this idea, arguing that even though long-term

financing is riskier, a bank's decision on short-term loans is also

based on the overall financial health of the borrower.

The fact that we are using a company-specific benchmark means that

the risk associated with providing a short-term loan to a company will

be reflected without any special adjustment. However, even where a

company-specific benchmark is not available, we do not believe it would

be appropriate to include a risk premium in the short-term benchmark

calculation. Short-term lending is less risky than long-term lending

and the inclusion of a risk premium in the short-term benchmark would

overcompensate for the commercial default risk. The risk of default in

short-term lending is minimal because short-term lending is usually

associated with specific transactions, and these transactions provide

security for the lender (albeit by means of a wide variety of legal

modalities). Thus, we have not adopted this suggestion.

We note that we have identified one situation where it would be

appropriate to include a risk premium in a short-term benchmark. This

would arise if we were forced to use a short-term interest rate as a

benchmark for long-term loans to an uncreditworthy company or as a

discount rate for allocating benefits received by an uncreditworthy

company.

Creditworthiness Analysis

Paragraph (a)(4) sets forth the standard for determining whether a

firm is uncreditworthy. In the 1997 Proposed Regulations, we made

certain modifications to Sec. 355.44(b)(6)(i) of the 1989 Proposed

Regulations to clarify the analysis we intended to undertake in

determining whether a company is creditworthy. Specifically, we adopted

a broader definition of ``uncreditworthiness'' where we would find a

company to be uncreditworthy if information available at the time the

terms of the government-provided loan were agreed upon indicated that

the firm could not have obtained long-term financing from conventional

commercial sources. In this context, the term ``conventional commercial

sources'' referred to bank loans and non-speculative grade bond issues.

Hence, uncreditworthy companies were those that would be forced to

resort to other sources, such as junk bonds, to raise funds. We also

listed factors we would consider in making a creditworthiness

determination. These factors focused on the financial position of the

firm receiving the government financing, without any consideration of

the purpose of the financing or whether different levels of risk might

be associated with different types of projects undertaken by the firm.

We received several comments on our definition of

``uncreditworthiness.'' Certain commenters urged the Department to

retain the definition of uncreditworthiness from the 1989 Proposed

Regulations, arguing that this standard was objective, uncontroversial,

and easy to administer. These commenters maintained that this standard

provided important guidance for petitioners who may have difficulties

obtaining information on the loan options available to respondents. The

commenters also argued that the new regulation would place a nearly

impossible burden of proof on petitioners to demonstrate that a

respondent is uncreditworthy.

We have not adopted this suggestion. As we stated in the preamble

to our 1997 Proposed Regulations, we changed the definition from the

1989 Proposed Regulations because we found that the old definition did

not contain a general principle to guide our determinations of

uncreditworthiness. Instead, the 1989 Proposed Regulation relied on a

formulaic approach to determining creditworthiness that was too

restrictive. We believe that the general principle adopted in these

regulations (i.e., an uncreditworthy firm is one which could not have

obtained long-term financing from conventional sources) will give us

the flexibility to address situations that would not have met the

formulaic approach for finding a company uncreditworthy.

However, although we changed the definition of uncreditworthiness,

we did not intend to change the standard for initiating an

investigation of a company's creditworthiness. Therefore, petitioners

may continue to provide the same type of information we have typically

relied upon.

Another commenter argued that the Department should not limit

itself to examining the creditworthiness of firms as a whole, but

should also give itself the flexibility to examine the creditworthiness

of individual projects. This commenter argued that some foreign

manufacturers, though creditworthy per se, are able to carry out new

development projects only because they obtain government financing. The

commenter argued that these manufacturers would not have been able to

secure financing from commercial sources for their huge development

projects because these projects are not commercially viable and would

be impossible to finance without government subsidies. The commenter

noted that, under the Department's traditional approach, the Department

would analyze the creditworthiness of the company as a whole, not the

creditworthiness of the specific project. Hence, the Department would

be likely to find the foreign manufacturer creditworthy, regardless of

the commercial viability of the project. The commenter argued that, in

this type of situation, the Department should focus on the

creditworthiness of the project, not the firm.

We share this commenter's concern and have amended the 1997

Proposed Regulations to allow for a project-specific analysis in

determining

[[Page 65367]]

creditworthiness. For example, for loans that are provided to fund a

large investment project into new products, processes, or capacity

(e.g., a plant expansion or new model or product line, where repayment

of a loan is contingent upon the success of the particular project

being funded), our traditional analysis focusing primarily on the

creditworthiness of the company as a whole may be inappropriate because

the risk associated with a new project may be much higher or lower than

the average risk of the company's existing operations. In these

situations, we would expect commercial lenders to place greater

emphasis on the expected return and risk of the project because the

success or failure of the project would be the most important indicator

of the borrowing firm's ability to repay the loan. This is not to say

that the financial position of the firm as a whole would be irrelevant

to the lender's decision, only that the primary focus would be on the

project itself. Therefore, paragraph (a)(4) now allows for the

possibility of focusing the creditworthiness analysis on the project

being financed rather than the company as a whole.

Significance of long-term commercial loans: In the 1997 Proposed

Regulations, paragraph (a)(4)(ii) provided that, if a privately-owned

company received long-term commercial loans without a government loan

guarantee, we would consider the presence of such commercial loans as

dispositive evidence that the company was not uncreditworthy.

Two commenters criticized the Department's proposed approach. These

commenters maintained that the presence of a long-term, commercial loan

does not prove that a company is creditworthy. Instead they urged the

Department to examine all the criteria listed in paragraphs (a)(4)(i)

(A), (B), (C), and (D) without treating one of these factors as

dispositive. One of the commenters argued that giving one criterion

dispositive status would constitute abuse of the Department's

discretion to implement the statute. The other commenter argued that

the Department's proposed approach would preclude an in-depth review of

the company as envisioned by the regulations. Both commenters stated

that making the presence of a commercial loan a dispositive indication

of creditworthiness would be particularly inappropriate if the

commercial loan had characteristics different from the government loan

(e.g., different requirements of security).

In general, we believe that if commercial banks are willing to

provide loans to the firm, we should not substitute our judgment and

find the firm to be uncreditworthy. This does not mean, however, that

if the firm has taken out a single commercial bank loan we would find

that loan to be dispositive evidence that the firm was creditworthy.

Instead, the intent of this paragraph is to indicate that, where the

firm has recourse to commercial sources for loans, as made evident by

the receipt of such loans, and the commercial loans are comparable with

the government loan, those loans will be dispositive of the firm's

creditworthiness. However, if, for example, the firm has obtained a

single commercial loan in the year in question for a relatively small

amount, and the loan has a short repayment term (e.g., less than two

years), or has unusual aspects, receipt of that loan will not be

dispositive of the firm's creditworthiness, and we will go on to

examine the other factors listed in paragraph (a)(4)(i) B through D.

We have also made a change from the 1997 Proposed Regulations

regarding the presence of guarantees and the firm's creditworthiness.

We have added ``explicit or implicit'' to modify ``government

guarantee.'' This serves to clarify our position that if either type of

guarantee is present, the commercial loans will not be viewed as

dispositive of the firm's creditworthiness. We may consider a

commercial loan to be covered by an implicit government guarantee where

the loan contributes to the financing of a project that is being

undertaken in conjunction with government loan funds or other types of

government participation such as development grants. In such a

scenario, while no explicit government guarantee is present, we believe

that banks are likely to assume that the government will stand behind

the project and ensure that creditors are repaid.

Finally, we note our longstanding practice that creditworthiness

determinations are made on a year-by-year basis. For example, if we are

trying to determine whether a firm is creditworthy in 1998, we will

look to whether the firm has negotiated commercial loans in 1998.

One commenter suggested that purchases of equity in a company by a

commercial institution should also constitute dispositive evidence of

creditworthiness. The commenter reasoned that a private entity willing

to invest in a company would presumably also be willing to lend money

to that company because investing is riskier than lending.

We have not adopted this suggestion. By its very terms, equity

differs from loans and, hence, the presence of equity investments (even

if made by private investors) is not necessarily indicative of whether

the firm could obtain loans from commercial sources. As an extreme

example, private owners may inject equity into their company because

the debt-to-equity ratio is so high that it has become virtually

impossible for the company to borrow funds. Clearly, in this situation,

the presence of equity purchases by the owners would not be indicative

of the firm's access to commercial loans.

We received two comments regarding the significance of the receipt

of a commercial loan where we are examining the creditworthiness of a

government-owned company. One commenter suggested that paragraph

(a)(4)(ii) should apply also to government-owned firms. Another

commenter took the opposite view, stating that it is not unusual to

find commercial lenders providing loans to government-owned companies

which are otherwise uncreditworthy.

We do not believe that the presence of commercial loans is

dispositive of whether a government-owned firm could have obtained

long-term financing from conventional commercial sources. This is

because, in our view, in the case of a government-owned firm, a bank is

likely to consider that the government will repay the loan in the event

of default. Accordingly, paragraph (a)(4)(ii) provides that the

presence of comparable commercial loans will be dispositive of

creditworthiness only for privately owned companies. For government-

owned firms, we will make our creditworthiness determination by

examining this factor and the other factors listed in paragraph

(a)(4)(i).

Significance of prior subsidies: Paragraph (a)(4)(iii) in the 1997

Proposed Regulations stated that we would ignore current and prior

countervailable subsidies in determining whether a firm is

uncreditworthy. In other words, we would not attempt to adjust a firm's

financial data for current and prior subsidies in making a

creditworthiness determination.

We received three comments on this issue, all of which urged the

Department to change its approach and adjust for prior subsidies when

examining a firm's creditworthiness. One of these commenters requested

that the Department take prior subsidies into account to the same

extent that a reasonable private lender would. This commenter argued

that, by ignoring prior subsidies, the Department is not adhering to

the standards of a reasonable private lender. The commenter maintained

that, if a

[[Page 65368]]

company's financial health is due to government assistance, a private

lender would examine the company's underlying performance independent

of subsidies. The private lender, who would then discover that the

company's financial health was superficial, might not lend money to the

company unless the lender was convinced that the government would

continue to provide subsidies in the future. A second commenter argued

that failure to consider prior subsidies when making a creditworthiness

determination underestimates the benefit received. This commenter urged

the Department to estimate the recipient company's financial situation

without subsidies and base its creditworthiness determination on this

estimate.

We have not adopted this suggestion. Our longstanding practice has

been not to take current or prior subsidies into account when

determining a company's creditworthiness. We believe that trying to

adjust a company's financial ratios for previously received subsidies

would be an extremely difficult and highly speculative exercise.

We have made one small amendment to paragraph (a)(4)(iv) addressing

the discount rate. We have changed ``non-recurring grant'' to ``non-

recurring benefit'' to conform with the new nomenclature used in

Sec. 351.524.

Calculation of Benefit From Long Term Variable Rate Loans

Paragraph (a)(5) deals with long-term variable rate loans and

codifies the methodology set forth in the GIA. Under paragraph

(a)(5)(i), which is unchanged from the 1997 Proposed Regulations, the

year in which the terms of the government-provided loan are set

establishes the reference point for comparing the government-provided

variable-rate loan with the comparable commercial variable-rate loan.

If the interest rate on the government-provided loan is lower than the

interest rate on the comparable commercial loan, a benefit exists. If

the interest rate on the government-provided loan is the same or

higher, no benefit exists. The rationale for basing the decision on the

first-year interest rate differential is that the interest rate spread,

if any, in that year generally will apply throughout the life of the

loan.

Paragraph (a)(5)(ii) recognizes that there may be situations where

the method described in paragraph (a)(5)(i) cannot be followed and

provides the Department with the discretion to modify that method. For

example, there may be no comparable commercial variable-rate loan to

use for comparison purposes, or the repayment structure of the

government-provided variable-rate loan may be such that the simple

interest rate comparison described in paragraph (a)(5)(i) would not

yield an accurate measure of the benefit.

Allegations

Paragraph (a)(6)(i) deals with the standard for initiating an

investigation of a respondent company's creditworthiness. It is

unchanged from the 1997 Proposed Regulations. In accordance with our

past practice, this paragraph states that the Secretary will normally

require a specific allegation before the Department will consider the

creditworthiness of a firm.

One commenter argued that the Department should not employ a

heightened initiation standard for investigating a company's

creditworthiness. Specifically, this commenter suggested that the

requirement that petitioners supply information ``establishing a

reasonable basis to believe or suspect'' that a company is

uncreditworthy be replaced with information ``reasonably available to

petitioners.''

We have not adopted this suggestion. The requirement that

petitioners establish ``a reasonable basis to believe or suspect''

uncreditworthiness rather than merely provide ``information reasonably

available'' to them dates back to the 1989 Proposed Regulations.

Because of the additional workload involved in investigating and

determining whether a company is uncreditworthy, we continue to believe

that it is appropriate to impose a higher standard for

uncreditworthiness allegations. This does not involve any change in our

past practice--the same types of allegations that we have accepted in

the past will still suffice to start a creditworthiness inquiry.

Paragraph (a)(6)(ii) establishes the evidentiary standard for

investigating loans extended by government-owned banks. In the 1997

Proposed Regulations, we made a distinction between government-owned

banks that are operated to meet special financing needs and government-

owned commercial banks. For special purpose banks (such as national

development banks), we asked that petitioners provide information

reasonably available to them indicating that loans provided by such

banks were specific and that the interest charged was not at commercial

rates. For government-owned commercial banks, we requested that

petitioners also provide information establishing a reasonable basis to

believe or suspect that the loans were something more than mere

commercial loans. In particular, we requested information suggesting

that such loans were provided at the direction of the government or

with funds provided by the government.

Several commenters objected to the higher initiation standard for

loans provided by government-owned commercial banks. They argued that

the additional information required by the Department for initiating an

investigation of loans from this category of banks is not reasonably

available to petitioners. They contended that it should be sufficient

for petitioners to demonstrate that a loan is specific and provided on

terms inconsistent with commercial considerations. They suggested that

the burden of proof be shifted to respondents to show that the loan

involves no government funds or government direction. Another commenter

asserted that the division of government-owned banks into two

categories is a new approach and not part of the Department's past

practice. The same commenter argued that the Department's 1997 Proposed

Regulations would create a loophole because the Department's threshold

for initiating an investigation of loans from government-owned

commercial banks would be higher than for initiating an investigation

of loans from privately-owned banks and government-owned special

purpose banks.

Based on our consideration of these comments, we have decided that

the distinction between government-owned special purpose banks and

government-owned commercial banks may not be helpful in this context

and that it is, therefore, not meaningful to retain different

initiation standards for investigating loans from these two categories

of banks. Paragraph (a)(6)(ii) has, thus, been changed and now provides

that, for loans provided by any government-owned bank, the Secretary

will require petitioners to present information reasonably available to

them indicating that the loans: (1) Are specific in accordance with

section 771(5A) of the Act, and (2) are provided on terms more

favorable than those the recipient would pay on a comparable commercial

loan that the recipient could actually obtain on the market. This

initiation standard is consistent with the initiation standard for most

subsidy allegations, i.e., petitioner must allege (and provide

reasonably available information in support of the allegation) that the

subsidy is specific and that it confers a benefit. We believe that, f

This text is long and has been trimmed here. Open the source document for the complete record.