General Electric Co. v. Environmental Protection Agency

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued December 3, 2001 Decided May 17, 2002

No. 00-1394

General Electric Company,
Petitioner

v.

Environmental Protection Agency,
Respondent

On Petition for Review of an Order of the
Environmental Protection Agency

Angus Macbeth argued the cause for petitioner. With him
on the briefs were Patricia K. Casano, Christopher L. Bell
and Timothy K. Webster.

H. Michael Semler, Attorney, U.S. Department of Justice,
argued the cause and filed the briefs for respondent.

Before: Ginsburg, Chief Judge, and Randolph and Tatel,
Circuit Judges.

Opinion for the Court filed by Chief Judge Ginsburg.

 Ginsburg, Chief Judge: General Electric Co. petitions for
review of the "PCB Risk Assessment Review Guidance Docu-
ment" issued by the Environmental Protection Agency. The
parties dispute (1) whether this case is ripe for review; (2)
whether the Document is a "rule" within the meaning of
s 19(a) of the Toxic Substances Control Act (TSCA), and
hence whether the court has jurisdiction to review its promul-
gation; and (3) whether the Agency should have followed the
procedures required for rulemaking in the TSCA and in the
Administrative Procedure Act when it promulgated the Docu-
ment. We conclude that the case is ripe for review, and that
the Guidance Document is a legislative rule such that the
court does have jurisdiction to entertain GE's petition and the
Document should not have been issued without prior notice
and an opportunity for public comment.

I. Background

The TSCA prohibits the manufacture, processing, distribu-
tion, and use (other than in a "totally enclosed manner") of
polychlorinated biphenyls (PCBs) unless the EPA determines
that the activity will not result in an "unreasonable risk of
injury to health or the environment." 15 U.S.C. s 2605(e)(2)
& (3). The Guidance Document governs the application of
two regulations promulgated by the EPA under the TSCA to
provide respectively for the cleanup and disposal of PCB
remediation waste and for the disposal of PCB bulk product
waste. See 40 C.F.R. ss 761.61 ("cleanup and disposal op-
tions for PCB remediation waste"), 761.62 (how "PCB bulk
product waste shall be disposed").

Under subsection (c) of each regulation a party may apply
for permission to use a method other than one of the generic
methods set out in the regulations for sampling, cleaning up,
or disposing of PCB remediation waste, or for sampling or
disposing of PCB bulk product waste. The EPA will approve
applications under these subsections if the alternative method
proposed does "not pose an unreasonable risk of injury to
health or the environment." Id. The regulations do not,

however, tell applicants how to conduct the necessary risk
assessment.*

That is where the Guidance Document comes in. It "pro-
vide[s] an overview of risk assessment techniques, and guid-
ance for reviewing risk assessment documents submitted
under the final PCB disposal rule." Guidance Document at
10. Of particular relevance to this case, in the Guidance
Document the EPA also explains that an applicant seeking to
use an alternative method under s 761.61(c) may take either
of two approaches to risk assessment. Id. at 21, 42. First,
the applicant may calculate cancer and non-cancer risks sepa-
rately. Id. To calculate cancer risks the applicant would
have to use a cancer potency factor recognized by the EPA.
Such cancer potency factors range, depending upon the expo-
sure pathway and upon the composition of the PCB mixture,
from .04 to 2.0 (mg/kg/day)-1. Id., Table 9, at 64. To
calculate the non-cancer risks a different type of toxicity
value--a reference dose, for example--would have to be used,
and certain specified non-cancer risks would have to be taken
into account. Id. at 21, 42.

The second approach endorsed in the Guidance Document
is to use a "total toxicity factor" of 4.0 (mg/kg/day)-1 to
account for cancer and non-cancer risks together. Id. In its
brief the EPA explains that this approach "provides the
applicant an opportunity to reduce the time and expense
associated with the risk assessment" because the Agency is
willing "to accept this 'default' toxicity value of 4.0 (mg/kg/
day)-1[ ] without requiring further justification."

II. Analysis

GE's primary argument is that the Guidance Document is a
legislative rule and therefore should have been promulgated
only after public notice and an opportunity for comment. In

__________
* The Guidance Document initially says it applies to both
s 761.61 and s 761.62. See Guidance Document at 10. Later in
the Guidance Document the EPA appears to use s 761.61 as
shorthand for both provisions. See id. at 21. We follow the
Agency's lead in referring henceforth only to s 761.61.

the alternative it contends that the Guidance Document is not
supported by substantial evidence. Before considering these
arguments about the merits, however, we must determine
whether the case is ripe for review and whether we have
jurisdiction to hear it.

A. Ripeness

To determine whether a controversy is ripe for judicial
review the court must evaluate "the fitness of the issues for
judicial decision and the hardship to the parties of withhold-
ing court consideration." Abbott Labs. v. Gardner, 387 U.S.
136, 149 (1967). "In determining the fitness of an issue for
judicial review we look to see whether the issue is purely
legal, whether consideration of the issue would benefit from a
more concrete setting, and whether the agency's action is
sufficiently final." Clean Air Implementation Project v.
EPA, 150 F.3d 1200, 1204 (D.C. Cir. 1998).

Here the EPA argues that "GE's claims satisfy neither
aspect of the 'fitness'/'hardship' standard under Abbott Labo-
ratories." Regarding fitness, the EPA argues that (1) GE is
asking the court to consider factual questions, such as how
the EPA would evaluate an application that did not use either
of the approaches to toxicity set out in the Guidance Docu-
ment; (2) the Guidance Document is not final agency action
because the Agency "is currently conducting an assessment of
the non-cancer risks of PCBs" and will be modifying the
Document "as needed"; (3) "the Court's consideration would
be aided by further application of the agency's position to
particular facts"; and (4) judicial review is premature because
"adjudication may well prove unnecessary."

We think the issues presented are fully fit for review.
First, whether the Guidance Document is a legislative rule is
largely a legal, not a factual, question, turning as it does in
this case primarily upon the text of the Document. GE does
rely in part upon the Agency's application of the Guidance
Document, but we need not reach that issue; we hold the
Guidance Document is a legislative rule because on its face it
purports to bind both applicants and the Agency with the
force of law.

 Second, it is clear that the Guidance Document is final
agency action because it marks the consummation of the
EPA's decisionmaking process and it determines the rights
and obligations of both applicants and the Agency. See
Bennett v. Spear, 520 U.S. 154, 178 (1997). The EPA argues
that the Guidance Document is not final because it is subject
to change and the "EPA has not completed its decisionmak-
ing process regarding the non-cancer impacts of PCBs." We
rejected a similar argument in Appalachian Power Co. v.
EPA, 208 F.3d 1015 (2000), stating: "The fact that a law may
be altered in the future has nothing to do with whether it is
subject to judicial review at the moment." Id. at 1022. If the
possibility (indeed, the probability) of future revision in fact
could make agency action non-final as a matter of law, then it
would be hard to imagine when any agency rule--and particu-
larly one that must be updated periodically to reflect ad-
vances in science--would ever be final as a matter of law.

In the same vein, the EPA contends that the Fifth Circuit's
decision in Central & South West Services, Inc. v. EPA, 220
F.3d 683, 695 (2000), remanding the Final Rule governing
PCB remediation and decontamination--which Rule the
Agency promulgated using the 4.0 (mg/kg/day)-1 toxicity fac-
tor--was "[i]n effect" a decision that the case was "not ripe
because EPA's position on this complex scientific issue was
not final." But the court there did not purport in the least to
hold GE's petition non-ripe or the Agency's Final Rule non-
final. Id. Rather, the court remanded the issue without
considering the merits of GE's petition because, in view of the
EPA's continuing assessment of the toxicity of PCBs, the
Agency had "no objection to a remand," and that was all the
relief GE was seeking. Id.*

__________
* This is all the Fifth Circuit had to say:

EPA is in the process of conducting a comprehensive assess-
ment of the non-cancer toxic effects of PCBs. According to
EPA, it promulgated the Final Rule before the assessment was
completed, in order to comply with the desires of the regulated
community to finalize the rulemaking as soon as possible.
However, EPA states that it has already committed to reexam-

 Third, we do not think "the Court's consideration would be
aided by further application of the agency's position to partic-
ular facts." We conclude below that the Guidance Document
should not have been issued without public notice and an
opportunity for comment because the Document purports on
its face to bind both applicants and the Agency. In this
situation, nothing would be gained from delaying review.

The EPA's final argument regarding the fitness of the
issues for review is that, if the court does not resolve this case
now, then it may never be necessary to decide the underlying
controversy. This contention rests upon the EPA's assertion
that it will apply the Guidance Document flexibly. That
assertion, however, simply restates a portion of the Agency's
argument that the Document is not binding, an argument we
reject below.

As for hardship, the EPA argues that GE will not be
harmed if the court defers review of the Guidance Document
because GE can later challenge under the APA any decision
of the EPA denying its application for a risk-based alterna-
tive. As we have previously explained, however, "[w]here the
first prong of the [Abbott Laboratories] ripeness test is met
and Congress has emphatically declared a preference for
immediate review ... no purpose is served by proceeding to
the second [or hardship] prong." George E. Warren Corp. v.
EPA, 159 F.3d 616, 622 (1998). In this case the TSCA
requires that any petition for review of a rule be filed within
60 days of the promulgation of the rule. 15 U.S.C.
s 2618(a)(1)(A). Consequently, having demonstrated that the
issue is fit for review, GE need not also show that delaying
review would work a hardship. For these reasons, we hold
the case ripe for review.

__________
ine the toxicity of PCBs and has no objection to a remand so
that it can consider the results of the assessment. Therefore,
we remand ss 761.61(a) and 761.79(b) to give EPA an opportu-
nity to complete its assessment and reconsider the Final Rule
in light of its study.

B. Jurisdiction under the TSCA

Before we can reach the merits, we must consider whether
the Document is a "rule" subject to our review under
s 19(a)(1)(A) of the TSCA, 15 U.S.C. s 2618(a)(1)(A). That
section provides:

Not later than 60 days after the date of the promulgation
of a rule under section ... 2605(e) ... of this title, ...
any person may file a petition for judicial review of such
rule with the United States Court of Appeal for the
District of Columbia Circuit.

GE contends that the term "rule" should be read broadly to
track the definition in the APA. See 5 U.S.C. s 551(4). The
EPA takes the narrower view that "direct appellate review is
limited to legislative rules, i.e., rules which were (or should
have been) promulgated through notice and comment rule-
making." See, e.g., Appalachian Power, 208 F.3d at 1020 &
n.11. We need not decide which interpretation of the term
"rule" in s 19(a)(1)(A) is correct because we conclude that the
Guidance Document is indeed a legislative rule.

GE argues that the Guidance Document is a legislative rule
rather than a statement of policy or an interpretive rule
because it gives substance to the vague language of 40 C.F.R.
s 761.61(c) ("unreasonable risk of injury to health or the
environment"), does so in an obligatory fashion, and is treated
by the EPA as "controlling in the field." See Community
Nutrition Inst. v. Young, 818 F.2d 943, 946 (D.C. Cir. 1987);
McLouth Steel Prods. Corp. v. Thomas, 838 F.2d 1317, 1320-
22 (D.C. Cir. 1988); Appalachian Power, 208 F.3d at 1021.
The EPA argues that under the three-part test applied in
Molycorp, Inc. v. EPA, 197 F.3d 543, 545 (D.C. Cir. 1999), the
Guidance Document is not a legislative rule. Although it is
not entirely clear what in the EPA's view the Document is,
the EPA comes closest to characterizing it as a statement of
policy; thus:

[T]he portion of the guidance at issue here is simply an
expression of EPA's policy judgment, based on the avail-
able scientific data and analysis, that when the "total

 toxicity" analysis is used, the 4.0 (mg/kg/day)-1 toxicity
value is appropriate to avoid an unreasonable risk to
health or the environment.

With the Agency's argument so understood, the question
before us can be framed as whether the Guidance Document
is a legislative rule or a statement of policy.

As GE argues, in cases where we have attempted to draw
the line between legislative rules and statements of policy, we
have considered whether the agency action (1) "impose[s] any
rights and obligations" or (2) "genuinely leaves the agency
and its decisionmakers free to exercise discretion." Commu-
nity Nutrition Inst., 818 F.2d at 946; Chamber of Commerce
v. Dep't of Labor, 174 F.3d 206, 212 (1999). In McLouth, we
recognized that "[i]n practice, there appears some overlap in
the Community Nutrition criteria" because "[i]f a statement
denies the decisionmaker discretion in the area of its cover-
age, so that [the agency] will automatically decline to enter-
tain challenges to the statement's position, then the statement
is binding, and creates rights or obligations." 838 F.2d at
1320. We emphasized that an agency announcement has
"present-day binding effect" if the agency is "simply unready
to hear new argument" in proceedings governed by the
announcement. Id. at 1321.

The EPA urges the court to consider three factors: "(1) the
Agency's own characterization of its action; (2) whether the
action was published in the Federal Register or the Code of
Federal Regulations; and (3) whether the action has binding
effects on private parties or on the agency." Molycorp, Inc.,
197 F.3d at 545; see also Florida Power & Light Co. v. EPA,
145 F.3d 1414, 1418 (D.C. Cir. 1998); American Portland
Cement Alliance v. EPA, 101 F.3d 772, 776 (D.C. Cir. 1996).
As the EPA concedes, however, the third factor is the most
important: "[T]he ultimate focus of the inquiry is whether the
agency action partakes of the fundamental characteristic of a
regulation, i.e., that it has the force of law." Molycorp, Inc.,
197 F.3d at 545.

The two tests overlap at step three of the Molycorp
formulation--in which the court determines whether the

agency action binds private parties or the agency itself with
the "force of law." This common standard has been well-
stated as follows:

If a document expresses a change in substantive law or
policy (that is not an interpretation) which the agency
intends to make binding, or administers with binding
effect, the agency may not rely upon the statutory ex-
emption for policy statements, but must observe the
APA's legislative rulemaking procedures.

Robert A. Anthony, Interpretive Rules, Policy Statements,
Guidances, Manuals, and the Like--Should Federal Agen-
cies Use Them to Bind the Public?, 41 Duke L.J. 1311, 1355
(1992).

Our cases likewise make clear that an agency pronounce-
ment will be considered binding as a practical matter if it
either appears on its face to be binding, Appalachian Power,
208 F.3d at 1023 ("[T]he entire Guidance, from beginning to
end ... reads like a ukase. It commands, it requires, it
orders, it dictates."), or is applied by the agency in a way that
indicates it is binding, McLouth, 838 F.2d at 1321. As
Professor Robert A. Anthony cogently comments, the manda-
tory language of a document alone can be sufficient to render
it binding:

A document will have practical binding effect before it is
actually applied if the affected private parties are reason-
ably led to believe that failure to conform will bring
adverse consequences, such as ... denial of an applica-
tion. If the document is couched in mandatory language,
or in terms indicating that it will be regularly applied, a
binding intent is strongly evidenced. In some circum-
stances, if the language of the document is such that
private parties can rely on it as a norm or safe harbor by
which to shape their actions, it can be binding as a
practical matter.

Interpretive Rules, 41 Duke L.J. at 1328-29.

GE argues that the Guidance Document is binding both
because it facially requires an applicant for a risk-based

variance to calculate toxicity by one of two methods--either
use a total toxicity factor of 4.0 (mg/kg/day)-1 or use a cancer
potency factor and account for the specified non-cancer health
risks--and because, considering the cost, delay, and uncer-
tainty entailed in the latter course, "[f]or all practical pur-
poses, the Guidance is a rule that directs PCB toxicity to be
measured by a 4.0 (mg/kg/day)-1 CPF."

The EPA counters that the Guidance Document lacks the
force of law because it does not purport to be binding and
because it has not been applied as though it were binding.
First, we are told, the Document "allows great flexibility"
because it not only "recognizes two broad approaches to risk
assessment," but also acknowledges (at 44) that

some risk assessments may have components that re-
quire the use of non-standard reference materials, unique
exposure scenarios or assumptions, or require the use of
unconventional methods for estimating risks. These risk
assessments will need to be addressed on a case-by-case
basis.

Second, the EPA says that it has not in practice "applied the
guidance document inflexibly, as if it were a rule or regula-
tion." By this, however, the EPA means only that it has
received and approved applications based upon the use of the
total toxicity factor and upon a separate analysis of cancer
and non-cancer risks--and even this limited assertion is
disputed by GE. Finally, the Agency contends that the
Guidance Document is "an expression of EPA's judgment on
values to be used in conducting risk assessments," much like
the data in the Agency's Integrated Risk Information System
(IRIS), which this court held are not subject to the require-
ments of notice and comment rulemaking. See Chemical
Mfrs. Ass'n v. EPA, 28 F.3d 1259, 1263 (1994).

We think it clear that the Guidance Document does purport
to bind applicants for approval of a risk-based cleanup plan
under 40 C.F.R. s 761.61(c). Consider the principal di-
rectives: "When developing a risk-based cleanup application
... both the cancer and non-cancer endpoints must be ad-
dressed...." Guidance Document at 21. If an applicant

chooses not to use the 4.0 total toxicity factor, then it "must,
at a minimum account for the risk from non-cancer endpoints
for neurotoxicity, reproductive and developmental toxicity,
immune system suppression, liver damage, skin irritation, and
endocrine disruption for each of the commercial mixtures
found at the cleanup site." Id. Although the Guidance
Document does, as noted, anticipate and acknowledge that
"some risk assessments may have components that require
the use of non-standard ... unique ... or unconventional
methods for estimating risk," id. at 44, that does not under-
mine the binding force of the Guidance Document in standard
cases. See McLouth, 838 F.2d at 1321 ("such a provision for
exceptions ... does not push it much in the direction of a
policy statement"). Furthermore, even though the Guidance
Document gives applicants the option of calculating risk in
either of two ways (assuming both are practical) it still
requires them to conform to one or the other, that is, not to
submit an application based upon a third way. And if an
applicant does choose to calculate cancer and non-cancer risks
separately, then it must consider the non-cancer risks speci-
fied in the Guidance Document. To the applicant reading the
Guidance Document the message is clear: in reviewing appli-
cations the Agency will not be open to considering approaches
other than those prescribed in the Document.

The Guidance Document also appears to bind the Agency
to accept applications using a total toxicity factor of 4.0
(mg/kg/day)-1 to calculate the risk from both cancer and non-
cancer endpoints. Guidance Document at 21. The EPA
recognized this in its principal brief: "By indicating that [a]
total toxicity value [of 4.0 (mg/kg/day)-1] will be accepted
without detailed justification, the guidance document offers
an applicant an opportunity to reduce the time and expense
associated with risk assessment." In its supplemental brief,
however, the EPA backs away from this statement, asserting
that the "EPA is not 'bound' to approve an application under
Section 761.61(c) if the applicant uses a total toxicity factor of
4.0 (mg/kg/day)-1, even assuming that the application falls
within the framework of the guidance document." How can
this be? According to the Agency, its position with respect to

the total toxicity factor "is a matter of policy" that "can be
changed at any time to respond to, inter alia, advances in
scientific knowledge." But the Guidance Document itself
says nothing of the sort. Clearly the EPA's initial response
more accurately describes the Agency's approach in the Doc-
ument: Stating without qualification that an applicant may
use a total toxicity factor of 4.0 (mg/kg/day)-1 strongly implies
that use of that value will not be questioned; an applicant
reasonably could rely upon that implication.

The EPA argues that the Guidance Document "neither
adds to EPA's prior position nor imposes any further obli-
gations on EPA or the regulated community" because the
Agency had used the toxicity factor of 4.0 (mg/kg/day)-1 when
it "establish[ed] the generic cleanup standards in the 1998
[PCB] regulations." In its supplemental brief, however, the
Agency explicitly states that it does not think its use of 4.0
(kg/mg/day)-1 in the 1998 regulations requires it to approve
use of that factor in an application under s 761.61(c). Be-
cause we conclude that the Guidance Document does bind the
Agency to accept use of 4.0 (kg/mg/day)-1, it follows that the
Document does indeed impose a "further obligation[ ] on the
EPA." In this way the Guidance Document is not like the
risk data at issue in Chemical Mfrs., which we held "constrain
no one until ... applied in a particular rule." 28 F.3d at
1263.

Furthermore, the EPA does not contend that in practice it
has not treated the Guidance Document as binding in the
ways described above. The EPA does not claim, for example,
that it has accepted any applications that (1) use neither of
the two methods of risk assessment approved in the Guidance
Document; or (2) calculate risk separately for cancer and
non-cancer endpoints, but fail to calculate endpoints for all
the non-cancer risks required by the Guidance Document to
be addressed. Nor does the EPA contend that it has ever
rejected an applicant's use of 4.0 (mg/kg/day)-1. Whether an
applicant has successfully used the second method of risk
assessment set out in the Guidance Document--as the EPA
asserts and GE disputes--is immaterial because, even if both

methods are practically available, the Document nonetheless
binds applicants and the Agency in the ways described above.

In sum, the commands of the Guidance Document indicate
that it has the force of law. On its face the Guidance
Document imposes binding obligations upon applicants to
submit applications that conform to the Document and upon
the Agency not to question an applicant's use of the 4.0
(mg/kg/day)-1 total toxicity factor. This is sufficient to render
it a legislative rule. Furthermore, the Agency's application of
the Document does nothing to demonstrate that the Docu-
ment has any lesser effect in practice. Consequently, we
conclude that the Guidance Document is a legislative rule.
The Guidance Document is therefore undisputedly a "rule"
for purposes of s 19(a)(1)(A) of the TSCA, and the manner of
its promulgation is subject to review.

C. The Merits

The EPA concedes that it did not comply with the proce-
dural requirements of the TSCA and of the APA. More
specifically, as GE points out, it failed to publish a notice of
proposed rulemaking, give interested parties an opportunity
to comment, and hold an informal hearing. See 15 U.S.C.
s 2605(e)(4); 15 U.S.C. s 2605(c)(2); 5 U.S.C. s 553. There-
fore, having held that the case is ripe for review and that the
Guidance Document is a "rule" for purposes of the TSCA, it is
clear that GE must prevail on the merits. The EPA agrees:
"Either the petition must be dismissed for lack of jurisdiction
or the PCB Guidance should be vacated." For this reason we
need not consider GE's alternative argument on the merits,
namely, that the Guidance Document is not supported by
substantial evidence.

III. Conclusion

GE's petition for review is granted because the EPA
promulgated a legislative rule without following the proce-
dures required by the TSCA and the APA. The Guidance
Document is accordingly

Vacated.